Wikipedia talk:Manual of Style/Medicine-related articles

Product pricing

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

---

Should there be a separate section or just the current brief mention? QuackGuru (talk) 17:31, 1 October 2019 (UTC)

• Would support such a section. Definitely an independent source is required. Doc James (talk · contribs · email) 20:05, 1 October 2019 (UTC)

• I think it merits discussion. One issue is that WP articles are usually about the active molecule and not about any product(s) containing that molecule. Another one is that list prices in cases of innovative drugs can be very misleading, as actual, negotiated prices are confidential and frequently carry rebates up to 70% (and more) from the list price. Then, there are also risk-sharing agreements that affect the effective price; pay-per-result arrangements; bundle pricing, etc. etc.

Until we agree on how we present the prices keeping in mind the requirement to be WP:GLOBAL, I suggest to leave out a separate section. Deinitely, adding it after just a few hours of "discussion" between just two participants, even if highly respected as they are, seems premature. — kashmīrī ^TALK 18:38, 2 October 2019 (UTC)

I was trying to find guidance on product pricing and I missed it. I moved one sentence and just added "The pharmaceutical industry has tried to conceal medication prices because of their continuing legal cases in the United States. There is a lack of transparency regarding medication and vaccine prices among non-governmental organizations." Numerous articles have content about pricing. It needs a separate section. If you disagree with the content I added you can just comment it out or delete it for now. QuackGuru (talk) 19:13, 2 October 2019 (UTC)

The prices in the BNF and from medicaid are fairly sound. Sure a bit of work may be required to find appropriate sources and craft the appropriate wording. Doc James (talk · contribs · email) 21:45, 2 October 2019 (UTC)

Again, the world is not just the US - in Europe, prices are not disclosed for an entirely different reason - and also the way you used "pharmaceutical industry" is incorrect: the majority of pharmaceutical companies are manufacturers of generic drugs and dietary supplements (sic!). — kashmīrī ^TALK 22:15, 2 October 2019 (UTC)

Please edit the section to improve the wording or you can just deleted the content I wrote. I'm not sure what the best wording should be. I think the separate section should remain because so many articles discuss pricing. QuackGuru (talk) 01:30, 3 October 2019 (UTC)

I have re-written it. WhatamIdoing (talk) 00:08, 12 October 2019 (UTC)

Thanks and looks good. Doc James (talk · contribs · email) 02:53, 24 October 2019 (UTC)

• I'm not convinced we need a pricing section or that the text there is understandable. Do we generally want prices on our drug articles or treatment sections? I think this section might encourage trivia. It isn't clear from the text that the do-this/don't-do-this follows. It asks lots of questions but the consequence would appear to be original-research. I certainly don't support QuackGuru's text, which was replaced by WhatamIdoing. The unreasonable increase in the cost of insulin is a notable issue in the US only and perhaps properly belongs in some article on drug company pricing. Is this sort of thing relevant in general (which a guideline should cover) or just a special case of including information per WP:WEIGHT. There may be merit in noting that some treatments are popular/uncommon due to pricing issues if that issue is covered by reliable sources. But to be honest, that seems to me to be pretty standard WP editing practice. -- Colin°^Talk 10:03, 12 November 2019 (UTC)
  • I think that the consequence of answering those questions, in most cases, is going to be accurate reporting of what your source says. One of the biggest problems in the foreseeable-and-solvable category is someone getting a good source on third-world wholesale prices and writing "The price is two cents per pill", which some reader is going to interpret as "My prescription will cost 60 (local) cents at the pharmacy". The meaning of those questions is "don't write 'the price'. Instead, write down which price your source is talking about". I do not think there is any scope for OR in that. I think that it militates against OR.
  • As to whether we want prices, we probably do, in some cases. Those cases are IMO mostly WHO essential medicines (for which "The estimated average wholesale price worldwide, according to the WHO, is around two cents per pill" is a not unreasonable answer) and fancy new medicines (for which the answer probably sounds like "The company announced a list price of $120,000K per year in the US, but the revenue per patient is expected to be lower due to negotiated discounts and the patient assistance program", assuming that my [probably business magazine] source mentioned that kind of reality). This information isn't being included for medical reasons. It's being included because some of our readers are interested in money. WhatamIdoing (talk) 15:29, 12 November 2019 (UTC)
    • In what way are they interested in money? Is it "Can I afford to buy this" (paying privately for a drug); "Will my insurance/health-service provide this?"; or just generally interest in which drugs are very very cheap, fairly cheap, expensive, eye-wateringly expensive... At the moment I can see the text provoking some kind of table with all countries and with wholesale, retail and discount prices if such were available. Ibuprofen lead (only) says "The wholesale cost in the developing world is between US$0.01 and US$0.04 per dose. In the United States, it costs about US$0.05 per dose". It isn't covered by the body nor do we list details about generic vs brand pricing. Aspirin also only covers cost in the lead, and only wholesale cost for "developing world". It lists the monthly cost in the US, but per your notes does not say whether that is wholesale, but also doesn't say what that is treating. Valproate is the same. Hmm, information about developing world prices added only to the lead. Hmm, could that possibly be done by a certain editor with a focus on a medical translation project? The prices are of course now 4-5 years old and not maintained. The fact that this is being only added by one editor, only to the lead, not maintained, very specific prices that get out-of-date, all rings alarm bells for whether Wikipedia really is the place for this. Isn't there a better way to cover this, and one that doesn't waste precious space in our lead sections for what is a trivia detail. The "very cheap ... borderline acceptable to NHS" scale may be significant, but whether the wholesale price in the US in 2014 was 4c or 5c really isn't. -- Colin°^Talk 21:44, 12 November 2019 (UTC)
      • The "money" audiences that I happen to care about are public policy (the "very cheap ... borderline acceptable to NHS" scale) and business stuff (e.g., someone trying to figure out why this company's in the news for over charging). If it's easier to write US$0.04 than "very cheap" (which is additionally a value judgment and therefore not something we could say without a source saying that), then that doesn't worry me.
        I'm not overly worried about keeping them up to date (although I mostly wouldn't put prices in the lead). As most of the prices for generics are stable over the years, being strictly up to date isn't terribly important. People aren't going to have radically different understandings of a drug if it turns out that the price has changed from one to four cents to now being two to five cents. For fancy new drugs, what gets the most attention in sources (and therefore what's most WP:DUE) is their announced initial price, which is always going to be its initial price, even though it will likely drop later. WhatamIdoing (talk) 02:31, 13 November 2019 (UTC)
        • Limited guidance on drug pricing was already in WP:PHARMMOS before the section in question was added here. Does it need covering at both MOSs? (My guess is "probably".) If not, which should be the one to talk about drug pricing?
          Having guidance on pricing might prove useful when discussing medical tourism, and the reasons why grey and black markets exist in medical care. Little pob (talk) 09:42, 13 November 2019 (UTC)
          • Although WP:PHARMMOS says "Economics: Global sales, distribution, cost in major English speaking countries, etc" it doesn't really offer guidance. I see now valproate does mention the cost later in a Cost section. I think "Econonomics" is a better title, which would encourage other economic data such as sales (particularly for blockbuster drugs that influence company shares) and whether the drug has been rejected on economic grounds (such as NICE rejection for drugs too expensive). I don't think there is any WP:WEIGHT/MOS:LEADREL argument to include specific cost details in the lead at all, though the other economic factors may be relevant for a small set of drugs. The cost-per-dose is not widely covered by reliable sources on drugs or medical treatments, largely because it is so variable per country and also over time. I think it is very difficult to justify including "The wholesale cost in the developing world is between US$0.01 and US$0.04 per dose.[13] In the United States, it costs about US$0.05 per dose." in the lead. This is noisy trivia which in a reliable source would merely get "this very cheap drug" or some such one/two word adjective. If we lack reliable sources to turn price details into an adjective, then probably better to simply remove from the lead on the grounds that nobody is commenting on the price as being something worth noting. I suspect it has been added to the lead because the medical translation project only translates the lead and the information was felt relevant to developing countries. I reject that argument, if that is the case, on the grounds that en:wp is here for its readers and not to be some source-code for other wp -- the translators can always be advised to include cost details from the body if that is felt relevant on other WPs. -- Colin°^Talk 10:38, 13 November 2019 (UTC)

Proposal

Add the following to the "product pricing" section: "Detailed cost information does not belong in the lead, though some significant economic points may be relevant for some drugs (blockbusters, drugs rejected as too expensive, etc)." The "product pricing" section be renamed "Economics" to fit with the earlier heading examples.

The current wording could be shortened to "The cost of medicine or procedures should include specific details regarding the kind of price (e.g., wholesale, retail, discounted), scope (e.g., US-specific, worldwide average) and the year." I don't think it worth including "if covered by reliable sources" because that is true of everything. I don't think the "indirect costs such as lost wages for the patient" is relevant to product pricing. This seems more relevant to disease articles, covering time off due to to sickness or to attend hospital for treatment or tests. -- Colin°^Talk 10:45, 13 November 2019 (UTC)

• Strong support. I am dead against having drug prices in the lead section and only mention them in the article body where it is specifically discussed in reliable sources as a matter of public interest (like Zolgensma being the most expensive drug in the world). — kashmīrī ^TALK 15:27, 13 November 2019 (UTC)

I've rewritten the section mostly per the proposal above. I agree with Kashmiri that it is hard to justify the routine addition of specific pricing details even in the article body. We have perhaps a difficulty translating raw figures into a more appropriate adjective that our sources would do. Few of our sources mention specific costs, so WP:WEIGHT makes it hard to argue for. There are of course drugs whose price is especially notable.

I've written the lead advice as: "Detailed cost information is rarely appropriate for inclusion in the lead." To be honest, I feel somewhat that we are having to include this guidance merely because one editor chose to ignore our basic rules for WP:LEAD: that they summarise article bodies. Any such prices in the lead could be moved/removed simply per WP:LEAD. -- Colin°^Talk 13:58, 14 November 2019 (UTC)

It is irrelevant you rewritten the content. That's not the consensus across med articles. QuackGuru (talk) 14:06, 14 November 2019 (UTC)

QuackGuru articles do not have "consensus"; editors do. That is why we are discussing this here. You tried to add the text "Cost information may be included in the lead, but this is done on a case-by-case basis.", this was reverted, and you restored it again without attempting to achieve consensus. That's edit warring, which will get you blocked. I see from your earlier edit to include "The pharmaceutical industry has tried to conceal medication prices because of their continuing legal cases in the United States. There is a lack of transparency regarding medication and vaccine prices among non-governmental organizations." that you have an agenda. That's even more reason for you to avoid editing or warring over guidelines. The statement you added about the lead is meaningless because all information may be included on a case-by-case basis. Further the specific issue is "detailed cost information" whereas in our reliable sources, a lead section/page would merely note if a drug is cheap or expensive: our reliable sources know how to summarise.

MEDMOS cannot supersede WP:LEAD. Let's examine this consensus of editors. I looked at the first bunch of anti-infective medicines on the WHO list:

• Albendazole added by Doc James.
• Ivermectin added by Doc James.
• Levamisole added by Doc James.
• Mebendazole added by Doc James.
• Niclosamide added by Doc James.
• Praziquantel added by Doc James.

So this is a practice followed by exactly one editor, who also has an agenda about information in the lead.

The information doesn't even seem to be correct. Albendazole says "The wholesale cost in the developing world is between 0.01 and 0.06 USD per dose". The 2014 source gives the dose at 400mg and the table strength at 200mg and from $0.0115 to $0.0341 per table (i.e., from $0.023 to 0.0682, which would be rounded to "0.02 and 0.07 USD per dose"). By the following year the highest price is only 0.04 USD, so already out-of-date. Who knows what 2018 or 2019 is. The article claims "in the developing world" though I can't see where this is is indicated in the source. Their own sources of buyers and suppliers does seem to be focused on such nations, but I don't know if that is representative of the developing world. The website does not seem to include prices any newer than 2015, which is a strong concern. For the US price, the source given is a database which is not an allowed source. Readers need to search through the database for the drug then average the individual records themselves.

I see this issue was discussed Wikipedia talk:WikiProject Medicine/Archive 84#Price of medications and concluded "Except in the cases where the sources note the significance of the pricing (which did have consensus), there is no consensus to add the pricing to the articles". Therefore it seems that Doc James and QuackGuru are editing against consensus. It's going to take me a while to read through all the discussions, but once again I see a few folk at WP:MED having ideas the rest of Wikipedia do not support. We don't include the prices of potatoes, televisions or package holidays to Greece. It seems fairly obvious that specific price information is not the role of an encyclopaedia for the general reader. We have a "source" for wholesale prices that appears to be no longer maintained as of 2015. This "source" material is raw data, not meaningful data or commentary that could be used in a lead summary. If we have no consensus to add pricing to articles, except for significant notable cases, then there is very much no consensus to add this to the lead. I request QuackGuru reverts themselves. -- Colin°^Talk 18:44, 14 November 2019 (UTC)

I see from Doc James talk page that this issue has become a recent hot topic. James edit warred four times with two editors. He also claimed that the RFC requirement "where the sources note the significance of the pricing" was met merely by a US Database including the drug price among its records. That is a strange way of assessing WP:WEIGHT. And a very deep rejection of what constitutes a "source" -- a pointer to the first page of database results showing 25,095 records. James writes "We know that the pharmaceutical industry is trying really hard to hide medication prices with ongoing legal cases in the United States. Many NGOs including Doctors Without Borders and UNICEF struggle with the lack of transparency around medication and vaccine prices. WP:NOTCENSORED applies here." I am reminded what a wise editor once wrote about people citing NOTCENSORED as a justification for including material: "Anyone who defends their edits by citing WP:NOTCENSORED doesn't have the first clue." The key requirement made by WhatAmIDoing in 2010 was "cost is discussed by significant sources". A database is not a discussion. A number in the BNF (a drug database) is not a discussion.

I think we should summarise and clarify (for those who think a database is a source) the RFC conclusion as "Except in the cases where reliable sources discuss the significance of drug pricing, drug price information is not appropriate in articles. In such cases where price information is relevant, the article should indicate why. The lead, as a summary of article content, should not include specific detailed price information." -- Colin°^Talk 19:15, 14 November 2019 (UTC)

It appears that QuackGuru does not wish to justify their edits or explain why the consensus at the RFC Wikipedia talk:WikiProject Medicine/Archive 84#Price of medications is being ignored. I'm also guessing that User:DocJames is ignoring the discussion because the current text supports his position. It seems very clear from the RFC that James does not understand that WP:WEIGHT is not met by the inclusion of a figure in a database. I can get multiple sources for the bus timetable and journey times, yet this raw data is not encyclopaedic. In their discussion with User:Rhododendrites, James also claims that a wide interest in drug prices (particularly those on the WHO essential list) is sufficient to include detailed wholesale pricing for all such medicines. But this is the difference between raw data and encyclopaedic information. The encyclopaedic issue that people are discussing is whether and where the price is low or high, why the price is enormously higher in some countries (e.g. US) and whether price has affected availability of the treatment to patients (e.g., NICE does not consider it's cost/benefit justified). This is the sort of information our articles might include, case by case. Most of the price sources suggested are either now no longer maintained, or incomplete or are merely a link to a database result-set of 25,000 records, none of which meet the standards required for sourcing. User:WhatamIdoing reminded us in 2010 that "cost is discussed by significant sources" is the requirement.

I think that unless those who lost the argument in the RFC desist from edit warring on this page, we may need some administrative action here or another community RFC. I'll reword my text from just above, since it seems some think general discussion of drug pricing justifies absolute inclusion of drug pricing everywhere. Here's what I think the RFC demands MEDMOS says:

"Except in the cases where reliable sources discuss the significance of the price of a particular drug, drug price information is not appropriate in an article. In such cases where price information is relevant, the article should explain why. The lead, as a summary of article content, should not include specific detailed price information."

-- Colin°^Talk 12:34, 16 November 2019 (UTC)

You pinged the incorrect user name. The sources being used qualify per WP:MEDRS. Lots of sources that discuss the price in greater detail are avaliable. There is an in depth discussion by WHO for example for each medication listed as essential. Sure I could write an entire section on this for every medication. Wikipedia is also a work in progress. Doc James (talk · contribs · email) 18:47, 21 November 2019 (UTC)

The 2016 RFC concluded, “Except in the cases where the sources note the significance of the pricing (which did have consensus), there is no consensus to add the pricing to the articles.”

Colin, could you have a look at Epipen as an example? Both of your versions are truer to the RFC and more helpful than QuackGuru’s version, which I oppose, but what a useless mess of information at Epipen, presented in an unhelpful fashion, while the pricing controversy is not summarized to the lead at all. Your phrase “... should not include specific detailed price information ...” is still warranted, as the Epipen pricing controversy could be summarized to the lead as a due weight portion of the article without getting into detailed and cumbersome pricing exammples. For an example like Epipen, where there was a pricing controversy, I prefer your 12:34 16 Nov proposal above to either of these.. QuackGuru’s detailed examples are just a repeat of basic editing practice, but are apparently misleading some editors to go against the RFC. Could you bold your new proposal please? SandyGeorgia (Talk) 14:58, 16 November 2019 (UTC)

I also prefer parts of what Colin wrote to what I wrote, especially his list of factors ["include specific details regarding the kind of price (e.g., wholesale, retail, discounted), scope (e.g., US-specific, worldwide average)"], rather than the questions I wrote. I think that including an example (although the specific example likely could be improved) will be useful to people whose writing skills are less developed than Colin's and SandyGeorgia's (and that's probably about 99% of us). WhatamIdoing (talk) 04:47, 19 November 2019 (UTC)

• Add prices I have commented at many of these drug price discussions over the years. Prices are necessary context for understanding drugs and treatment. Consumers, physicians, journalists, and policy makers all want this information. Challenges for Wikipedia are that (1) prices vary by time, place, and by medical insurance (2) there are international price catalogs, but these are challenging for individual readers to interpret and (3) Wikipedia still does not have good policy of when to include or exclude primary information, and pricing has to come from primary sources. The information that I really want is context of the order of magnitude. Wikipedia articles should clearly communicate whether the price of a dose of a drug is closer to US$0.001, 0.01, 0.10, 1.00, 10, 100, or 1000. If the reader is left thinking that a $100 dose drug is generally affordable, or that a very inexpensive drug is inaccessible, then that reader is lacking a fundamental understanding of the nature of the drug. The most reasonable proposal we have for communicating order of magnitude pricing is to import and publish some price catalog for everything. Blue Rasberry (talk) 17:28, 16 November 2019 (UTC)

What I wrote originally was:

"The cost of medicine or procedures should include specific details regarding the kind of price (e.g., wholesale, retail, discounted), scope (e.g., US-specific, worldwide average) and the year. Some medicines contribute significantly to a pharmaceutical company's turnover (blockbuster drugs), or their high price is a factor in their rejection or difficult acceptance by state health services or insurance companies. Detailed cost information is rarely appropriate for inclusion in the lead."

Upon reading the RFC and thinking more about what a lead should be, I proposed:

"Except in the cases where reliable sources discuss the significance of the price of a particular drug, drug price information is not appropriate in an article. In such cases where price information is relevant, the article should explain why. The lead, as a summary of article content, should not include specific detailed price information."

There is perhaps merit in merging the two. I agree Epipen is a mess and a sort of OR. Same with the WHO essential drug leads that James has added prices to. Here we have the leads of drug articles juxtaposing "developing world" and "US" prices, making some point but lacking a source that makes any point. Bluerasberry, I'm afraid I disagree with some of your points and don't think this is the place to rerun the RFC discussion. Wikipedia does have policies against raw data being dumped in articles and does have guidelines about leads. While price of drugs clearly is a concern, if our sources aren't making commentary on international drug price of drug XXX, then our articles also should not include that information. This is just basic WP:WEIGHT stuff. We are only writing this price information so badly because of using bad sources (online databases containing multiple prices from multiple sources) and the lack of skill of the writer. It smells of the kind of Wikipedia editing where folk just add random factoids and data without thinking about how to present information to the reader and where to put it in the article. Currently we have QuackGuru edit warring and just chucking out all of what I wrote and inserting his own agenda. There doesn't appear to be anyone who likes his edits, but nobody fixing it. I don't edit war, and right now I have a few real-world issues on my plate. So I hope you can together try to agree on a form of words that respects that RFC and our WP:WEIGHT and WP:LEAD guidelines. If you don't like the outcome of the RFC then it needs to be raised again for the whole community to comment, not just the handful with this page watchlisted. -- Colin°^Talk 13:11, 19 November 2019 (UTC)

When faced with edit warring, I am disinclined to revert the damage, but support anyone who attempts to restore to RFC-based consensus wording. Also, I will be traveling over the US holiday and unlikely to follow this for a while, so hope someone addresses the issue. I, too, disagree with Bluerasberry's points. SandyGeorgia (Talk) 14:54, 19 November 2019 (UTC)

@Colin: I agree with you on all points about the difficulty of adding price: we lack quality data, the data requires intense regular updating at massive scale, our diverse audience will have challenges interpreting data, and we in Wikipedia lack the social norms, technical infrastructure, and labor base to manage this to our normal standards of quality. In so many ways you have said all these things and your case is well made.

Somehow, this is the development path I want:

1. We have discussion where somehow, we place a value on including price information. I value general price information highly, and roughly, would like to place all drugs into categories of "nearly free, accessible by most, moderate price, expensive, very expensive", or some such scale. I care less about particular prices and more about informing readers about cost barriers. Other people may have another goal in mind, such as exact price information, but I care more about interpreting price for general readers than having exact financial data at points in time and place. The point here is to develop consensus about what is appropriate to include in infoboxes.
2. We scope the ideal price system in Wikipedia, which we would have if we had all data, technology, etc.
3. We identify whomever has the best drug price communication system in the world, and compare ourselves to them, and not necessarily to that ideal pricing system
4. Based on the existing global norms of whom communicates pricing and how, we compromise from our ideal, and start experiments in communicating pricing in some circumstances with intent to develop our price communication policy. All wiki practices develop over years, and if we will communicate price in 10 years, the time to experiment is now.

Colin, this conversation could go lots of ways. Which way seems most productive to you? What do you see as the long term ending consensus in this? Blue Rasberry (talk) 14:58, 19 November 2019 (UTC)

Recognizing and understanding your wants/wishes/desires as good faith, nowhere in your list of wishes do you mention the policies and guidelines Wikipedia has in place that cover content. Colin has named several of those policies and guidelines ... I can name more (WP:NOT, WP:RS, WP:WEIGHT, WP:V, WP:LEAD, etc.). The best way forward (in this, and in any discussion) is to focus not on wants and wishes, but actual policies and guidelines governing content. Colin's position is rooted in policy and guideline. I gave Epipen as an example hopefully to help focus on the policy-driven addition of pricing information. When MEDRS sources discuss pricing, we can include it, to the extent we also respect WP:WEIGHT, WP:NOT and WP:LEAD. Getting ahead of Wikipedia-wide policy and guideline via "the time to experiment is now" would be folly for WPMED. SandyGeorgia (Talk) 15:30, 19 November 2019 (UTC)

Bluerasberry I appreciate your list of desirable information. It represents an aim that a publisher of original material could target if they hire the right people and pay for access or research. But on Wikipedia we are led by sources. We have seen the rather odd interpretation of WP:WEIGHT and 'reliable sources' to mean that just because someone has a database of arbitrarily-sourced prices from a handful of developing countries in 2014, or a database of US wholesale prices in any given week, that means this information is warranted in not only articles but also the lead. We can only give price information the weight that reliable sources do, when discussing price for that article. A database cannot 'discuss' and has no weight. I'm not sure that infoboxes will work for this, nor does that meet the RFC consensus. The price of a drug is too nuanced and varied to just add 'Low' to an infobox. This is a wiki, however, so it is possible that in future we do identify a good source or measure of weight for this. We currently don't appear to have one. --Colin°^Talk 10:24, 21 November 2019 (UTC)

@SandyGeorgia and Colin: I agree with both of what you are saying about the present. Colin expresses that he understands me by saying that we could have reliable drug information from databases in the future, and in the future we might have our choice of expressing that information in any way we like. My objective in this discussion is to seek comments on what realistic expectations anyone has for including prices in Wikipedia if we have the right content quality and technology. I take for granted that consumers need price information, such as what Consumer Reports and similar publish, and that Wikipedia / Wikidata are going to have this information. If an argument needs to be made that reliable sources publish prices and expect consumers to have access to them, then it is possible to make a strong case for that. We are talking about drug prices and the pharma industry here, so it is plausible to imagine that there could be a multi-million dollar investment in Wikipedia just to sort prices any way that we imagined was appropriate if we had the best of everything available to us. I am not suggesting a price roll out everywhere now, but I want experiments and policy development now, and I want to talk through what we would do and how we would do it if we could be the best at this. 10 years ago Wikipedia had to generally dismiss database content, but with the advent of AI publishing / web 3.0 / Semantic Web, publishing is getting weird fast, and algorithmic interpretation is becoming routine in reliable publishing. I see no harm in small controlled live experiments and for anyone who wishes to speculate to develop policy which does not fit now, but which could in the future. Blue Rasberry (talk) 13:44, 23 November 2019 (UTC)

Here's my attempt at merging what we have currently and Colin's two paragraphs above: Unless reliable sources discuss the significance of the price of a particular drug or procedure, it is not appropriate to include that price information in an article. In articles where price information is relevant, the article must explain why it is being mentioned (e.g. claims of price gouging). Some medicines contribute significantly to a pharmaceutical company's turnover (blockbuster drugs), or their high price is a factor in their rejection or difficult acceptance by health services or insurance companies. Prices of medications and procedures can vary significantly between countries, for different brand names, and other factors. So include the specific details regarding the kind of price (e.g., wholesale, retail, discounted), scope (e.g., US-specific, worldwide average) and the year. The lead, as a summary of article content, should not include specific or detailed price information. Little pob (talk) 13:46, 20 November 2019 (UTC)

WhatamIdoing, Bluerasberry, SandyGeorgia, what do you think of Little pob's proposal? I think it is a good mix and also adds 'price gouging' to 'blockbuster' and the health-service/insurance rejection issues. So we have some pointers as to what kind of commentary editors should look for when thinking about price. If we have a consensus then we can move forward to updating the guideline. Currently QuackGuru is refusing to engage in discussion and citing 'competence is required' in his edits, which is a clear insult towards us. Let's demonstrate how collaborative editors work. --Colin°^Talk 10:24, 21 November 2019 (UTC)

On iphone at airport, briefly agree with Little pob but first sentence merely repeats policy ... must we restate policy to get the point across? I guess so ... but it seems strange to me that we even need to say any of this ... it is policy. But I defer so we can move forward. SandyGeorgia (Talk) 12:09, 21 November 2019 (UTC)

PS, prices in infobox, never, adamant oppose. SandyGeorgia (Talk) 12:11, 21 November 2019 (UTC)

I like much of what Little pob wrote.

User:SandyGeorgia, it is sometimes useful to repeat policy.

The main point that concerns me this morning is that I'm not entirely certain that "Unless reliable sources discuss the significance of the price of a particular drug or procedure" is something that holds as strong a consensus as it did during the last major discussion, which was nearly a decade ago. Since that time, there has been so much discussion about the cost of drugs in general, or the cost of drugs in terms of categories (e.g., the new anti-hypertensives are so x% more expensive than the old anti-hypertensives) that I think editors might think that drug costs are more worth mentioning.

They might also hold a different opinion about the costs of specific procedures. These can be quite interesting and discussed in detail by reliable sources (e.g., ultrasounds are much cheaper in Japan than in the US; MRI prices can vary by an order of magnitude within the same metropolitan area in the US), but there is much less emphasis on these. (Also, prices aren't Wikipedia:Biomedical information, so a business magazine is a reliable source for most price information.)

The sentence is also incomplete, because it doesn't mention disease costs. The economic costs of diseases is a routine calculation in the medical literature (and reflected in headlines about "Depression costs the national economy billions of dollars per year"), but that's an easy fix.

I think we might need to go back to editors and ask them where they fall among a range of options. I don't think we're going to get a response of "never, even if extraordinarily well-sourced" or "always, assuming you have any half-decent source at all", but I'm uncertain whether their answer will be closer to "usually yes" or "usually no" these days. To put it another way, encyclopedia articles need to cover the topic comprehensively (the best we can, which is constrained by the existence of sources), and I don't know whether editors currently believe that having some information about the price would be necessary for a comprehensive understanding of the topic. WhatamIdoing (talk) 18:24, 21 November 2019 (UTC)

Yes, I agree with the little pob text here. There are more issues to sort - this little pob text is for prose in Wikipedia, and the biggest part of this price discussion is about the circumstances under which we put prices in the infobox. That infobox discussion is another issue, and while we use default policy for infoboxes usually now, I am expecting another set of policy for infoboxes in the future and in development now. Blue Rasberry (talk) 13:27, 23 November 2019 (UTC)

Thanks for comments. I've taken on board what has been said, and hopefully addressed the majority of the points:

second pass

When adding price information to medical articles, it is important to establish the reason why the price is being stated. Some examples of when it might be appropriate to mention cost information in an article include: The drug is on the WHO Model List of Essential Medicines There have been claims of price gouging or anti-competitive practices Has blockbuster drug status, or is a high-value medication that contributes greatly to a pharmaceutical company's turnover The high price of a drug, or procedure, was a factor in it being rejected or received difficult acceptance by a health service or insurance company Prices of medications and procedures can vary significantly; so include the specific details regarding the kind of price (e.g., wholesale, retail, discounted), scope (e.g., US-specific, worldwide average) and the year: Y Do this: In the US, between 1987 and 2017, the wholesale price of long-acting insulin increased from US$170 to $1,400 per vial. N Don't do this: Insulin costs US$1,400. The lead, as a summary of article content, should not include specific or detailed price information.

I switched the examples to a list layout for easier reading and stylistic consistency with the rest of the MOS. Copied in the do and don't from current, as I think they're useful. I'm not 100% sold on that last sentence, but I'm not sure why. Might just be where I've placed it, or it could be that it's written as an absolute yet there may well be exceptions needed. Little pob (talk) 12:59, 23 November 2019 (UTC)

See "When adding price information to medical articles, it is important to establish the reason why the price is being stated." I don't get this. This suggests if I can't establish the reason why the price is being stated then it should not be included.

See "The lead, as a summary of article content, should not include specific or detailed price information." Prices are already in the lede in numerous med articles. The Current wording does not have these obstacles in place. Please stop trying to add content to this page that would result in deleting pricing from the lede.

Including the examples is fine. QuackGuru (talk) 13:45, 23 November 2019 (UTC)

@QuackGuru: This suggests if I can't establish the reason why the price is being stated then it should not be included. Precisely! If any editor can't justify the reason for adding content – i.e. give the reliable sources without interpretation – they shouldn't be adding that content. It should go without saying, but for clarity; I'm in no way trying to say that same content can't be added by an editor who does have the sources.

Prices are already in the lede in numerous med articles. That doesn't mean they should be in there. Several policies and guidelines have been listed, by much more competent editors than I, as to why there should not be detailed price information in the lead. It's also important to understand that the inclusion of "specific or detailed" is significant, used deliberately, and shouldn't be ignored. <aside>I think it's the fact that the sentence changes meaning if you do, accidentally, skim over the "specific or detailed" bit that has me unhappy with it. Thank you, QuackGuru, for helping clarify that thought process.</aside>

Please stop trying to add content to this page that would result in deleting pricing from the lede. As this is the talk page, no. Discussion around having prices in the lead is ongoing. Some editors are not seeing an issue. Some are strongly against prices in the lead (and apparently infoboxes). Some, myself included, think there is a middle ground to be found. I don't know what that middle ground is, wouldn't dare to guess what numbers each viewpoint has, and nor would I ask those whose opinion differs from mine to hold their tongue. That's not how we find consensus. Little pob (talk) 15:13, 23 November 2019 (UTC)

Proposing content that puts stiff barriers in place is not a middle ground. QuackGuru (talk) 13:31, 27 November 2019 (UTC)

Couple of points, as I know we've butted-heads previously elsewhere:

Re "stiff barriers" specifically – This might come back to bite me on the arse; but WP:IAR would still apply regardless of the wording we all settle on.

Re "the whole": Middle ground is found through discourse. We have to allow all concerned editors to set out their stall, and be willing to engage with them, to find that fabled land.

Little pob (talk) 15:13, 27 November 2019 (UTC)

• Add prices per Blue. Prices in the infobox would be excellent, especially if we can automatically upload them based on high quality sources. User:Seppi333 was working on something like this. Doc James (talk · contribs · email) 18:27, 21 November 2019 (UTC)
• add prices per Blue Rasberry--Ozzie10aaaa (talk) 19:08, 21 November 2019 (UTC)
• Add prices is the way forward. Adding specific wording to this guideline to delete prices from all med articles is not going to work. QuackGuru (talk) 20:12, 21 November 2019 (UTC)
  • Could we please stop voting? Really, this situation is lot more complicated than just "add prices". You can "just add prices" to articles about some subjects, but can any of you premature voters tell me what the price is for a pregnancy?
    The conversation that we need to have here is (a) a conversation, not a vote, and (b) about all the complicated and diverse details, not just some one-size-fits-none "add prices and quit bothering me about the details" opinion. I don't necessarily mind if you want to add prices, but you need to tell me which prices you want to add, and which articles you want to add them to. WhatamIdoing (talk) 21:43, 21 November 2019 (UTC)
    • You are asking what is the typical "cost" of an inhospital delivery in various countries? Average in the US in 2013 was $32,000 but it varies fairly widely.[1] Doc James (talk · contribs · email) 22:34, 21 November 2019 (UTC)
      • No, that's the price of hospital-based childbirth and postpartum care in the US. If it's so easy to "just add prices" that we don't need to have an actual conversation about which prices to add, which articles benefit from prices, and what kinds of situations we should be including and excluding, then you should be able to tell me what the price of the pregnancy itself is. There's millions of pregnancies each year. You want to "add prices". Okay, what price would you add to Pregnancy to explain the costs of getting pregnant? WhatamIdoing (talk) 22:50, 21 November 2019 (UTC)
        • One is directly by the avaliable sources. There is the field of health economics that revolves around this. Would need to look at the sources. We have an entire article called Price of oil. Doc James (talk · contribs · email) 00:16, 22 November 2019 (UTC)
          • Okay, so that's something we should be talking about: "Add prices", but only if and when we've got good sources. But it's more complicated than that, isn't it? The price of pregnancy can be free. It can be hundreds or thousands or tens or even hundreds of thousands of dollars, if the couple is infertile. The price can also be measured in terms of morbidity and mortality for the pregnant woman: the price is weeks of vomiting, or months of depression, or years of worsened body dysmorphia, not to mention the number of women who die by suicide or murder as a result of an unintended pregnancy. And then there are straight economic costs, of lost wages and lost jobs and increased expenses for everything from bigger clothes to more food to prenatal healthcare costs (if any is obtained). So you want to "add prices". Pregnancy is well-studied; we can get sources for any of these. Which of these prices do you actually want to add? WhatamIdoing (talk) 00:36, 22 November 2019 (UTC)
            • Most stuff is much more well defined like the average price of a C-section versus in hospital vaginal birth or the price of a hip replacement or the price of a single medication. Doc James (talk · contribs · email) 01:59, 22 November 2019 (UTC)
              • Does that mean that you want to "add prices" only to these more discrete items, and not to "just add prices" to every article? Do you want to add only certain kinds of prices, like the price of the surgical procedure or the hip prosthesis, or do you want to include the price of the pain, medical equipment, caregiving, physical therapy, lost wages, etc. that go along with that hip replacement surgery? WhatamIdoing (talk) 02:28, 22 November 2019 (UTC)
                • • Include the "price of pain"? Not sure if you have sources for that but I do not think I do. Price of lost wages is avaliable for some conditions as is the price of caregiving. Doc James (talk · contribs · email) 21:25, 22 November 2019 (UTC)

@WhatamIdoing: Some years ago I looked for and could not find "price of pain" or total cost of medical care info.

• Wikipedia:Articles for deletion/Medical harm
• Undesirable health care outcomes
• Talk:Iatrogenesis#Requested_move_Iatrogenesis_→_Medical_harm

I still think we need a general article on non-financial costs. I am not aware of any source which presents this concept. Blue Rasberry (talk) 13:19, 23 November 2019 (UTC)

Perhaps you had the wrong key words for your search? If you are trying to put a dollar value on the experience, I undestand that you measure the pain in DALYs and multiply that by whatever number your economist tells you is the relevant price for a year of (healthy) potential life lost. You can also state the cost directly: "Most people experience moderate to severe levels of pain for n weeks after the surgery". This may be preferable, just like "most people can't work for n weeks" is usually more informative to individuals (especially individuals outside the US) than "People lose US $X in wages". WhatamIdoing (talk) 02:59, 24 November 2019 (UTC)

We include a lot of DALY data. Agree for surgery the typical length of pain is more useful when looking at that measure. The average cost or range of costs for a procedure is also useful. Doc James (talk · contribs · email) 14:47, 26 November 2019 (UTC)

Informative to some (e.g., Americans), but less to others (e.g., anyone in the developed world). "Two-hour surgery that can usually be handled at an outpatient surgical center" translates across time and place better than "Cost US$12,000 in the US in 2015". The same isn't really true for mass-produced products (drugs, supplies, devices, etc.), although if we wanted to avoid numbers altogether, it'd be possible to place most products in some sort of price range ("higher price than most similar devices"). WhatamIdoing (talk) 15:54, 26 November 2019 (UTC)

What's wrong

No editor has explained what is the problem with the current wording and if an editor wants to add content that appears to violate WP:CREEP then that editor should explain why they want to violate CREEP. Telling editors what should or should not be included in every med article is not productive. For example, Onasemnogene abeparvovec says "It carries a list price of US$2.125 million per treatment, making it the most expensive medication in the world as of 2019.[6] In its first four months of sales US$160 million of medical was sold.[7]" We should not add content that would delete this from the article for no good reason. If anyone wants to add content to this guideline that would delete the US$2.125 million price tag from the article then that would be a violation of CREEP. QuackGuru (talk) 14:02, 21 November 2019 (UTC)

Quack, you are not actually entitled to an explanation that you can understand or that will WP:SATISFY you. At some level, I want you to remember that Wikipedia:Policy writing is hard and that it's a skillset that most editors, including you, don't have to any significant degree. But since you have asked, an incomplete list of the problems includes:

• Redundancy: Some versions say the same thing twice. For example (NB: I'm only giving one example of this problem. This is not an exhaustive list of all instances of redundancy), the current version tells editors to mention the "scope (e.g., US-specific, worldwide average)" and then asks them to consider "Is it in a particular country, or a worldwide average?" Saying this twice is pointless.
• Incompleteness: Some versions do not include all of the information that editors think would be helpful. For example (NB: again, just one example), some versions do not mention whether (or under which circumstances) a price should be mentioned in the lead.
• Divergence from consensus: Some versions do not align with the current consensus in all points. In some cases, I am even uncertain what editors would recommend. For example (NB: just one example), perhaps editors have changed their minds, and now they think that the price of drugs is worth mentioning in a greater proportion of articles than they thought several years ago.

The bottom line is that even though these problems are (apparently) not obvious to you, they do exist, and they are worth fixing. WhatamIdoing (talk) 18:08, 21 November 2019 (UTC)

Current wording

Economics may include prices of medications or the cost of procedures, if covered by reliable sources. Cost information may be included in the lead, but this is done on a case-by-case basis. The cost of medicine or procedures can include specific details regarding the kind of price (e.g., wholesale, retail, discounted), scope (e.g., US-specific, worldwide average) and the date or year. Certain medicines contribute significantly to a pharmaceutical company's turnover (blockbuster drugs), or their high price is a factor in their rejection or difficult acceptance by state health services or insurance companies. Prices may vary significantly between countries, for different brand names, and other factors, so include specific information. For example, is your source reporting the list price or the actual selling price? Is it in a particular country, or a worldwide average? Is it the total cost, including indirect costs such as lost wages for the patient, or the initial price?

Y Do this: In the US, between 1987 and 2017, the wholesale price of long-acting insulin increased from US$170 to $1,400 per vial.

N Don't do this: Insulin costs US$1,400.

• Support current wording and current section name. What is the problem with the current wording? If it ain't broken then don't fix it. QuackGuru (talk) 14:02, 21 November 2019 (UTC)

• Comment: please guys let's do a proper RfC, a local consensus cannot overturn a previous RfC, so this vote is bound to be pointless anyway. Please let's revert back to the RfC state and follow a proper BRD procedure, the MED community can do better than that. --Signimu (talk) 15:32, 21 November 2019 (UTC)
  • WP:RFCs do not result in binding decisions. An RFC is an advertising process for an otherwise normal talk-page discussion. Wikipedia:Consensus can change (that's the actual policy; neither WP:RFC nor WP:BRD are policies), and if the consensus disagrees with the outcome of a previous RFC-advertised discussion, then any convincing discussion is adequate to "overturn" it – just like any other discussion. WhatamIdoing (talk) 17:18, 21 November 2019 (UTC)

    Ah just saw your reply, we discussed this on my talk page, but just to write it down here too, yes consensus can change, but it should be at least on a similar level of "strongness" to avoid relying on a WP:SHAM consensus. It's just that I'm sure we can reach a better consensus by discussing more, and more broadly Signimu (talk) 19:34, 27 November 2019 (UTC)

• This isn't the best we can do. Let's keep talking above about how to improve it, rather than just voting. WhatamIdoing (talk) 17:52, 21 November 2019 (UTC)

• Looks like a good start. Doc James (talk · contribs · email) 21:27, 22 November 2019 (UTC)

It is complicated and you are forgetting WP:WEIGHT

I'm really disappointed at the level of argument from QuackGuru and DocJames. QuackGuru insists the "current wording" be left because "If it ain't broken then don't fix it". We didn't have anything on Price a month ago, but it is worth reminding ourselves of what QuackGuru added:

"Economics may include prices of medications or cost of procedures if covered by reliable sources. The pharmaceutical industry has tried to conceal medication prices because of their continuing legal cases in the United States. There is a lack of transparency regarding medication and vaccine prices among non-governmental organizations."

The first sentence is in direct contradiction to and misunderstanding of policy. Our policy doesn't just say (as James further above claims) that if a "reliable source" mentions something, then we can also mention it, in whatever detail we choose and however prominently we choose. Whether or not to include things, and whether they are significant enough to appear in the lead, and how much detail to afford to them, is determined by WP:WEIGHT: "in proportion to the prominence of each viewpoint in the published, reliable sources". Our policy goes on to say "Undue weight can be given in several ways, including but not limited to depth of detail, quantity of text, prominence of placement, juxtaposition of statements and imagery". Weight isn't just about viewpoints, but about any aspect of a subject and how it is presented. This is explained in policy: "An article should not give undue weight to minor aspects of its subject, but should strive to treat each aspect with a weight proportional to its treatment in the body of reliable, published material on the subject". The second two sentences added by QuackGuru are his own agenda and not appropriate text for any guideline on Wikipedia. The editorialising was quickly removed by WhatamIdoing but the initial faulty sentence was retained. QuackGuru then added another statement:

"Cost information may be included in the lead, but this is done on a case-by-case basis"

The reason for this would appear to be to defend Doc James addition of detailed price information in the lead of mainly the WHO essential drugs. It isn't supported by any policy, guideline or consensus of editors on Wikipedia. Indeed, the RFC of two years ago (not 10 as claimed above) rejected the general inclusion of price information in articles. James has found a source that lists a handful of prices from a limited small set of suppliers of a given drug back in 2014 and has extrapolated that to be "The wholesale cost in the developing world" (e.g. Mebendazole). This is a statement that is simply untrue. Nobody has published "The wholesale cost in the developing world". There is a database that contains specific wholesale prices from a very small collection of suppliers who mostly target the developing world. They have got the price information from whoever makes that price information available to them: they don't claim this is actually representative of world-wide pricing or developing-world pricing. There could be huge gaps in that data due to price being withheld or simply hard to retrieve. Perhaps it is likely to be in the approximate ballpark, but no more than that.

Because the source does not perform any statistical analysis on the price, neither can James. But he chooses the smallest and largest price in the database of six suppliers. Picking the smallest and largest value in a dataset is not a typical method used to statistically analyse data and present it to people, because it is prone to giving undue weight to outliers. Performing analysis on a tiny dataset is prone to drawing incorrect conclusions because information is missing. Some of the suppliers in this small dataset could be orders of magnitude larger (and thus more significant) than others, yet we treat them all equally by ignoring all except the cheapest and dearest. That is simply statistical and economic incompetence. The data is certainly now out-of-date as the website has not been maintained for five years. We are told the price in the developing world is between XX and YY. That means we claim to have a reliable source that says the cheapest price in the whole developing world is XX and the most expensive price is YY. But in fact we don't. We have a source that merely gives a price from six suppliers who do not cover the entire developing world. And a price that is 5 years old. And let's not even start on a definition of "Developing country" which may or may not include South Africa, India, China and Russia depending on which measure you use.

James directly quotes these arbitrary prices in dollars and cents. We have in our lead the claim "between USD 0.004 and 0.04 per dose". Let's return to WP:WEIGHT. Do our reliable sources give weight to the lowest and highest prices in the IMP Price Guide of 2014 for this drug? Do our reliable sources quote those prices in tenths of a cent precision? Do they think that the price, to tenths of a cent, charged by German charity Action Medeor International Healthcare to one African country, Tanzania, is important enough to mention in the lead? I hope you agree that is ridiculous. The only reason our lead, as written by one Wikipedian with an agenda, mentions the price in Tanzania by a German charity, is because that's all that James found convenient when web surfing. Not a single one of our reliable sources on Mebendazole mention the price, in tenths of a cent in Tanzania in 2014, in their leads. So neither should we.

Then there is the question of what we even mean by "dose". The article doesn't explain or link. The source gives the price per tablet and also gives a defined daily dose. For example, Mebendazole, the lowest price is $0.0035 per tablet of 100mg and a "defined daily dose" of 200mg. Our article just says "$0.004... per dose". What our article neglects to mention, and most of our reliable sources do mention, is that mebendazole might only need a single dose of 100mg to treat threadworms, though a second dose after two weeks is used if reinfection occurs (which is common). For some other infections, the treatment is 100mg twice a day for three days. Add to that, the advice is often to treat the whole family. Now, we generally don't go into specifics about mg dosage, but it is encyclopaedic that this is a very short and quick treatment (compare e.g. fungal nail infection which might take 6-12 months of treatment daily). Which figure is useful to our readers? The single dose taken three times a day, the 3x combined daily dose, the "defined daily dose" standard "used to standardize the comparison of drug usage between different drugs or between different health care environments" (which is not the therapeutic dose), or the total amount needed to fully treat the condition in the patient, or the amount for the whole family? So, just like wholesale vs retail pricing, we aren't really defining what we mean by dose. Should the lowest price in the lead be $0.0035 tablet in Tanzania, 0.007 per defined daily dose, or give different values for threadworms vs other infections, or total treatment cost?

The article also claims the price in the UK is around $5, citing an FT article I can't read. The BNF says 100mg tablets from Janssen-Cilag Ltd costs £1.34 for a pack of 6 tablets, which is wildly different from $5. The UK retail price is £8 or £9 for a pack of four from two pharmacies I found, which would likely require two packs (£16-18) to treat the family of four with two doses two weeks apart. Why are we mentioning $5 in the UK, when that bears no resemblance to any price whatsoever? It is just wrong.

There is notable commentary in reliable sources about the US price hike. Also this article from 2011 mentions dodgy accounting by reputable charities wrt the price of this medicine.

In summary, we are making patently false and unsupported claims in our leads: nearly every fact presented is just plain false and our wording ambiguous. We are giving way too much weight and precision to prices from an arbitrary and tiny collection of sources, and our major source for such prices has not updated their data for 5 years. And to add to that, our Mebendazole article doesn't mention far more relevant encyclopaedic information such as that just two doses of the tablet, taken two weeks apart, will clear a threadworm infection, or that family members are encouraged to be treated even if they don't show symptoms.

The claim that such information is supported by reliable sources takes a deliberately naive view of how sourcing should influence our article content. I'm sure everyone here knows that people turn up on Wikipedia with claims made by some mouse experiment in such and such a journal. They are perplexed that although we agree that journal is a reliable source for what the experiment did, it isn't a reliable source from which we can make health claims in humans. Similarly, a 5-year-old neglected database of prices from six random organisations is not a reliable source from which we can present the very complex business of worldwide drug pricing to our readers.

Our lead should contain a summary of only the most vital and robust information about the article topic. It is, after all, what is read out to us by our electronic devices and what is translated to other Wikis. Instead we simply have nonsense written by incompetent amateurs who are misinterpreting their inadequate and out-of-date sources. It couldn't really, be much worse. -- Colin°^Talk 21:32, 23 November 2019 (UTC)

Well, there is some precedent for this. It depends on whether you think that the price of a product or service is more like a birth date or more like a criminal conviction. We routinely include bare facts in the name of comprehensiveness even when almost no source mentions them. In biographies, that means that we include people's middle names, birth dates, parents' names, cities of birth, high schools, etc., even when no source actually cares about those facts. In articles about chemicals, we include bare facts such as molecular weight, even though we know that nobody's actually going write an article about how it's so interesting that the molar mass of chlorine trifluoride is 92 instead of some other number. We do this in medical articles, too: if there's no known treatment, then we would usually have a section called ==Treatment== that makes a brief statement that there's no known treatment. We just feel that these articles would be incomplete without this kind of information, and the difference in terms of WEIGHT is whether you mention these things in passing, or if you put a lot of emphasis on them – not whether it's mentioned at all. WhatamIdoing (talk) 02:39, 24 November 2019 (UTC)

Much more like a criminal conviction. A birth date or molar mass is a pure, bare statement of fact. The molar mass of chlorine triflouride isn't one thing in San Diego and something very different in Rio de Janiero; it's just 92, everywhere, every time. If you were born on 1 January 1970, that's your birthday, period, everywhere. The price of a given medicine very well might be different everywhere, and even different depending on how it's being procured or purchased. That's not something we can just slap a number on; it would require contextualization and discussion by a reliable source, not just grabbing a number out of a database. There are cases where reliable sources have extensively discussed the prices of certain medications or treatments, and I'm all for including such well-sourced information in the article, but it shouldn't be a routine practice to add to all or most articles. Seraphimblade ^{Talk to me} 04:19, 24 November 2019 (UTC)

Agree with Seraphimblade. Someone's middle name, DoB, etc are the sort of information one expects in an encyclopaedia and are either correct or not. The price of a drug is just about the most nebulous thing we can consider: where, when, how much, how delivered all cause huge fluctuations. Consider the price of a mango:

• Standard whole fresh fruit. 69p
• Giant whole fresh fruit. £1.80
• Twin pack wrapped in plastic and ripened ready to eat. £1.89
• Dried mango slices, 200g. £4.00
• Mango slices in syrup in tin can, 425g. £1 or 3 for £2.50
• Mango pulp in tin can, 850g. £1.75

And that's just from one major supermarket in the UK in November. Now imagine that I wrote "The price of mango in the developed world is between £0.69 and £4.00" and inserted that into the lead, and edit warred to retain it there and then went and edit warred at WP:MOSFRUIT to ensure the price of mangos was required. The database James found is just a collection of data values from limited sources, just as MySupermarket.co.uk is a collection of data values from limited sources (it doesn't include the price at markets and small shops, just large chain supermarkets). James has engaged in WP:OR to make claims that this dataset represents the lowest and highest prices in the developing world and presented it in a way we don't really know what the price is for.

The price of mebendazole is not notable except in the case of the US price hike. Without that hike, the price follows the pattern of any cheap-to-produce out-of-patent pill around the world. Further, it is remarkable for being effective with just a very short treatment regime -- something our article does not mention. Time and again I see WP:MED arguing to include factoids or awful videos but not even getting the basics of a text encyclopaedia right. These embarrassingly bad price statements are present on dozens, hundreds, of our key drug articles, and they are all, every one of them, as wrong in so many ways as a Trump tweet. -- Colin°^Talk 12:34, 24 November 2019 (UTC)

And yet if I added the annual average FOB (~wholesale) prices for imported mangoes according to the USDA to Mango, I doubt that anyone would tell me that was not encyclopedic information and that readers would be mad to discover that they weren't paying US$4.60 to 5.74 per box in the US during 2018 and therefore all information about prices urgently needed to be removed to stop people from thinking that the wholesale price was their local retail price, or that it was terribly confusing to let the experts at the USDA weight conventional and organic, and different sizes of mangoes, and different ports of entry, and different seasons, and the different countries of origin into a single price range all by themselves. I might get complaints that it's only the US (which is fair; 8 of the top 10 mango producers in the world don't export mangoes to the US in any considerable quantity, but that suggests expanding it with a more prices for different countries, not removing the prices that we do have), and I might have to point out that the US grows very little mango domestically, so the import price basically is "the" price, but I don't think that anyone would say that the overall economics of a popular fruit crop are unimportant unless I'm citing a source that talks about those prices in great detail. WhatamIdoing (talk) 16:57, 24 November 2019 (UTC)

• I just set up Wikipedia:Defining data. I intend this to be the start of some guidance about what sort of information should go into articles, seemingly in circumvention of the guidance at WP:NOT, WP:RS, WP:WEIGHT, WP:V, WP:LEAD, and the rest. WhatamIdoing, I think this is what you were mentioning above. Colin, I think you are saying that price is not "defining data", and I am saying that some information about price is defining. Again, I agree - we have great difficulty actually getting this data, but if we had it in sufficiently high quality, then I would like to present it as fundamental. I see price to be as fundamental to drugs as concepts like nationality and century of birth are to a biography. Obviously we need more discussion. Blue Rasberry (talk) 15:20, 24 November 2019 (UTC)
  • Bluerasberry the concept of "Defining data" is interesting. I don't think that wiki projects have always been sensible about what to include in info boxes. For example, WP:MED for years put lots of external database links in info boxes, and I see that is now demoted to a collection at the bottom of the article. Even now it tries to include some rather complex concepts in the infobox. Perhaps a good characteristic of the data we should include in info boxes is whether that data describes something simple and straightforward rather than nuanced and complex. Does it need a lot of explanation? Even so, it can be simple for some subjects and hard for others. The infobox for Measles is fairly straightforward. The Tuberculosis one breaks down a bit with its Frequency datapoint being hard to explain and the value given is probably not what most people would consider (numbers with illness). Infoboxes become downright silly with Cancer and Epilepsy which are really groups of diseases that may have very little in common.

But surely "defining data" is information about a subject that if it were significantly different, it would be a different subject. Someone's DoB, parents, partners, children, occupations and year of death all go to define a person. The defining attributes of mebendazole include its chemical and therapeutic properties. It may also include the company that invented it and may still hold a patent on it. Whether it is licensed in Europe or US. In the UK we might be concerned if it is available only by prescription (PoM) or only from a licensed pharmacy (PM) or widely (GSL). In the US you care much more about price than we do in the UK. Depending on your circumstances and age, prescribed medicines are either free of charge or require a fixed £9 charge for about 1 month's supply. Whether a bottle of pills has a wholesale cost of £1 or £2 or £10 is really only the concern of my pharmacist.

The recent astronomical drug price hikes in the US (and occasionally in the UK) demonstrates in fact that price is very much not a defining characteristic of a medicine. In fact, it is more a representation of the economic systems of countries, and the price regimes their governments do or do not enforce. Has drug company mergers led to decreased competition, have monopoly regulators been asleep? These are not, in fact, questions about a drug at all. The drug price is just a pawn in a bigger game. -- Colin°^Talk 17:10, 24 November 2019 (UTC)

Don't you think that whether the wholesale cost is £2 or £200 is also a matter of interest for taxpayers and policy makers? Perhaps just as it doesn't actually matter whether some politician's birthday is January 1st or January 2nd (although, for better or worse, we do report that level of detail), the exact details aren't that important, but the overall order of magnitude could be important to the healthcare system and its users (not to mention students writing about the drug for school).

Which brings me to the other point: Prices matter to the business end of healthcare. Any individual consumer or taxpayer might not care about the general price range for a given drug, especially since so many of them are generic drugs, but investors should and do care. Are we actually writing comprehensive articles if we're omitting key information about that POV? I notice that s:1911 Encyclopædia Britannica/Tea mentions the early price of tea in England and the then-present estimated capital value of those big businesses, so I think it's fair to assume that some business matters are a proper subject for encyclopedias. WhatamIdoing (talk) 17:24, 24 November 2019 (UTC)

And that's great, and there are places those can be found in many cases (and if it can't be found, well, then it's unverifiable so we couldn't include anything about it either). It's useful is explicitly not an argument for inclusion in an article. Bus, train, and flight timetables and schedules are indisputably useful and of interest to many people, but we still do not include them, both because they're a greater level of detail than is generally encyclopedic, and they often change and would quickly become out of date. That's the same reason we do not generally include prices. Now, with something like the treatment mentioned above, which reliable sources have extensively noted is the most expensive in existence, sure, then we should mention that, in the same context the sources do—its extreme price is part of what's notable about it. But we shouldn't do that for run-of-the-mill drugs or treatments where their prices are not of any particular note. So, while there are exceptions, the general rule of other articles about products should also apply here: No pricing. Seraphimblade ^{Talk to me} 18:40, 24 November 2019 (UTC)

WAID, I just noticed your comment above. My mango comment wasn't made to get side-tracked in to how to price mangoes, but to demonstrate the silliness that I'd found a website quoting a handful of mango prices and had done OR on that to make claims that really aren't true. WAID, we don't have a source for the cheapest price of mebendazole in developing nations. We don't have a source for the dearest price of mebendazole in developing nations. Our source doesn't even claim to represent "developing nations", whatever definition you might use. Nor do we state what the price is actually for. Surely you agree that "what is this price for" is crucial? All our important drug articles make price claims that are simply not true and not supported by the source given. Are you not concerned about that? Consider Ibuprofen:

We can normally take two 200mg ibuprofen tablets up to 4 times a day. The defined daily dose of ibuprofen is 1.2g. So by "dose" do we mean a 200mg tablet, 400mg individual dose, 1200 defined daily dose, or 1600mg maximum normal daily dose?

• Ibuprofen says "The wholesale cost in the developing world is between US$0.01 and US$0.04 per dose"
• The source has "Lowest Price: 0.0046/tab-cap", "Highest Price: 0.0080/tab-cap" (and explains that lowest/highest is just among their 9 suppliers, not "the developing world"). The tablet "Strength: 200 mg". The "Defined Daily Dose: 1.2 G".
• If our dose is a 200mg tablet then "$0.005 to $0.008"
• If our dose is two tablets, 400mg, then "$0.009 to $0.016" (or $0.01 to $0.02 if we round).
• If our dose is defined daily dose, 1200mg, then "$0.028 to $0.048" (or $0.03 to $0.05 if we round)
• If our dose is maximum daily dose, 1600mg, then "$0.037 to $0.064" to $ (or $0.04 to $0.06 if we round).

None of these ranges bear any resemblance to what our article claims. The article also claims "In the United States, it costs about US$0.05 per dose." without saying if that is a wholesale price. The source does not, AFAICS, include any price information whatsoever. Consider the epilepsy drug valproate.

• Valproate says "The wholesale cost in the developing world is between US$0.14 and US$0.52 per day" added by Doc James in 2015.
• The source has one supplier price (IDA) at $0.1486 per 500mg tablet. If we might multiply by 3 to get the 1.5g defined daily dose, it comes to $0.45 per day. The other two prices it lists are buyer prices, for Peru and Sudan. These work out at $0.23 per day (Peru) and $0.79 per day (Sudan). There isn't AFAICS any price or price calculation that gives "between US$0.14 and US$0.52 per day". The price quoted in our article is similar but not exactly the same as two 500mg tablets, but the defined daily dose would be three 500mg tablets.

Honestly, the more examples you look at, the more one wonders if this is some experiment in trolling Wikipedia to include random false facts and see how long the community takes to notice. You ask if our readers might get confused. No, I think here we actually have an editor who is confused, who really hasn't a clue about what they are writing, and has spread their confusion over all our major drug articles. Are any of our drug prices correct? -- Colin°^Talk 18:43, 24 November 2019 (UTC)

Like WAID said, "Don't you think that whether the wholesale cost is £2 or £200 is also a matter of interest for taxpayers and policy makers?" I agree that reporting bus fare and the difference between $0.03 to $0.05 is silly, so let's leave that out because that was never the point. Now you guys say something about uncertainty about whether something is approximately $2 or $200, which is the difference between accessible and inaccessible. Why not communicate this? We seem likely to be able to do so soon with good data and technical capacity, and if people want to experiment with the process now, then why not? Blue Rasberry (talk) 22:21, 24 November 2019 (UTC)

That's...not how any of this works. Wikipedia follows sources. They, not us, should be doing the experimenting. If they decide to go into a lot of detail on the price of certain medication, we follow their lead and do the same. If they don't, we don't either. If we will one day have sufficient data, we will use that data after it exists, not presume one day it will. Wikipedia follows, never leads. Seraphimblade ^{Talk to me} 00:45, 25 November 2019 (UTC)

I think we have two different conversations going on. WhatamIdoing and Bluerasberry are still trying to make the case for including a cost figure and not really engaging on the problem that (a) we don't have sources for what they want and (b) all our major drug articles have incorrect drug cost figures. WhatamIdoing, there are several parties who care about price. You say "prices matter to the business end of healthcare". I return to the question "the price for what?" Clearly our current description of pricing is not actually stating "what" because "dose" is not defined and can be defined in many ways (though defined daily dose would appear to be the relevant one for financial comparison). But for a drug company, that only has meaning when combined with how many doses a person takes (or keeps taking for life), how many people get ill and need the drug, whether other drugs compete or will compete, what competition there is in the market for this drug (either on patent or off patent but nobody else making it), what agreements they already have with healthcare providers that control prices, etc, and whether you want to spend eternity in hell for charging $400 rather than $0.04 for a pill to get rid of threadworms. Our cost information will not help anyone make that decision, and frankly, nor can they be trusted. I think if you want to make the case that cost information should be presented in our drug articles, and describe how it should be presented, then you need to start with an example of best practice. MEDMOS shouldn't be based only on speculative information we might not be able to reliably get, nor based on the very bad practice described above, but on best practice. -- Colin°^Talk 15:08, 25 November 2019 (UTC)

Seraphimblade, the idea of "experimenting" is to experiment with different approaches for reporting what the sources say. For example, would editors be happier with a properly sourced sentence that says "The wholesale cost of generic ibuprofen tablets is usually about a penny each", and leave it to the reader to guess that we're talking about the standard 200mg pill, or would we be happier with a properly sourced bot-controlled item in an infobox that gives the price for an extremely specific definition, such as the "US Average wholesale price (pharmaceuticals) for the Defined daily dose on the specified date"? US AWPs have been reported for four decades for drugs that require long-term treatment, and DDDs are standardized worldwide (that's the point of the DDD, although for ibuprofen, it'd require more than one entry, because there's more than one DDD), so we can source those prices. The "experimental" part here would be deciding whether we'd prefer something approximate in prose vs something precise (we can get the AWP for ibuprofen down to the thousandths of a cent on any specific day, which might be overkill) vs some other approach (e.g., list price at time of introduction, in the ==History== section for a drug). WhatamIdoing (talk) 16:47, 25 November 2019 (UTC)

WhatamIdoing, that's a false dichotomy. The general rule for products is that we do not include pricing at all, and I see no reason to deviate from that here. Exceptions are made in which the price, in and of itself, is one of the notable features of the product, but if it's just something that's sold and no sources talk about the price being of particular note, it should be omitted entirely. So, in most cases, I would prefer to see neither. Seraphimblade ^{Talk to me} 19:38, 25 November 2019 (UTC)

That is simple not correct. We have whole pages on prices.Price of oil Doc James (talk · contribs · email) 21:28, 25 November 2019 (UTC)

Seraphimblade, I'm not sure that the former general rule has as much widespread support as it used to, either in this subject area or in any other. Template:Infobox camera has a parameter for the list price. Template:Infobox battery includes a measure of cost-effectiveness (the amount of energy compared to the price paid by end users). Template:Infobox aircraft type includes the price of an aircraft, and they've clearly put some effort into deciding exactly which price they want to include. Looking at some FAs, Sega Genesis lists the initial or list price for multiple products, plus the clearance price for one of them. 32X has the initial price in three currencies in the infobox. Science Fantasy (magazine) reports the cover price. NeXT names the price for one of their products in the lead. House (TV series)#Distribution tells you how much iTunes is currently charging to buy an episode. If editors genuinely didn't want prices in articles, then I don't think that we'd find it in infoboxes and Featured Articles. WhatamIdoing (talk) 22:00, 25 November 2019 (UTC)

Featured article samples

My apologies for just now getting to this, as I was traveling during Thanksgiving. IMO, these FA samples are not indicative that support for NOTPRICE has necessarily changed. And even FAs are subject to WP:OTHERSTUFFEXISTS-- even more so with the decline in participation at FAC and FAR since 2011. Even in the best of FAC times, it was near impossible to get top reviewers to engage with gaming articles, so one gaming article passing FA with prices in the infobox doesn't indicate much to me about Wikiwide trends.

Looking at each of these examples:

1. Sega Genesis promoted Dec 2013 seems to discuss historical prices in keeping with the restrictions and exceptions of WP:NOTPRICE, with reasoning for the price information being significant put in context relative to history, competition, marketing, etc. and from independent sources. This is different than using database drug prices with no context for why we are reporting the price.
2. 32X promoted March 2014 is a great example of WP:OTHERSTUFFEXISTS even in FAs. It is a very weak, WikiCup FAC, with clear fan support (which was a typical issue during WikiCup) and limited independent or expert review, such that it would not have been promoted FA in the days of more active participation at FAC without more strenuous review. (FAC delegates who followed my tenure had to deal with declining reviews, and have less help in confronting effects of fan supports from inexperienced reviewers such as occur during WikiCup.) This weak gaming article is not a strong indication of anything Wiki-wide.
3. Science Fantasy (magazine) promoted May 2011 by me was written by a top-notch FA writer and reviewer, Mike Christie, who writes about (among other things) historical science fiction magazines, generally no longer in print. Pricing issues mentioned have to do with the historical context of business decisions made, and are in keeping with WP:NOTPRICE. Again, this appears to a policy-compliant use of pricing data, very different from what is being proposed with adding current prices to drug articles without context or significant sourcing.
4. NeXT was promoted in 2008 (before the time when the bot indicated who promoted it, but it was most likely promoted by me), and the FA promoted version did not have that information in the lead. The article has not been reviewed at FAR (most FAs these days have not been reviewed at FAR, because FAR is dead, which is why most FAs today are not actually FAs at all). In other words, this is a very old, unreviewed FA ... and the mention in the lead, added since its FA promotion, appears to comply with WP:NOTPRICE because of the context given. Again, this is very different from what is being proposed with drug prices.
5. That House (TV series)#Distribution "tells you how much iTunes is currently charging to buy an episode" (in 2008) shows just what an embarrassing debacle that editor's FACs were. This FA was written by an editor who was later revealed to be a prolific sockmaster, with supports for promotion of his articles coming from his socks. It was promoted by me in June 2009, that editor never wrote an FA I thought was up to standard, but my job was to determine consensus among reviewers, so he had numerous FAs passed based on his sock supports. I thought his work was dreadful, and this article passed FAC based on his own sock supports. That this editor was one of my main FAC frustrations during my tenure is obvious in that I let a line that talks about a "current" price get by me, and that inaccurate line (since it does not specify an "as of" date) is still in the article. This is another great, but embarrassing for me, example of WP:OTHERSTUFFEXISTS, and not an indication of anything Wiki-wide. Its editor was most unpleasant to deal with and his work eventually could attract very few reviewers, so he socked and did his own.

In summary, I am not seeing strong evidence of a trend here. SandyGeorgia (Talk) 06:34, 8 December 2019 (UTC)

Sandy, another good WP:OTHERSTUFFEXISTS example: In the edit war at ivermectin, a "fourth opinion" editor cited the fact that it appears most of our drug articles have prices as evidence that Wikipedia supports prices in drug articles. But since all (as far as I can see) prices in drug articles were added and occasionally maintained by a single editor and that editor was slow edit-warring to retain the price in that article (which suggests strongly that further examples could be found) one must be careful not to use bad practice as evidence that it is accepted. WhatamIdoing, I would be far more impressed if WP:MED could bring a dozen drug articles (not trival ones, but ones like have been discussed here and at WP:MED, which are typical rather than edge-cases) bring them up to first-class standard wrt source->text compliance with policy. If that was done then the matter of "do we include drug prices" would be a much simpler matter of consensus opinion. I really think "first show us how you'd do it" is the obvious step. And you are welcome to try. -- Colin°^Talk 10:59, 8 December 2019 (UTC)

What's found in the articles about the Genesis (and some comparable articles, such as Nintendo Switch) is the original list price. I can see why that might be included, as generally the introductory price of a game console will indeed be discussed pretty heavily by sources immediately preceding and following the launch. I would be amazed to find there isn't a bunch of source material about it. On the other hand, something like the iPhone 11 lacks it (though ironically, it had a spam link for it I cleaned up), because it doesn't even really have a particular price. Sure, there's a list price if you just buy one, but very few people outright purchase at that price; most get them through some kind of deal with their carrier. Additionally, the Genesis article doesn't go into what getting hold of a Genesis would cost you today, so that is a static figure that won't become outdated. Those are all very different circumstances than medications. They're much more similar to the phone—ask ten people what they paid for them, and you'll likely get ten different answers. The same is true of oil prices; that in itself is a notable subject. The subject of medication pricing might also be worth an article if we don't have one; I suspect we could find a fair bit of sourcing for that. But that's different than trying to put current prices in individual articles. Seraphimblade ^{Talk to me} 22:32, 25 November 2019 (UTC)

You said "On the other hand, something like the iPhone 11 lacks it (though ironically, it had a spam link for it I cleaned up), because it doesn't even really have a particular price." The reason it lacks pricing in the article is because you deleted it from the article.[2] QuackGuru (talk) 00:38, 26 November 2019 (UTC)

QuackGuru, you attempted to put it in after I made that comment. Check the timestamps, please. Seraphimblade ^{Talk to me} 07:39, 26 November 2019 (UTC)

Seraphimblade, the original list price for every model of iPhone is listed in the large table at iPhone#History and availability (i.e., the first section). Also, I believe that "deal from your carrier" thing really only happens in the US. Most of the world pays cash on the barrel head for their phones.

It would be trivial to find equivalent sources for the original (US) list price of any new drug. WhatamIdoing (talk) 16:00, 26 November 2019 (UTC)

They certainly are, and with no sourcing to indicate any type of significance. I don't have time to clean up such a large table right now, but I'll correct that once I do. In any case, the correct venue for this, ultimately, is an RfC at WT:NOT. If there's consensus that WP:NOPRICE should be changed or removed, well, then such as it is and we'll go forward with that. But until and unless that happens, that is the global policy on prices and articles, and that rule is "With a few exceptions, don't include them." A local consensus here couldn't change that in any case, so this really is just an academic discussion. If you'd like to work on formulating an RfC, I'd be happy to give my thoughts on wording. Seraphimblade ^{Talk to me} 17:25, 26 November 2019 (UTC)

WhatamIdoing with an iphone the original list price doesn't require any maths or judgement about doses or indications. You say it would be trivial to find original US list prices of any new drug. But all you can find is the price for a product with a bar code: a particular pack size of a particular strength in a particular formulation of a particular brand. I can't just transform that into "price for a drug". Also, "any new drug" doesn't help us with the 99% of our drug articles that are existing drugs. -- Colin°^Talk 17:33, 26 November 2019 (UTC)

The original list price of a MiniMed artificial pancreas doesn't require any maths or judgment about doses or indications, either. It's just an electronic device, complete with bar code. What we recommend about how to describe prices should encompass more than just drugs. WhatamIdoing (talk) 00:20, 27 November 2019 (UTC)

That's very nice, but what we recommend now is "Don't include them". If you'd like to see that change, the aforementioned RfC is still needed. Seraphimblade ^{Talk to me} 04:00, 27 November 2019 (UTC)

See WP:NOPRICE: "An article should not include product pricing or availability information unless there is an independent source and a justified reason for the mention." It does not conclude never include pricing. There is independent sources and there is a justified reason to include the pricing for the IPhone 11 but I believe you violated WP:NOPRICE.[3] There is significant independent press coverage. That is a justified reason. QuackGuru (talk) 12:32, 27 November 2019 (UTC)

International Medical Products Price Guide

The source for most "developing country" prices in our drug articles is the "International Medical Products Price Guide". This is typically expressed in our articles by a statement like "The wholesale cost in the developing world is between US$0.14 and US$0.52 per day." or "The wholesale cost in the developing world is between US$0.01 and US$0.04 per dose" which appears in both the lead and a Cost section. Although cost statements are widespread in our drug articles, particularly the WHO Essential Medicines, only one user, Doc James, added them, mostly in 2015.

The guide "provides a spectrum of prices from nonprofit suppliers and commercial procurement agencies, based on their current catalogs or price lists. It also contains prices obtained from international development agencies and from government agencies." Its purpose is "to improve procurement of medicines of assured quality for the lowest possible price". It claims "comparative price information helps in price negotiations, in locating new supply sources, and in assessing efficiency of local procurement systems". It cautions "This Guide is intended as a comparative reference only", "The vendors included are not intended to be a comprehensive list of potential suppliers".

There are two types of price in the guide. Buyer and Supplier prices. Buyers are usually government agencies and the price is what they obtained after negotiation, competitive bidding, etc and are not available to anyone else. The buyer prices are indicative only, and can't be used for comparison or reference purposes. Suppliers are organisations who have a warehouse and supply to customers. The prices are further complicated as there are other cost factors to consider when ordering medicines: insurance, transport (air, sea, inland), import duties, unloading costs, handling fee, currency conversion fees, minimum order.

The guide is aimed at the person ordering medicines for their hospital, charity or government department, who is being offered various prices and wants to know if that is reasonable ("for comparative purposes only"). If much higher than the guide prices, then there may be room to negotiate a better price. If much cheaper than the guide prices, questions should be raised about quality.

What the guide does not claim to be is a comprehensive reference of prices in the developing world. It doesn't weigh the prices offered by a huge international supplier as any more significant than the prices offered by a small regional one. For some drugs, formulations or sizes, there may be no price at all, or only one Buyer or one Supplier price.

The guide gives a unit price: per tablet or capsule, or per ml if liquid, or per g if a cream, etc. Each of these will correspond to a particular amount of medicine (often in mg). The actual dose taken by a patient is likely to be unrelated to the unit price: they may take two tablets, measure out 20ml or spread less than a gram on their skin. The guide makes no attempt to related this unit cost to the actual cost incurred by a therapeutic dose or to fully treat a condition. Instead it offers the defined daily dose. This is the "average maintenance dose per day for a medicine used for its main indication in adults". This could therefore be quite different to the dose actually used or when used for another indication. Its usefulness here is in estimating the likely quantity of a medicine that a hospital may go through each day, and thus need to regularly procure. Do I need to buy 10,000 100mg pills a month or 20,000. For some medicines and formulations, the defined daily dose is not available. It also doesn't indicate the duration of treatment.

The guide mostly presents raw data: the package size (100 tablets), the package price ($5.85), the unit price ($0.0585/tablet), the tablet strength (200mg) and the defined daily dose (1.5g). It does perform a little statistical analysis if several prices are available. It gives the lowest unit price, highest unit price, median unit price and high/low ratio. The guide explains that the most useful indicative value is the median unit price. It explains that a simple average (mean) is too influenced by outlier prices. Similarly the lowest / highest prices are themselves most likely to be outliers. The high/low ratio gives an indication of how close or spread the price is. So the most useful price, if we were to present it to readers, would be this median. And it would serve merely as an indication of a typical wholesale unit price available to purchase on the international market.

But a unit price (per tablet or per ml) isn't very useful to our readers, particularly as we don't tell them how many tablets or ml to take. Instead our current wording usually talks of "dose". This is defined to be the amount of drug administered at one time. For ibuprofen, the BNF tells me: "Initially 300–400 mg 3–4 times a day; increased if necessary up to 600 mg 4 times a day; maintenance 200–400 mg 3 times a day, may be adequate". For children, it gives different dose values for seven different age groups. And that's just for the most common indication. Some pills are available in an extended-release formulation, which requires only a once-a-day dose. Other medicines may be designed to take effect rapidly (e.g., buccal midazolam). There are endless permutations of single-dose, three times a day for two weeks, once a day for the rest of your life. As noted above, there is a defined daily dose but that is intended to be used to help the purchasing team know how many pills to order for the hospital, not how much of a medicine a reader might actually use.

Since the guide is based on purchasing a particular form of a medicine, it may actually have many entries for the one drug. For example Valproate.

• Sodium valproate 200mg ec tablet. 100 tablets. Anticonvulsants/antiepileptics. 6 suppliers. 0 buyers.
• Sodium valproate 25mg/5ml liquid. 120ml bottle. Anticonvulsants/antiepileptics. 0 suppliers. 1 buyer.
• Sodium valproate 500mg ec tablet. 100 tablets. Medicines used in bipolar disorders. 1 supplier. 0 buyers.
• Valproic acid 150mg. 100 tablets. Anticonvulsants/antiepileptics. 1 supplier. 0 buyers.
• Valproic acid 200mg ec tablet. 100 tablets Anticonvulsants/antiepileptics. 0 suppliers. 1 buyer.
• Valproic acid 200mg/5ml. 300ml bottle. Anticonvulsants/antiepileptics. 0 suppliers. 1 buyer.
• Semisodium valproate (a drug with equimolar amounts of sodium valproate and valproic acid) -- no information.

So we have 150mg, 200mg or 500mg tablets, and 25mg/5ml or 200mg/5ml syrups. We even have enteric coated and normal tablets. Which of these is "the price"? Most of them have data only for one supplier or one buyer. The drug is used for three main conditions: epilepsy, migraine and bipolar disorder. The recommended initial dose, typical dose and maximum dose is different for each condition. There is no formula to convert unit price to dose.

The International Medical Products Price Guide has not been updated since 2015. I have emailed them to ask why and if there is likely to be a future update. Perhaps they are changing to update less frequently than annually. While the guide provides a lot of data, turning this into information that our readers might use would require a lot of original research and there is no clear algorithm that would take these data values and offer a simple number to the reader. Our source does not give any indication of "dose", nor could it, so we can't quote a price per dose. Although our source does perhaps give a "per day" price (by multiplying the unit price to achieve the defined daily dose), it very much cautions against using this metric for any purpose other than stock control. And our source offers multiple prices for different formulations and no way to choose which is relevant. We could give the median unit price for a 200mg tablet in 2014, but our reader would have no way to interpret what a "200mg tablet" represented for a given condition being treated.

So I think that while this guide is invaluable for its intended purpose, I cannot see any way we can use the raw data it presents in a way that is meaningful to present to our readers of a drug article. It is clear that current use of this source is

• misrepresenting it as a comprehensive reference of developing nation prices
• giving too much weight to the two potentially-outlier values of lowest and highest price
• using the ambiguous term "dose" or "per day", which is not supported by the source, which only gives unit prices
• simply incorrect -- I cannot find examples where the quote prices match the source under any interpretation of the data values

I propose that all use of this source to present the prices in developing nations be removed from Wikipedia. Statements like "The wholesale cost in the developing world is between US$0.01 and US$0.04 per dose" are factually incorrect and misleading in every single word and every single number.

I see similar problems with our prices per dose for the US. These link to Drugs.com monographs which do not actually contain prices. Instead the price is on another page (e.g. diazepam, valproic acid). Those are retail prices "based on using the Drugs.com discount card which is accepted at most U.S. pharmacies." So our lead juxtaposes "developing nation" wholesale prices with US retail prices for a certain discount card holder. And again we have the whole mess of different formulations, different conditions being treated and different pill sizes or syrup strengths and no way to define one and only one "dose". So it seems likely that we should need to remove the US dose prices as well. -- Colin°^Talk 18:22, 25 November 2019 (UTC)

Websites frequently shuffle stuff around. It was present on the page in question when the source was used.[4]

Here we have the "National Average Drug Acquisition Cost"[5] Doc James (talk · contribs · email) 18:33, 25 November 2019 (UTC)

Colin, this particular comment feels like it's more about the application of MEDRS to a specific source than about what general advice to give in MEDMOS. Maybe this particular discussion should be moved to RSN or WT:MED instead? WhatamIdoing (talk) 18:51, 25 November 2019 (UTC)

We could change it to "The wholesale cost in the developing world is about US$0.40 per defined daily dose as of 2015."

But in my opinion "The wholesale cost in the developing world is about US$0.40 per day as of 2015." is reasonable.

So these values are published in a book by WHO. The specific number for valproate comes from the International Dispensary Association Foundation an NGO founded in 1972 that specializes in the delivery of medications to the developing world, so it applies to that area of the world. Doc James (talk · contribs · email) 20:30, 25 November 2019 (UTC)

WhatamIdoing, the problem is that MEDMOS is currently pushing the non-consensus of price/cost into drug articles, whereas MEDRS does not. Price is an economic factor, and not really MEDRS's concern. Perhaps that's why WP:MED is making such a hash of it -- we need some accountants, statisticians and basic maths ability here.

Doc James defined daily dose is explicitly not a useful figure for our readers. See WHO Definition. It exists merely to indicate consumption that affects hospital supplies or consumption in a population. "The DDD is sometimes a dose that is rarely or never prescribed because it is an average of two or more commonly used doses." and "DDDs are not established for all medicines" and "DDD is a unit of measurement and does not necessarily correspond to the recommended or Prescribed Daily Dose". It also corresponds to the "its main indication in adults" -- I have no idea whether the main indication in adults for valproate is epilepsy or manic depression and nor are our readers told. The source doesn't say.

I don't think you are carefully reading what the source does and doesn't include. James continues to talk of a "wholesale cost" as though that is one thing for one drug (e.g. valproate). It isn't. The source gives, for example, an indicative cost of buying bottles of 100 500mg tablets in bulk. There's a different price for 150mg tablets and these are approx 2x more expensive per mg than the 500g tablets. The syrup is even more expensive per mg of drug. These aren't reference prices for "the developing world" but really just indicators from whatever sources they could manage to get that year. So the 200 mg/5 ml syrup only has the Buyer price for South Africa. This isn't a price anyone can buy -- they didn't find any Suppliers with public prices at all. It is the unique price the SA government negotiated and reflects only their contract. And the syrup formulation of valproate is important, as that is the primary form used to treat epilepsy in children.

James, you claim now you used the defined-daily-dose to work out the prices in articles. But as I worked out above, the price you included in the articles does not match that figure. I see you have updated the valproate article. You haven't explained why the figures were wrong before.

What is happening here is a gross over simplification of the issue of "cost". It isn't "developing world" it isn't "one cost" it isn't "lowest" or "highest" and it isn't "per dose" or "per day" in any actual patient and it isn't "one indication/use". A drug has multiples of all these and you are trying to find one number.

WhatamIdoing, this reminds me of Brexit -- all things come back to Brexit in the UK. A slim majority of the UK voted to leave the EU but nobody was told what kind of Brexit we would get. Some promised we would stay in the customs union or the single market, others said we could leave both and still get all the benefits for free. It now turns out that there is no form of Brexit that a majority would be happy with, though it seems like we are heading for a no-deal Brexit in December 2020. You might all wish to include "the cost" as a nice simple figure for readers. But when you actually look at the sources at the fundamentals of how drugs are delivered, indicated, dosed and used, there isn't one cost. -- Colin°^Talk 09:32, 26 November 2019 (UTC)

Colin, we're kind of spread all over the map here, and it might be useful to get a little more precise. Whether any specific source is being accurately represented in a specific article isn't really MEDMOS's concern. MEDMOS, like all manuals of style, is primarily concerned with how you say something, after you've already decided that it belongs in the article in the first place.

The things that should be covered by MEDMOS are things like these:

• Is this something that should normally be included (e.g., similar to side effects; we do not require any sort of extraordinary coverage of side effects to put those in a drug article), or only under certain circumstances (which should be listed)?
• How much precision should be included? The US Medicaid program says that the National Average Drug Acquisition Cost for 250 mg capsules of valproic acid is running between 15 and 25 cents per pill right now, but it reports it to the one-thousandths of a cent. How much do you think is appropriate for most circumstances? Dollars, dimes, pennies, fractions of pennies? The first two significant figures?
• Should costs be presented in monetary terms, or in real-world terms? For example, do you say "Recovery costs US$n in lost wages" or "People can usually return to work after n weeks"? This question is not at all specific to drug costs. Assuming that ideal sources exist, would we rather say that:
  • depression represents a drag of US$n trillions on the global economy, or that
  • it results in US$x in medical costs, US$y in increased morbidity, and US$z in excess mortality, or that
  • depression causes in US$x in medical costs, worsens the quality of life, and shaves y years off the lifespan of affected people?
• When you describe a product or service with significant variations, how do you describe it? There's a lot more variation in laptops than there is in valproic acid, but Laptop#Market Share lists an average sales price. How could we do better?
• Which prices belong where? Presumably, if you're on this kind of list, then it goes in the lead plus elsewhere in the article. But in other cases, where's the best place? Does it depend upon the nature of the information, e.g., with recent prices going in ===Economics=== and original list prices going in ==History==?
• How should prices be described? Consider this business source. WP:MEDPOP approves of the popular press for financial information, and that's what this is: Prosecutors said that Fagron's subsidiary "typically sold the fluticasone propionate for about $160 per gram". It's a generic drug, the daily dose is the same for (almost) everyone, the price only applies to the US, and it's specific to one company. What else would you need to know to describe that price? What about the price for Asfotase alfa, which is only made by one company and whose dose varies by weight [6]? There are a lot of ways to calculate prices for drugs in particular, so which one would you use, and how would you explain it, if you needed to include that information in an article?

Completely apart from the question of whether prices should be in many articles or few, I think we could do some good work here in describing how to write about them. WhatamIdoing (talk) 00:12, 27 November 2019 (UTC)

WhatamIdoing, the discussion on the MSHPriceGuide is ongoing at WP:MED so perhaps that's a better place for it than here. The questions you ask may be valid. But I'm interested in the text in actual articles right now being very very wrong, rather than specifying about how may digits of precision some hypothetical price might be. We have an awful lot of basic "text is not supported by the source" problems at the moment, such as a claim that the price offered by a supplier of Ethosuximide only in the Democratic Republic of Congo is "the price in the developing world". It is all well and good speculating about how we might describe prices, if we can't actually do that given the sources we have. -- Colin°^Talk 09:01, 27 November 2019 (UTC)

• 500mg tablet median cost x DDD = $0.40
• 200mg tablet median cost x DDD = $0.52
• 150mg tablet median cost x DDD = $0.88
• 200 mg/5 ml median cost x DDD = $1.05
• 250 mg/5 ml median cost x DDD = $1.18

James claims the "The wholesale cost in the developing world is about US$0.40 per day as of 2015." But that is just for enteric coated 500mg tablets from IDA Foundation. The price in any other form or for other suppliers or in some developing nations is wildly different, even if we misuse the DDD in the way James wants (which is Original Research). -- Colin°^Talk 09:47, 26 November 2019 (UTC)

Introduction to Drug Utilization Research 2.6 Drug costs lists several cost metrics:

• total drug costs
• cost per prescription
• cost per treatment day, month or year
• cost per defined daily dose (DDD)
• cost per prescribed daily dose (PDD)
• cost as a proportion of gross national product
• cost as a proportion of total health costs
• cost as a proportion of average income
• net cost per health outcome (cost-effectiveness ratio)
• net cost per quality adjusted life-year (cost-utility-ratio)

These are all the possible costs we might consider in the encyclopaedia. James has (apparently, though not supported by any evidence) chosen cost per DDD. Here's what WHO say about that: "For example, the cost per DDD can usually be used to compare the costs of two formulations of the same drug. However, it is usually inappropriate to use this metric to compare the costs of different drugs or drug groups as the relationship between DDD and PDD may vary."'. So my above table showing that 500mg is the cheapest per DDD and the 250mg/5ml syrup is the dearest is a valid usage. But it is only for one formulation and it would be wrong of us to use it to indicate "the cost of valproate" compared with eg. "the cost of carbamazipine".

Problems using the defined daily dose (DDD) as statistical basis for drug pricing and reimbursement: "The defined daily dose is an artificially and arbitrarily created statistical measurement used for research purposes in comparing the utilization of drugs.", "in most cases, the DDD differs greatly from the typical PDD of the drug in question. In some cases, this gap may be exacerbated by the fact that a drug may be prescribed in two vastly different dosages and the DDD represents the average of those outliers." The paper goes on to explain why cost per DDD is a misuse of the metric.

From these and other sources, it seems that presenting a cost per DDD in the lead of our drug articles would be exactly the misuse of DDD that WHO and others caution against. It is an arbitrary unit for doing research into consumption. Presenting this information in our leads is a clear example of why we do not allow original research. -- Colin°^Talk 10:13, 26 November 2019 (UTC)

IDA Foundation sells at these rates to more than 130 LMIC countries. So this represents the approximate wholesale cost in a lot of places.

Per WHO Defined daily dose "is the assumed average maintenance dose per day for a drug used for its main indication in adults."[7]

So this is the approximate or typical dose in adults.

Additionally $0.40 per day is approximately the same as $0.82 per day.

What it is NOT the same as is $638 per dose.[8]

If these approximates both you I would be happy to switch to "generally less than US$1 per day", "$US1 to 2" "2 to 5, 5 to 10, 10 to 20, 20 to 50, 50 to 100, 100 to 500, 500 to 1000" etc.

Sure there are other numbers we could also provide. And I would welcome you finding sources and doing so. Doc James (talk · contribs · email) 14:42, 26 November 2019 (UTC)

James, you keep mentioning IDA foundation as though somehow that makes your numbers right. IDA isn't even listed as a supplier for all formulations. If you want to just use a single source for pricing, then say "IDA Foundation sell ..." rather than misleading our readers that this represents developing world prices. If the IDA foundation price was all that mattered, then there would be no need for the MSH Price guide.

You are just plain wrong about DDD. Please read what I wrote and linked to.

No, $0.40 is not the same approximately as $0.82 per day. Remember these are wholesale prices. The average daily wage in India is $5.50. So a tablet at $0.40 and a syrup at $1.18 + retail markup are hugely different prices to someone in the developing world. You are just waving your hands about at the moment with this "these are pocket change numbers" comment. It very much looks like you don't care about the actual numbers... perhaps that's why every example I looked at is wrong.

I've no idea what you linked to that was $638 per dose. The link doesn't indicate any drug name. I think we can agree cents are not the same as hundreds of dollars, but if you want to make a point about a treatment being cheap vs extortionate you need a source making that point.

I have looked at lots of sources. James, you are missing the big point. When you say, for example, "IDA Foundation sells at these rates" you miss out the "sells what". What are they selling? They are selling 500mg tablets in bottles of 100 in bulk orders. You have arbitrarily chosen just one formulation of this drug and quote that price. It is as daft as me saying "The price of strawberry yoghurt is 50p". I didn't say how much yoghurt and there is no one size for yoghurt and there are all kinds from Greek style to French set to Organic.

What you want to say "The wholesale price of valproate in 2014 was $...." cannot be said. It makes no sense. You are conducting original research in an area well beyond your competence, even if we allowed editors to do so.

James, there is a reason why none of our sources, in their leads, say things like "The wholesale price of valproate in the developing world is $0.40". They don't say it because it is nonsense. Instead, we have some sources that give raw data for prices for specific product items with bar codes.

I think perhaps we need to open this discussion up to parts of Wikipedia with mathematical competence, rather than just WP:MED. -- Colin°^Talk 15:35, 26 November 2019 (UTC)

When you are interested in having a civil discussion let me know. Otherwise please do not ping me. Doc James (talk · contribs · email) 15:53, 26 November 2019 (UTC)

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Restart

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

---

Discussion above seems to have gone into the weeds a bit. I believe the best way forward might be to clearly understand the purpose and context of including pricing. For example:

• Increases in the wholesale price of insulin, noted above, are clearly a pressing public health concern and noted by large numbers of sources.
• Variations in costs of drugs between nations may be a matter of public of concern (e.g. salbutamol / albuterol at $30 - $60 per inhaler according to various sources whereas the NHS estimates the annual cost of treatment to the NHS at about £22, with a per-item prescription fee payable by working-age adults of £9 per item). I would suggest this is most relevant on drugs on the WHO essentials list, where there is a notably wide disparity as reflected in independent sources such as news articles.
• The price itself, if it is not exceptional, would not seem to be relevant or encyclopaedic, unless it is specifically called out as extraordinary (e.g. a new cancer drug that costs hundreds of thousands of dollars for a course of treatment but does not yet have a sufficient body of evidence to unambiguously support use - a perennial problem in the UK - or perhaps aspirin, which is pretty much the cheapest drug in the world and is constantly finding new uses).

In other words, it seems to me that we need to show a reason to care about cost: either it's sharply increased or decreased, or there are large and important disparities between nations, or it's called out as extraordinarily high or low. In all cases the key is that sources we normally trust - serious mainstream newspapers, in-depth journalistic reporting by fact-finding TV shows, published journal articles - note the price as being a significant matter. What would people think of using a benchmark expectation of three or more serious sources mentioning the price or one devoted entirely to the price as a significant concern or factor? Guy (help!) 11:43, 5 December 2019 (UTC)

Addendum The questions I think I'd like to be able to answer are:

1. Should we include pricing as a matter of course, or only where it's been discussed as a significant issue by RS?
2. If as a matter of course, based on what part of WP:5P?
3. If we should include as a matter of course, how should we manage the problem of global variation?
4. If we should include only based on RS commentary, what level of commentary do we need?
5. If we include pricing, should it be only from secondary sources or is a primary source acceptable or (because it may be more accurate) preferred?

I think if we have confident answers to these questions, everything else comes out in the wash. Guy (help!) 12:38, 5 December 2019 (UTC)

Guy I fully agree with you that only exceptional pricing is encyclopaedic. This is also WP:NOT policy. And also WP:WEIGHT policy -- if nobody is commenting on the wholesale price of a drug in DRC vs the retail price in the US, then neither should we. James, unfortunately, contests that WP:WEIGHT is supported by database records of raw price data, or by the general concern about drug pricing means it has WEIGHT in every drug article.

However, the only reason this conversation was on this page was because pricing was very recently added to MEDMOS and QuackGuru started edit warring to support James's position of putting detailed prices in the lead. WhatAmIDoing suggested another forum might be better. So it continued at Wikipedia talk:WikiProject Medicine. If you want to start a new RFC on drug pricing, then that may be useful. WP:MED ignored the result of the last one, though. The issue of an RFC is similar to Brexit. The last RFC contained a lot of untruths about pricing sources and wishful thinking about prices being useful, and rather ignored the problem that we don't have sources of anything but raw data, and so require original research to make any statement. As folk say about damn lies and statistics, it is unfortunately a problem that one can make any point one likes about drug pricing merely by selecting the right database records to use. -- Colin°^Talk 12:18, 5 December 2019 (UTC)

Colin, You were doing so well right up to the point that you started attacking other editors again. Could you maybe not do that please? Guy (help!) 12:25, 5 December 2019 (UTC)

Guy I've given an example of behavioural problems and open agenda pushing in a sequence of diffs over at ANI.- Colin°^Talk 13:21, 5 December 2019 (UTC)

Basically, we're just saying to follow existing policy: NOT, POV, OR. --Ronz (talk) 19:17, 5 December 2019 (UTC)

NOT does not apply to medical content. QuackGuru (talk) 19:53, 5 December 2019 (UTC)

I'm afraid you'll find no consensus for that, or are you just trolling? --Ronz (talk) 20:11, 5 December 2019 (UTC)

I'm dying to know what that assertion is based on. – Levivich 21:37, 5 December 2019 (UTC)

@QuackGuru:, could you please explain the comment that "NOT does not apply to medical content"? SandyGeorgia (Talk) 18:04, 6 December 2019 (UTC)

Where does Not mention drug prices? One size does not fit all. QuackGuru (talk) 18:15, 6 December 2019 (UTC)

Your statement was "NOT does not apply to medical content"; your answer does not address that. Also, almost all of NOT specifically addresses topics just like drug pricing, and the idea that the page should list every one of those items explicitly is faulty. SandyGeorgia (Talk) 18:22, 6 December 2019 (UTC)

Where does the WP:NOT policy mention prices? Seriously? At WP:NOTPRICES:

An article should not include product pricing or availability information unless there is an independent source and a justified reason for the mention. Encyclopedic significance may be indicated if mainstream media sources (not just product reviews) provide commentary on these details instead of just passing mention. Prices and product availability can vary widely from place to place and over time. Wikipedia is not a price comparison service to compare the prices of competing products, or the prices and availability of a single product from different vendors or retailers.

– Levivich 18:37, 6 December 2019 (UTC)

And I agree with that. All the sources being used to discuss prices are independent. And we have sources that justify the importance of pricing information. Doc James (talk · contribs · email) 20:40, 6 December 2019 (UTC)

The problem with the wording above at WP:NOTPRICES is that there is no mention of WP:WEIGHT. We don't typically add information to Wikipedia based on one source. That guideline needs attention, although I suggest that the subsequent sentence does use a plural and clarifies the singular in the first sentence. The not a price comparison service seems to a definition of where WPMED practice currently falls. SandyGeorgia (Talk) 15:38, 7 December 2019 (UTC)

I fully agree with Guy and Colin on this. Exceptional pricing that attracted media coverage should be mentioned, but otherwise Wikipedia is not a drug sales catalogue. As I mentioned elsewhere, pharmaceuticals pricing is among the most opaque ones in the world – a medication will have a list price, a reference price, a benchmarking price, the effective price per payer/country, an out-of-bundle price, and many others. They also vary significantly – a drug with a US list price of $1000 and a regional reference price of $800 may be sold for an effective price of $200 or less after volume discounts, risk sharing agreements and/on bundling or capping arrangements. For this reason list prices are irrelevant countries where drugs are purchased by public payers. Listing US retail prices will usually be confusing to the international reader. Again, I'd like to remind Doc James that there is an entire world outside of the simple buy/sell reality of the US! — kashmīrī ^TALK 22:52, 5 December 2019 (UTC)

The World Health Organization states "Data on drug costs will always be important in managing policy related to drug supply, pricing and use." They list a few price metrics including cost per treatment day, month or year and cost per defined daily dose (DDD). Doc James (talk · contribs · email) 03:21, 6 December 2019 (UTC)

WHO doesn't dictate Wikipedia's content. It may be important, but it may not be important enough for what's been done or proposed for product pricing. Nor does it mean it is suitable for presentation in an encyclopedia article.

Best focus on Guy's initial comment and addendum. --Ronz (talk) 04:21, 6 December 2019 (UTC)

It is best to oppose Guy's proposal that is against WHO's position.

WHO confirms drug prices are important. Drug prices is basic information readers are seeking. QuackGuru (talk) 04:33, 6 December 2019 (UTC)

No one is against WHO's position. --Ronz (talk) 04:48, 6 December 2019 (UTC)

Excluding drug prices except in rare cases is saying drug prices are not important. WHO says drug prices are important. QuackGuru (talk) 05:10, 6 December 2019 (UTC)

QuackGuru, no it's not, it's just saying that Wikipedia does not play a role in drug policy. Guy (help!) 09:10, 6 December 2019 (UTC)

Yes, it is saying that drug prices is unimportant. Essential information is drug pricing according to WHO. Excluding essential information is against WHO's position. QuackGuru (talk) 12:56, 6 December 2019 (UTC)

QuackGuru, no it's not. It is saying that drug pricing is important to some purpose other than an encyclopaedic description of the drug. Guy (help!) 14:22, 6 December 2019 (UTC)

• (from ANI) Re Guy's proposal, I think include the pricing only where it's been discussed as a significant issue by RS. The level of commentary should be "discussed as a significant issue by multiple RS"–more than just a mention, and more than just one source treating it in depth. Maybe 2+ RSes giving it significant coverage. The RSes supporting pricing should be secondary, but could be complemented by primary sources in appropriate circumstances. For example, 2+ secondary sources discussing the increase in insulin prices in order to include it in the article, but we might supplement with a primary source to give the most up-to-date price. – Levivich 05:03, 6 December 2019 (UTC)

Being mentioned in a single reliable source is sufficient IMO. There is discussion of prices for every medication more or less but some is easier or harder to find.

For example lenalidomide and bortezomib this WHO text says "Monthly pharmacy costs included in the total monthly cost in the unadjusted analysis were 4101 $US (SD 1931) and 4855 $US (SD 2431) for lenalidomide and bortezomib, respectively."[9] as part of a 6 page discussion of costs.

This is done for all medications added to the WHO Essential medicines list. Doc James (talk · contribs · email) 06:14, 6 December 2019 (UTC)

Doc James, If the price is the primary focus of an independent RS, then I would agree. If it's mentioned as a side issue, then I'd look for more than one. I don't think that's unreasonable, do you? Guy (help!) 09:11, 6 December 2019 (UTC)

User:JzG my thoughts on your questions:

We have plenty of sources that discuss prices.[10][11] Doc James (talk · contribs · email) 20:44, 6 December 2019 (UTC)

1. Should we include pricing as a matter of course, or only where it's been discussed as a significant issue by RS?
   • WHO states prices are an important aspect of medications and thus IMO we should generally include content on this when covered by a reliable source.
2. If as a matter of course, based on what part of WP:5P?
   • The content is encyclopedic. The cost of a medication is a core part of cost benefit analysis and a key aspect of public health.
3. If we should include as a matter of course, how should we manage the problem of global variation?
   • There are generally three prices ranges globally. That for LMIC and listed as international reference pricing. The US wholesale price of which there are a number but in my opinion the National Average Drug Acquisition Cost from Medicaid is a good source.[12]
4. If we include pricing, should it be only from secondary sources or is a primary source acceptable or (because it may be more accurate) preferred?
   • We have textbooks that list these as well as WHO and the US government. Doc James (talk · contribs · email) 06:22, 6 December 2019 (UTC)

     Doc James, Cost-benefit analysis is not part of Wikipedia's remit, though. We're an encyclopaedia not a textbook or prescribing guide. You and I agree that drugs such as insulin, where costs and cost disparities are identified as a pressing public health concern, should unquestionably be included. But I'm not persuaded that just being on the WHO register qualifies to discuss cost unless other sources do too. Normally I would not expect it to be particularly difficult to find sources discussing cost disparities for essential medicines, given the prices paid in the US compared with other countries. If there's evidence that price, specifically, is a focus of RS coverage for a substantial proportion of the WHO list then I'd be more inclined to support your position. Can you show that? Not price books or anything else, but actual RS coverage showing that costs of WHO essential drugs are considered independently significant? I suspect this does exist and we may be able to word a consensus position for the WHO essential list. Guy (help!) 09:18, 6 December 2019 (UTC)

   user:JzG As I mentioned above we have a 6 page discussion in a WHO document of the price of lenalidomide and bortezomib among others. Would you consider that to be suitable for showing notability? Easier to get this information for newly added medications to the list as documents for when older medications were added are harder to find. If we limit cost discussion to only those picked up by the popular press we will end up with primarily US prices for new medications. And our content will become more a US perspective. Doc James (talk · contribs · email) 20:38, 6 December 2019 (UTC)

   we have a 6 page discussion Can you please link these discussions? --Ronz (talk) 20:45, 6 December 2019 (UTC)

Here starting on page 51[13] Doc James (talk · contribs · email) 02:49, 9 December 2019 (UTC)

Guy, James, this 'discussion' is an application for three drugs to be included in the list of WHO essential medicines (which they were this year). It is in WHO's archives and not AFAICS published elsewhere. I believe Guy asked for independent RS, not internal WHO documents in their archive. Naturally cost and cost-effectiveness is going to feature in that decision. And it is a hard decision because cost is hard thing to consider and weighing that cost against sickness is a hard thing to do. The document is 76 dense pages long and includes discussion of complex pricing models. Cost changes throughout the many years of treatment, there's relapse and some patients are refractory to treatment. The drugs are taken in combination. The descriptions in that document blow my mind and I can't begin to think how to explain the cost of treatment with multiple therapies simultaneously.

Yet on our article Lenalidomide we say "In the United States it costs about US$16,000 to US$21,000 per month". But what we don't tell the reader is that's a retail price, and most other drugs we've looked at cite a wholesale "cost". The Drugs.com article explains all sorts of discounts on offer, possibly as low as $25 a prescription if your insurance is good enough (though I may be misreading as I'm not familiar with US drug pricing).

And we juxtapose that in the lead with "In the United Kingdom this amount costs the NHS about £3,400 to 4,400." This cites the BNF which gives an "Drug tariff price". But hold on a second, shouldn't we be comparing retail prices? Firstly cancer patients in the UK get their prescriptions free (normally £9 for a month's supply). So the true juxtaposition of the US retail price is £0!! Yay for socialist healthcare. Secondly, the UK has a special arrangement price for this drug. These arrangements vary from year to year, but this article from 2014 notes that "The manufacturer proposed the patient access scheme, which provides a price discount: the NHS receives the drug for free for any patients remaining on the drug after 26 cycles.". But when bidding for the drug to be a first line treatment (rather than third), the patient access scheme discount was part of the reason negotiations took so long. This article in 2019 says "The drug’s list price in England is up to £4,368 a month, or over £52,400 a year per patient, with the confidential PAS discount cutting this substantially.".

Lenalidomide is one of those drugs that is on the borderline for affordability even in the rich UK and the high price issue is discussed in the body of literature. However, the statement "costs the NHS about £3,400 to 4,400 per month" was derived from original research reading simplistic drug pricing in the BNF. If a professional expert was writing about the cost of lenalidomide to the UK's NHS, they'd have known about the patient access scheme and explained that any official list price is not the true cost to the NHS, which in this case is confidential. Our wiki article currently has a very incorrect wholesale list price for the UK juxtaposed with a US retail list price, which I really don't know what proportion of the population are actually paying. Honestly, even in articles where price is notable, we still go and ruin it by doing original research on raw data sources. We slip up every time. -- Colin°^Talk 19:18, 9 December 2019 (UTC)

This may be the place to introduce Colin to US drug pricing. First, every different insurance company negotiates different deals, to such an extent that if one has a chronic condition, one may look to a specific insurance company whose policy is best depending on that condition. Independent insurance brokers gain clients by studying the pharmaceutical pricing structure of each company so they are in good position to offer recommendations. Ditto with Medicare supplement plans (our social security does not cover all medical costs, so many buy supplemental insurance to cover drugs). We personally pay $ 000000 (ZERO) for many prescriptions, because we chose the right plan.

Second, almost EVERY pharmacy I know of has what is called a $4 plan ... that is, regardless of the retail, wholesale, or whatever price, they offer a 30-day prescription to many common drugs for only $4. This is marketing-- to get customers in the door. They similarly offer 90 days for $10 for many drugs, and their plans are typically posted right next to where you submit your prescription.

Third, coupons are available all over the internet from specific pharmacies for specific drugs-- again, marketing to get clients in the door. (My husband saved hundreds of dollars recently by simply googling the drug he needed.)

And finally everybody with a phone uses the app, GoodRX, which finds the cheapest source of a given drug in any geographic location, and even coughs up the coupons. And by everybody, I mean ... I volunteer as a Spanish-language interpreter in a free clinic that serves hispanic migrants without insurance, and everyone with a phone knows about GoodRX-- even those who don't speak English. To help the uninsured save money, when a physician writes a scrip in the clinic, they first pull up the $4 list and if they don't find something there, they next go to GoodRX.

In short, prices are extremely variable, subject to many factors, and very few people pay retail. Having said that, epipen is expensive for everyone, which is why I brought up that example (had to do with shortages or something.) SandyGeorgia (Talk) 20:18, 9 December 2019 (UTC)

Further adding to the complexity is the variation between insurance plans for drug formularies. Insurance may make one drug free while requiring people to pay out of pocket for another. And I know from personal experience that the classification of a single drug can vary not only from one insurance provider to another, but between different insurance plans from a single provider (bronze, silver, gold, platinum). --Tryptofish (talk) 22:28, 9 December 2019 (UTC)

• The price of all things is important to somebody. I can find sources or indeed highly authoritative reports from respected bodies that say the price of ____ is important. Food, heating, housing, travel, clothing, broadband, alcohol, pet food, kitchens and bathrooms... I don't think the argument that the WHO say prices are important or even that some authority considers the WHO List of essential medicine prices to be important is actually a valid argument for including in an encyclopaedia. Someone mentioned Wikipedia's place in the information<-->knowledge spectrum. The problem with James's sources are they are raw database records. They tell you that a pack of 60 5mg wonderpam-sodium enteric-coated tablets from DrugMaker corp on the week ending 06/12/2019 cost $xx [insert some country-specific definition of an arbitrary "wholesale" price here]. What James has tried to do and claims we can do, is give 'a price' for 'a drug'. We can't. Above it is claimed "Drug prices is basic information readers are seeking" but read kashmīrī's comment above about all the different concepts of "price" in the US. James claims we can use the BNF for UK pricing. It is no better. It gives several prices for each specific thing with a bar code you might get from a pharmacy. It lists two prices, not one, one of which is indicative of what a pharmacy might pay, and the other an indication of what the NHS would refund the pharmacist -- but that explanation hugely hugely oversimplifies the actual pricing. These prices depend on whether the drug has several generic alternatives or only one supplier, whether the GP wrote the brand name on the prescription or not, they vary depending on supplier shortages, which are not uncommon. They vary in dose size, formulation, coatings, etc. I'm afraid the claims "we have ...." sources that would provide an easy "the price for wonderpam is $3 a day in the US" do not stack up to scrutiny. We have raw data and way way too much original research needed to even begin to simplify the price, and no one concept of "price" from country to country. -- Colin°^Talk 11:37, 6 December 2019 (UTC)

  Colin, it's a great argument for an article series on medical costs in $COUNTRY, though. Guy (help!) 14:25, 6 December 2019 (UTC)

  I love this argument. Why don't we add national prices to articles on cement, timber, steel, copper, gas, electricity, agricultural land, university tuition fees, public transport, bread, housing, postal services etc.? For sure there are UN sources that list many of these prices as highly important in some area of human activity. Why stopping at drugs? Let's make Wikipedia a registry of Very Important Prices! — kashmīrī ^TALK 17:49, 6 December 2019 (UTC)

• Reliable sources do say it is important and reasons have been given for including them. Prices of the same drug vary in different countries.[14] Drug price monopolies raise drug prices.[15] QuackGuru (talk) 12:56, 6 December 2019 (UTC)

  QuackGuru, also true of cars. You are framing opinion as fact and not citing policy to support it. Guy (help!) 14:19, 6 December 2019 (UTC)

  I cited sources that support it is important. That is framing a fact as a fact. QuackGuru (talk) 14:22, 6 December 2019 (UTC)

  QuackGuru, see WP:ITSIMPORTANT. You cited sources that say it's important to drug policy. Wikipedia has no role in that. Guy (help!) 14:26, 6 December 2019 (UTC)

  Drug prices fluctuate and there is a controversy over the increasing drug prices. QuackGuru (talk) 14:30, 6 December 2019 (UTC)

  QuackGuru, which is either significant or not depending on the importance of the drug, the amount of the increase, and other factors. And we establish that from reliable secondary sources about the drug specifically. What we don't do is to decide that because prescription drug prices in the United States is an important topic, so the price in the United States is important to every drug. Guy (help!) 14:50, 6 December 2019 (UTC)

  Drug prices vary globally for the same drug. This has led to a controversy. Therefore, it is important to include the differences in drug prices. QuackGuru (talk) 14:56, 6 December 2019 (UTC)

  QuackGuru, Your logical fallacy is: begging the question. Guy (help!) 11:28, 7 December 2019 (UTC)

• (Saw this on ANI). Not my area of expertise, but I just wanted to challenge the argument that because the WHO considers that "Data on drug costs will always be important in managing policy related to drug supply, pricing and use" it does not follow that "Data on drug costs will always be important in informing a reader of an encyclopaedia article about a drug" which is our remit. We might decide it is, we might decide it's not, but as (hopefully) our articles will not be used to manage policy re drug supply, pricing and use, we should consider things by our own requirements. Scribolt (talk) 13:58, 6 December 2019 (UTC)
  • Indeed it is framed that "WHO say X therefore Y". For example James says "[WHO] list a few price metrics including cost per treatment day, month or year and cost per defined daily dose". Yes they do list them, but that doesn't mean they recommend them, or that one has to be very careful when using one. For example the "cost per DDD" is extremely contentious and WHO warn that DDD is essentially an arbitrary number for population studies. They say "the cost per DDD can usually be used to compare the costs of two formulations of the same drug. However, it is usually inappropriate to use this metric to compare the costs of different drugs or drug groups as the relationship between DDD and PDD may vary." (PDD is the prescribed daily dose for an individual with a particular condition). So you can use the arbitrary DDD to do some maths to compare that a person needing 100mg DDD with two 50mg doses a day would in the UK cost the same as someone taking 100mg once a day (because the few UK prices I've seen seem proportional to dose -- not claiming that as a fact) whereas in the US it would likely be twice as expensive (again, based on my experience with US prices being the same for various sizes, though with some odd values here and there). Of course all I just did is original research and just because WHO say an expert performing Drug Utilization Research (the source James cites) could do that sort of maths, doesn't mean that Wikipedians can or should. -- Colin°^Talk 16:28, 6 December 2019 (UTC)
• Guy, you might be starting to realise that there are two sides to this who are completely talking past one another. One side is trying to frame an argument based on polices, on what individual drug articles might warrant, and what they might reasonably expect editors to be able to say based on very limited sources, and an appreciation of just how complex the issue of cost is. The other side is framing their argument based on a political agenda against Big Pharma secrecy, price extortion, etc, based on feelings around drug costs in general, and never budges from considering that cost is a simple basic thing all readers want to know. It is like placing a Brexit party MEP in the same room as a Lib Dem MEP. Each have totally separate value-systems when it comes to emotion/truth/fairness/equalty/etc,

Further, as we know from Brexit, the Devil has the best lines. A punchy dishonest/misleading soundbite zings whereas you need a five-paragraph BBC Fact Check web page explaining why, in fact, it isn't as simple as that and no there never were going to be 40 hospitals or 20,000 police officers. The Truth is complex and messy whereas Lies are simple and easy. I don't have a solution for that.

If we keep talking past each other, as has happened with me and James, and with Ronz and James, then we will not achieve consensus, and any poll in the end will just becomes a numbers game. It may result in another no-consensus result, and it may once again result in a problem where one side says things the other side regard as patently false or obviously irrelevant. Further, a poll merely asking "should all our drug articles have prices", still doesn't solve the problem of which prices and how do we get them without original research or relying on now defunct websites or over-simplifying. When you have editors who are totally happy with juxtaposing wholesale prices from 2015 with retail prices from 2019, the problem is bigger than "should we have prices". Like politics, it is all well and good if a majority want the moon on a stick, we need some agreement on what those drug prices might look like, and acceptance that the methodology used to get them passes our fundamental policies. I don't think that has been demonstrated yet, not even close. -- Colin°^Talk 16:28, 6 December 2019 (UTC)

I think there are sound (and source supported) arguments in both sides, that's why this issue is so difficult to solve. We need to be careful and open in our approach here, depending on how we work it out it may be beneficial to the encyclopedia, if only just to create new articles dedicated to the subject. Signimu (talk) 20:11, 9 December 2019 (UTC)

Regarding the importance of pricing: Where is it covered in depth in Wikipedia? Medication costs, Prescription drug prices in the United States, Prescription charges. I expect there are more. There seems to be some disorganization and outdated links currently in articles. Regardless, we clearly think pricing is important and are treating it so. --Ronz (talk) 17:58, 6 December 2019 (UTC)

Thanks for the restart, JzG. I went away over the US Thanksgiving holiday, and lost track of the discussion here (will not try to review the content added since I last weighed in, and appreciate the restart). Colin, henceforth please let the bigger problems speak for themselves (ala ROPE); I believe they are self-evident. I agree with JzG, Colin, Levivich, Ronz, and Scribolt (because their positions address Wikipedia policy). I also agree with Kashmiri (an unusual position for me). I disagree with Doc James and QuackGuru (because those positions do not encompass policy). I also submit that with so much dangerously defective information on Wikipedia, WPMED should get back to working collaboratively to improve content rather than focusing on adding information that is not policy compliant. SandyGeorgia (Talk) 18:13, 6 December 2019 (UTC)

MEDRS is not policy. It is a guideline like this page. QuackGuru (talk) 18:17, 6 December 2019 (UTC)

QuackGuru, it has much greater weight than this page. Guy (help!) 11:28, 7 December 2019 (UTC)

QuackGuru, the relationship is that guidelines explain how to apply policy. Guidelines cannot/should not deviate from policy. SandyGeorgia (Talk) 15:33, 7 December 2019 (UTC)

We have an overview on the relationship between the different price estimates used in the United States here. Indicates as we all agree there is more than one way to present a price. Even those prices for medications are rough figures, they are still useful. $US10 is very different from $US10,000 or $US2.1 million. Doc James (talk · contribs · email) 20:49, 6 December 2019 (UTC)

Try to register that drugs are mostly available for free in much of Europe and parts of Asia and Africa and have entirely different retail prices in most of the developing world. Why would anyone want to stick your American deliberations on Medicaid, etc., into a global encyclopaedia? If a resident of the US, India or Ghana is interested in a particular drug price, he or she can always look up an online pharmacy in own country or ask in real world around the corner. The differences in drug prices between countries can be even 1000-fold. — kashmīrī ^TALK 23:51, 6 December 2019 (UTC)

In our struggle to GLOBALIZE, are we ignoring how drugs/medical products are typically supplied in the US? Insurance companies negotiate drug and medical product prices and the differences within the US can be significant depending on insurance company. How many people pay retail price in the US? SandyGeorgia (Talk) 15:44, 7 December 2019 (UTC)

"up to 90% of the population in low- and middle-income countries must pay for medicines out of pocket" For the United States out of pocket costs are about 48 to 67%.[16] Doc James (talk · contribs · email) 02:46, 9 December 2019 (UTC)

Thanks for this, Doc, but ... First, the best I can tell, that is 1998 data, which is useless post-Obamacare. Second, the full text of the cited source about US out of pocket costs reveals another layer of complexity: age (not surprising considering how insurance plans work). Annual average out-of-pocket prescription drug expenditures for all adults are $177, but people age 65 and older pay much more for their medications. People age 65 to 79 pay $456 out-of-pocket. People age 80 and older pay even more (see Figure 4). Adults pay almost half — 48 percent — of their expenses for prescription drugs out-of-pocket, but persons age 65 to 79 pay 56 percent and those age 80 and older pay 67 percent of their total drug expenditures out-of-pocket. Doc, with this much variation in drug prices in the US, and this new layer of complexity, could you please take one clear example and plainly explain what numbers you are using, how they can be representative of anything, and how they are not original research? As you are the person seeking to make this change to policy, it would help all of us and expedite matters if you would clearly explain how the data you want to use relates in any way to prices people actually pay. SandyGeorgia (Talk) 03:21, 9 December 2019 (UTC)

Sure. The DDD for valproate is 1.5 grams. The external reference price is 0.1339 per 500 mg tab. 90 tabs per month at this tablet size. So about 12 USD per month at DDD.[17] Doc James (talk · contribs · email) 23:50, 9 December 2019 (UTC)

You didn't answer Sandy's question, which asked you to explain how what you are doing is not original research and why you picked numbers that represent anything useful. You picked the 500mg tablet which has only one supplier. The 200mg tablet has five suppliers, which suggests to me you picked the wrong one. Doses above 250mg should be taken as divided doses according to Drugs.com. A 500mg enteric coated tablet cannot be split, so isn't very useful in practice while titrating dose from an initial 700mg, say. You've been misled by the 1.5 DDD into thinking that is nicely 3 tablets, when real living patients might be on any dose from 700mg to 2.5g and so require smaller tablets. And 500mg is essentially useless for paediatric epilepsy, which is a major use for this drug. You claim "the external reference price". That's just not true, James. The MSH can be a source of external reference prices, plural, and that 500mg tablet is weak in terms of data strength, in only having one supplier. It is "an external reference price", chosen arbitrarily, and the difference is huge and to say "the" is wrong. There are other prices, for syrups and other tablet sizes, but only the 200mg tablet seems to be widely available in the developing world. The 200mg tablet doesn't divide perfectly into 1.5, which itself is just an arbitrary average figure. Lets assume 1.4 for a real patient, which is 7 tablets. That's $0.4865 per day according to your original-research method, and $14.60 per month (or $16.68 for 8 tablets/day). And you don't have 2018 or 2019 data, because the MSH price guide appears to be discontinued. -- Colin°^Talk 09:46, 10 December 2019 (UTC)

"The MSH can be a source of external reference prices, plural" I think there is a misunderstanding. When sources or I guess editors here refer to "MSH reference price" singular, they often/always refer to the median supplier price, which is the recommended methodology by the WHO/HAI. An example is the Lancet paper. The MSH median supplier prices are the basis that was used to build the subsequent WHO/HAI medprices database which uses them to compare drug prices internationally and between generics and original brand packaging. The use of MSH median supplier prices as the basis for surveys, analyses or policies is widely accepted, essentially any document mentioning the use of the "WHO/HAI methodology" is doing that [18][19]. --Signimu (talk) 13:36, 10 December 2019 (UTC)

Signimu, to pick the valproate example, you are right we can have an external reference price for 200mg tablet using the median of the five supplier prices. This is what the MSH guide suggest to use. We could, in negotiations with a supplier, also cite each of five records as examples of external reference prices. I have said this multiple times, and am pleased you say so because so far I've been a lone voice in the wilderness on that one. Our articles mostly use a lowest and a highest price taken from both the supplier and buyer price lists. Gasp with me in statistical horror. Sometimes there are no suppliers for a particular drug size (or at all) and only buyer prices. Sometimes there is only one supplier or one buyer. Sometimes the supplier is international but sometimes they only supply one country (e.g. DRC). We have consistently cited MSH as an authority of "developing world cost" even in this very suboptimal cases. But what the MSH can't give us, is an external reference price, singular, on valproate. I am not aware of any methodology approved by WHO that allows us to select which pill or syrup, or whether to use sodium valproate or valproic acid, or whether crushable or enteric coated. These all have their own price. It has been repeatedly claimed that the MSH data can give us an international reference price or an external reference price or a price in the developing world for "a drug". And of course, the next step after that is to conduct original research to get a treatment price. -- Colin°^Talk 16:54, 10 December 2019 (UTC)

I agree, using the range of min-max without providing the median is meaningless, the median is more informative (and more commonly accepted metric) than simply the min-max, although the min is also used in policies (for obvious reasons of trying to pressure to drive prices down). Yes, mixing supplier and buyer price should not be done, but it's unclear which one we should prefer for an encyclopedia, maybe both have their place if it's clearly described along with the pricing. About I am not aware of any methodology approved by WHO that allows us to select which pill or syrup, or whether to use sodium valproate or valproic acid, or whether crushable or enteric coated, there is one, which is to cluster biosimilar compounds according to the ATC classification. The WHO/HAI database, and most databases and studies in fact, use the ATC or a similar classification to cluster products. Comparing prices between biosimilar ATC (what I mean by biosimilar is explained here) is often used by studies and policies alike to compare generics vs originator brands, and sometimes to compare prices internationally, and there is evidence that clustering biosimilar products allows to actually have a better representation of prices (so it is methodologically advised to do so), so this may be a possible solution. Signimu (talk) 17:14, 10 December 2019 (UTC)

Signimu, I'll reply in your other post mentioning HAI. I would really really have preferred if wiki could have a bog standard source->text discussion like you and I are doing, where novel sources are offered, and advice about statistical approach given, and an agreement reached. We haven't had that to date, but it would help narrow down the choices of acceptable prices before any RFC on them. -- Colin°^Talk 18:37, 10 December 2019 (UTC)

Extended content

Wrt the 90% figure "out of pocket". Firstly, when you were doing your Black Friday Christmas sales shopping, and the store said "Up to 50% off all prices", what does the "up to" bit mean? It's a get-out-of-jail-free statistic. Let's look at WHO's source, which is another WHO document that says "In many low-income countries in particular, private out-of-pocket spending accounts for 50%-90% of pharmaceutical sales.". So that would be 50-90% then. And these figures are more than 10 years old.

Secondly, "out of pocket" is a retail price. The WHO source says "Duties, taxes, mark-ups, distribution costs and dispensing fees are often high, regularly constituting between 30 to 45% of retail prices, but occasionally up to 80% or more of the total" But all our developing world costs for drugs are given as a wholesale price (and not even a well-defined one, see MSH for the two kinds of prices they have). If we are concerned to supply detailed dollars and cents "out of pocket" prices for drugs then here's what WHO says about MSH: "Medicine price indicator guides [they cite MSH] provide the sales prices from large wholesalers of generically equivalent medicines to governments. However, they do not give the price patients must pay in either the public or private sectors and often do not include new, essential but patented medicines."

That WHO document also says "Prices of the same medicines frequently vary between countries" which makes it hard to justify citing a price in Democratic Republic of Congo in 2014 or the price in Costa Rica in 2015 and claiming this is "the price in the developing world" (which we do). Wrt MSH it has been claimed "what we are using is an accepted international reference price" (singular). But MSH provide "international reference prices" plural. These are multiple prices that then require expert care to use properly. They explicitly state that the Buyer prices "should not be used as international reference prices", yet a Buyer price is frequently given in our articles. So MSH is a source of IR prices, plural, if you know what you are doing, that has some value to someone -- mainly the folk who cite them (plural) in order to negotiate a better price for their government. The closest we might get to "an international reference price" for a 100mg tablet, is to cite the median of the Supplier prices (which is what MSH recommend) but that only has value if the sample size is reasonable (perhaps half a dozen suppliers, and not just one). And that's just the price of a 100mg tablet, not "the drug" in general. The dose-cost ratio of higher and lower dose tablets varies considerably between countries.

Returning to the WHO document, and "out of pocket" prices. WHO's 2007 survey concluded "medicine prices are high, especially in the private sector (e.g. over 80 times an international reference price); availability can be low, particularly in the public sector (including no stocks of essential medicines); government procurement can be inefficient (e.g. buying expensive originator brands as well as cheaper generics); mark-ups in the distribution chain can be excessive; and numerous taxes and duties are being applied to medicines."

WHO give two examples:

• "Salbutamol inhaler – an important medicine used to treat asthma – is virtually unavailable in the public sector of many countries (where medicines are generally cheaper or even free) and when purchased from the private sector, can cost the lowest-paid, unskilled government worker several days’ wages."
• "The price of originator brand atenolol 50 mg tablets is over 20 times the international reference price in all the countries except India (where it is still high at 5 times the reference price) and Kazakhstan. Even the lowest-priced generic is very expensive in all countries, and there are some huge brand premiums, e.g. in Uganda the originator brand is about 13 times the price of the generic."

I'm not getting a good feeling that the IRPs bear anything other than a tenuous "less than" relationship with the "out of pocket price" that "50-90%" of people in low income might pay. -- Colin°^Talk 09:02, 9 December 2019 (UTC)

References

Agreeing on WP:NOT

On the AN/I discussion, User:WhatamIdoing suggest that perhaps community feeling about WP:NOPRICES might have changed. However, look at the above discussion for the quote box in pastel shade. User:Levivich quotes the existing NOPRICES text and James replies "And I agree with that. All the sources being used to discuss prices are independent. And we have sources that justify the importance of pricing information". This is why I said above that it seems that two sides are talking past each other, and have different value systems that are framing their reading of policy. James thinks that raw database records in an MSH database (independent source) combined with general WHO statements that "drug pricing is important" (justified reason for the mention) together satisfy existing policy at WP:NOT for every drug article on Wikipedia. Most others here take the view that WP:NOT currently requires significant commentary about an individual drug, for example, price extortion due to limited generic suppliers in the US, or the high price of a barely effective cancer treatment not satisfying NICE that it is justified for use in the UK's NHS. So while we each continue to actually interpret current policy differently, it seems premature to suggest we have a vote to change policy. -- Colin°^Talk 11:57, 7 December 2019 (UTC)

No evidence has been offered for statements about the community being divided on the application of NOT to pricing, and I think all RfCs show strong support for NOT. --Ronz (talk) 15:54, 7 December 2019 (UTC)

This, I think, is a question that could be settled through an RFC. Is it enough to have multiple sources that justify the importance of drug pricing in general, or do we need a source that justifies the importance of pricing for this specific drug?

There are broadly three possibilities here:

• People think that sources about the importance of drug prices in general justify some prices for most/any drugs.
• People think that sources about the importance of a class of drugs justify prices for most/any drugs in that class. For example, if sources say that the price of antibiotics is particularly important, then we add information to articles about antibiotics, but not to other drugs.
• People think that (only) sources about the importance of prices of specific drugs justify prices only for the individual drugs named in the sources. For example, if the source says that the end-user price of aspirin is important, then we would write about end-user prices in the article on aspirin.

WhatamIdoing (talk) 07:39, 12 December 2019 (UTC)

WhatamIdoing, this is similar to the Q1 below. Firstly I don't think there is anything unique to medicines in this. Take the example I gave below about food. The price of food is hugely important both in terms of feeding one's family but also in production as it is an important source of income for much of the world. Indeed, for some foods, the price is commoditised and traded and speculated in markets. And the price of some foods is also of special interest. In Scotland, there is a minimum price for alcohol at retail, and a ban on multi-buy offers to try to address a drink problem. The price of healthy fresh fruit and vegetables vs junk food is also frequently discussed. In other countries, the price of the staple carb is key. The welfare and environmental issues surrounding meat are associated with its historically low price. And if one wants to, with Google, one can probably find documents discussing the price of any individual thing from the cost of Sainsbury's premium mince pies to Aldi's bargain Scotch. I don't see anyone campaigning that every food article should have "a price".

For example, every single drug in the WHO list of essential medicines will go through a proposal/approval process that involves cost effectiveness which involves some mind blowingly complex statistics to estimate the cost of treatment. James linked to one in the WHO archives for a group of three recent drugs; older drugs are probably not online. And every single new drug is assessed by the NHS and a price negotiated and the cost-effectiveness weighed in terms of whether it should be used vs older drugs. Is the notability of price in our articles dependent solely then on the happenchance that these memos and reports and applications are in an archive that Google can find and permitted for public access?

An RFC gauging opinions on all these thing may end up being largely pointless if simply subverted as seen in the above "I agree" to the existing WP:NOTPRICES. And I can't see your 2nd bullet "class" discussion being constrained to the format of an RFC if there are so many classes of drugs each with their own importance to someone. Elsewhere it was suggested for example, that a discussion of cheap/expensive drugs is likely more appropriate in an article on the drug class, and could then easily be sourced to secondary/tertiary sources. For example, at Anticoagulant: "NOACs are a lot more expensive than warfarin, after having taken into consideration the cost of frequent blood testing associated with warfarin" (unsourced BTW). This is a heck of a lot more useful than original research producing dollars and cents prices of each NOAC and warfarin and expecting our patients to compare for themselves, without the caveat that warfarin has hidden costs, and the efficacy and tolerability both also factor into cost effectiveness. This is why Wikipedia medical articles are built on top of expert knowledge and opinion, not our own editors doing their own research.

We have so greatly abused the raw-database online resources, simply because they are accessible and give an illusion of comprehensiveness, and this has led to the idea that we can put a price on any drug article. It is a mirage. Price is too complex for that, and this approach always requires original research. We need to take it a level up, stop the original research and presentation of essentially random numbers "accurate" to the cent to our readers, and offer them expert-sourced higher-level cost information. --Colin°^Talk 08:56, 12 December 2019 (UTC)

Colin, your response is one that a debate teacher would admire – it hits all the points in contention, so that none can be struck as out of play for the next round – but if we want to settle things, I really do think that we need several small RFCs, instead of a huge, multi-part RFC. One of the separable assertions is that "Sources: Drug prices are important!" is enough to "justify" (in the sense of NOTPRICES) prices in articles about individual drugs. Other editors say that's not the case, and that you need to have "Sources: The price of aspirin is important!" to add the price to Aspirin. (There is the possibility of some middle ground, and since RFCs these days tend to be mere votes, I think it's worth calling that out, by way of encouraging people to identify where they land on the spectrum of possibilities.) My point is: even if you're 100% right about the other aspects, they don't have direct bearing on this question. We could actually settle the dispute over this question. WhatamIdoing (talk) 16:00, 12 December 2019 (UTC)

WhatamIdoing, well maybe that set of questions is approaching things from the wrong angle. Most people already think WP:NOTPRICES is your option 3. We already have WP:LEAD, WP:WEIGHT and WP:NOR all of which firmly disallow what is being done in our leads of drug articles. For some reason, these policies are being chucked in the bin. By asking 1 and 2, you are just setting fire to the bin. A better way is to look at what people are trying to say in articles and examine if that is appropriate per these non-contentious policies. You raise the question of what to ask, and this is being discussed far far below, so perhaps this conversation is better done there. There is an open request at the NOR noticeboard for input on the NOR issue in particular. Currently not a lot of input on that, which is odd since examining source->text issues is basic wiki stuff. --Colin°^Talk 16:23, 12 December 2019 (UTC)

I think WAID proposition of asking this question (and with the proposed formulation) makes sense. We should separate this issue from NOR, as I will later and in a separate section show that it's possible to use each of these kinds of sources with 0 OR.

I can also provide some sources for each proposition:

1. Drug prices are important in general: [20][21][22][23]
2. Prices for classes of medicines: [24] review is quite comprehensive although not exhaustive, it covers generics vs originator, high-priced/orphan medicines, medicines for specific indications (such as HIV/AIDS, cancer, hepatitis C, cardiovascular, Alzheimer, etc), hospital medicines, non-prescription medicines.
3. Prices for specific drugs: either use databases to simply get a price, or papers to get an interpretation (eg, comparison between countries, evolution over time, etc), this chapter and the following ones provide lists of several studies per geographical regions[25]. Signimu (talk) 16:42, 12 December 2019 (UTC)

PS: in my mind, if WAID's proposition is integrated in the RfC, it should of course be accompanied anyway with concrete examples, else it's just too broad and impractical. --Signimu (talk) 16:47, 12 December 2019 (UTC)

Well we are repeating the discussion below, where I already suggested any classification of rules really needs to include examples of articles that would be excluded/included by that proposal. I think you will find editors willing to argue any drug meets any "rules" you care to invent. I still have a huge problem with the concept that a drug has a price, in the general case -- our sources clearly suggest not. So would be very interested in how you escape the NOR issues to date. Btw, I can't read the book you link and it seems from the abstracts I read to only confirm that claims such as "the price in the developing world is..." are going to be rubbish. --Colin°^Talk 17:00, 12 December 2019 (UTC)

Yes my PS was a callback to the discussion below I will go back there, I have finished reading new material to try to answer some of the questions you and others have pertinently raised. You can access all chapters using a website which name starts with sc Alternatively I can send it to you, it's really a great source material, we can mostly use only this source as a reference to help us make sense of the different possibilities, it contains pretty much everything about drug prices and is up to date (2019). --Signimu (talk) 17:22, 12 December 2019 (UTC)

Proposed questions for central RfC

Archived ANI leading to RFC

There appears to be consensus for a central RfC to settle this issue, with the RfC questions decided here.

please focus on the questions to be asked, not the answers you think the RfC should return.

Opening proposal

Q1. In articles on medicines, should pricing be included:

1. Only where cost has been the focus of significant discussion in reliable independent sources (excluding trade press);
2. For all drugs on the WHO essential medicines list;
3. For all drugs.

Q2. Where pricing is included, should it be referenced to:

1. Authoritative primary sources;
2. Secondary sources only.

Q3. Where pricing is included, should it be:

1. In the info boxes;
2. In the info boxes only where global variation can be documented, otherwise in narrative text;
3. Only in narrative text;
4. Via templated external links.

Discussion

Please suggest additional questions and options for the questions, or modifications to the questions or options. Again,. please focus on the questions we should be asking the community, not the answers you think the community should return. Guy (help!) 11:40, 7 December 2019 (UTC)

• As I note above, I think you need to clarify your first question to make it clear the "significant discussion" is about that drug and that drug alone. James appears to have interpreted the existing WP:NOPRICES requirement for justification and commentary on these details as being already satisfied in the general case for all drugs.
• For the primary/secondary sources question, again I think this needs clarified. Is the MSH international price guide, the National Average Drug Acquisition Cost (NADAC) weekly reference data, the BNF online, and Drugs.com (retail) a primary or secondary source? None of them are the actual supplier/pharmacy website. So in that sense, their information is secondary. Some of them have processed the data (the NADAC is an average) and some haven't (the MSH and BNF list each supplier, though MHS provide a median, and BNF an NHS reference price). The Tarascon Pocket Pharmacopoeia price in $, $, $, etc symbols is perhaps even a tertiary source, though in my view, mass reproducing their symbolic "rough guide only" price (whether symbolicly or by writing $50-100), is likely to foul copyright on databases as these are not hard facts. James is now claiming Tarascon's prices are "inflated" so he doesn't want to use it any more. So I'm not sure we have any examples of what many of us would regard as secondary sources that could be directly inserted into articles with just paraphrasing and no original research.
• The location question doesn't include "in the lead". Nor explain what detail. Perhaps some examples for each option would help. For example, many people might be very happy with the lead summarising economic information in the form of the words "cheap" or "expensive" or "very affordable" inserted into the description of the drug. They may however, not think that citing several $xx.xx prices in the lead is appropriate.
• I don't understand the second infobox question. Perhaps examples would help.

Overall, however, I think an RFC is premature though certainly required at some point. Surely before the RFC we could get 6-10 articles (including, say, the drugs already discussed here and at WT:MED) to a standard that meets our policy (OR, V, NPOV, etc) and sourcing guidelines (MEDRS if claiming treatment costs). Those articles could then be used as examples of "Is this what you want?" If those wanting prices can't even come up with example text or mock infoboxes that satisfy the community on the basics, then we are wasting our time having a divisive vote on the issue. Perhaps we could invite people who are currently respected in reviewing such policy and article texts to examine the proposed pricing lead, body and infobox, and for us to reach a point that says "OK, if Wikipedia is going to include drug pricing in every article, this is the high-quality example you might follow". After all, policy and guideline should be based on best-practice, not wishful thinking. -- Colin°^Talk 12:22, 7 December 2019 (UTC)

• Related to Q3, on including drug and medical pricing info in external link templates: MEDMOS used to recommend DMOZ specifically as an external link, but that was removed in 2018 because DMOZ no longer existed. The new CURLIE was never added back in.[26] Could we discuss or have a specific RFC question about adding a drug pricing template to external links for much the same reasoning we were using DMOZ/Curlie? Can pricing be handled similarly? SandyGeorgia (Talk) 15:57, 7 December 2019 (UTC)

  SandyGeorgia, As a refinement of Q3.4, yes? I'm for it. Guy (help!) 08:49, 9 December 2019 (UTC)

My suggestions:

Q4. Where pricing is included, what should be the geographical area covered?

• Any country that sources can be found for
• Only major economies, notably .... [to be agreed on]
• Only countries or areas where prices of particular drugs have become a matter of significant public interest (e.g., due to controversies)

Q5. Where pricing is included, what prices should be used?

• product cost (e.g., per package)
• treatment cost (e.g., per day/month/year/treatment course)

Q6. If per-package pricing is included, what prices should be used?

• Ex-factory prices (manufacturer's list prices)
• Wholesale (distributor) prices averaged using the following method: ... [please propose]
• Retail prices averaged using the following method: ... [please propose]
• Effective prices per package or per treatment (i.e., the money effectively received by the manufacturer per drug package sold, after all rebates, discounts, free packages, risk-sharing arrangements, etc.)

• Where effective prices cannot be calculated (e.g., due to bundle deals, budget capping or other RSS methods), should this be mentioned if confirmed by sources?

Q7. If US prices are primarily used, should they be accompanied by a note stating that US prices are not usually representative of prices in most other countries (as is the case according to many RS)?

• Yes, always
• Yes, if differences are indeed observed in pricing data
• No

— kashmīrī ^TALK 12:28, 7 December 2019 (UTC)

• The product/treatment choice needs to be a concrete one with examples that meet our policy. For example, if price per package (or per pill), then how does one select which pill and package size if a source offers many? James has, for example, claimed he chose the cheapest pill-dose to meet the Defined Daily Dose, but this also requires knowledge of how often the dose is taken per day, which also requires knowledge of which indication we are talking about, and requires agreement that the DDD is a reasonable measure to use [its an arbitrary value, contentious, and not available for all drugs]. In one example we had, the 100mg DDD would require a hard-to-get and oddly expensive 100mg pill in the US, whereas two 50mg pills would be much cheaper, provided you accept it is typically taken twice a day. None of this methodology is currently documented in the article -- why a particular dose of pill was chosen -- and there are strong grounds for arguing that this is original research. It gets worse of course when you go onto treatment cost. Again, none of our current prices-per-treatment (or monthly cost for ongoing therapy) indicate how this was chosen in the article/footnote and require original research to achieve it. A drug may be used for depression, anxiety and neuropathic pain, and we give a cost per month currently without explaining to the reader which indication that cost is for [there are no sources I have found, that indicate which condition the DDD is supposed to represent, and "the most common indication" may well differ from US to UK to developing world].

So offering the community a choice: Would you like cost-per-month for longterm use & cost per treatment for short use, as appropriate [again requiring OR to pick] begs the question: how do we achieve it. So I really think all of these question options have to be grounded in an in-article example, where some respected and knowlegable independent wikipedians, agree that it is reasonable to do this. Our RFC cannot rewrite the rules on original research or misuse of raw primary data, even if folk wanted to. -- Colin°^Talk 14:23, 7 December 2019 (UTC)

• Related to Q6 and Q7, in the US, insurance companies negotiate prices and pricing varies considerably depending on insurance. How relevant are retail prices in the US anyway, and if prices are included for the US, how do we factor in the price paid by most people per the largest insurance companies? How many consumers pay retail for medical products/drugs in the US? Let's take prostate cancer as an example. It occurs generally in older populations, so we get into medicare pricing and medicare supplement plans, and the issue that most elderly men with prostate cancer will opt for a medicare insurance supplement that covers cancer drugs/treatment. The variation in those plans is considerable. So what price do we use in this instance for the US? SandyGeorgia (Talk) 16:04, 7 December 2019 (UTC)

• Related to Q6. Just as a heads up, there is an ongoing switch to a pay-for-results pricing model (also known as outcome-based risk sharing), especially in new super expensive medical technologies (including gene and cell therapies) – i.e., payments are due only relative to the health benefit observed in each treated patient. Simple "let's add some price to each drug article" will certainly not capture this; and by not capturing, it will falsify the real treatment cost. I am mentioning this to show how complicated the area of drug pricing is once one starts looking beyond common medications. That's also why I oppose the simplistic arguments on some "WHO publication" coming from Doc James et al. — kashmīrī ^TALK 17:18, 7 December 2019 (UTC)
  • See the "WP:OR at Diclofenac"/"Diclofenac in more detail" at WT:MED, where James is starting to appreciate there is more than one "wholesale cost" and has now decided the US list cost is "inflated" so would prefer to use a price after discounts/rebates. However, on the MSH "developing world" source, the Supplier prices James cites as "international reference prices" are supplier list prices, not prices negotiated after discount/rebates, etc, and MSH reckon buyers should add 10% to those list prices to cover transport and other costs. The UK's BNF lists two prices but I'm not expert enough to describe what they mean, and a given pill doesn't always have both prices in the BNF. So a question of "Do you want A or B or C cost figures" might not be achievable for all of e.g. US, UK, developing world. This is why I really feel it is up to those wanting to include costs in articles, to demonstrate that the options they want to ask us about can actually be achieved with the rules of policy. -- Colin°^Talk 17:53, 7 December 2019 (UTC)
    • Yeah, I didn't really check all the sources, but from my own area, 10 x 2g oral tablets of salbutamol cost $55.47 in the US[27] and INR1.08 = $0.015 in India[28] - a 3,700-fold difference. Of course such differences would need to be captured, but do we really want Wikipedia to serve as a price comparison website? — kashmīrī ^TALK 18:05, 7 December 2019 (UTC)
• Developing world the only source I've seen used so far for "developing world" is the MSH Price Guide. It is a respected source of raw primary data. Though its data comes from a number of suppliers, any given drug or drug pill size, might have few or no supplier prices. We have examples given where "the price in the developing world" was actually just the price paid by the government of Costa Rica in 2015, or the price of a supplier who only supplies war-torn Democratic Republic of Congo in 2014. But bigger than that is that it was updated yearly but has not been updated since 2015. I have tried multiple ways of contacting them and GMail has informed me that the contact email address given on their website is not responding. Perhaps someone in the US can phone them to enquire if there is likely to be a future update. Maybe they can't afford yearly updates and are moving to 5-yearly. I'm just guessing. As we enter 2020, I don't think Wikipedia should be committing to sourcing prices from "the developing world" to a source that will be already 5 years out of date and stale. -- Colin°^Talk 17:52, 7 December 2019 (UTC)

It takes time to gather data. When I first began working with the source the data was from 2014. They have since updated to 2015. Doc James (talk · contribs · email) 02:52, 9 December 2019 (UTC)

The 2015 price guide was published in 2016, as one would expect, and "it takes time to gather data" does not appear to be the case. And it had been published annually since 1986. There has been no guide published since 2015. Something is up. -- Colin°^Talk 10:03, 9 December 2019 (UTC)

From my readings, I understand that the WHO was tasked in 2001[29] to monitor more actively pharmaceutical prices and provide new methods to do so on a consistent basis, which led them to finance the MSH Price Guide in 2005[30] and to the WHO/HAI methodology and set of survey tools in 2008[31], geared mostly towards low-and-moderate-income countries as prices and availability infos are not centralized there[32]. For high income countries, the WHO rather pushes for them to create their own centralized systems, such as Euripid in Europe[33], which contains all drug prices of EU members but is inaccessible to the public unfortunately. Also of note, 2014 is the time when EU began to implement drug pricing policies systematically following this report: [34], and 2009 the first financial contribution of Euripid (following the first WHO/HAI report). So it may be that simply this phase of price monitoring by the WHO is ending, and is progressively shifting towards the countries themselves using these provided tools to make their own platforms, as is happening in EU. Signimu (talk) 13:55, 9 December 2019 (UTC)

Coming back to the fact that it's not updated since 2015, the WHO now lists the MSH update frequency as "regular" instead of "annually" before, and WHO/HAI database as "sporadically" updated. However, I don't think this is really an issue, sure it would be better to have more recent info, but it's not uncommon for economical metrics to not be updated annually, see for example the OECD pharmaceutical expenditures[35], the latest data for several countries dates back to 2015 and possibly before (didn't check all countries), and yet figures are made to present the expenditure "at 2018 or latest data", hence mixing up data from 2018 for some countries with earlier data for other countries. Signimu (talk) 18:32, 10 December 2019 (UTC)

• Colin, the historical data is interesting enough as an external link IMO. See http://mshpriceguide.org/en/single-drug-information/?DMFId=38&searchYear=2015 (I use this for palliative treatment of C7 radiculopathy). The sharp trending increase is interesting to note. Guy (help!) 08:53, 9 December 2019 (UTC)

  Guy this is exactly why we don't allow original research and I wonder if MSH is really suitable for "the general reader". Looking at the chart on the link you give and the price sharply increases between 2014 and 2015. See 2014 data and 2015 data. Firstly both records only have "Buyer" prices. See MSH Price Sources for an explanation of Buyer vs Supplier -- Buyer prices are less interesting/useful. In 2014 we only have the price the South African Department of Health agreed to pay, which was $0.0226 per 10mg tablet when bought as a bottle of 100. In 2015, the same South African Department of Health agreed to pay $0.0197 per 10mg tablet in a bottle of 100. The price went down 13%. However the 2015 database also has a record for the price the Sudan National Medical Supplies Fund will pay. Strangely the record is for a single tablet, not a bottle of 100 and the price is $0.1330 per 10mg tablet. The median price (which MSH recommend we use, and they use for the chart) breaks down if your dataset is as tiny as this. It ends up being the mean of the two prices. So the "sharp trending increase [that] is interesting to note" is fake, an anomaly of the dataset and artefact of limited sampling. And in my experience a 1 tablet price record is kinda weird for a medicine taken daily for continuous therapy. So I wonder if there's actually a mistake with the data.

  That $0.13 a tablet price for 10mg is far higher than any price for 50mg and 25mg. In fact the 25mg record is far far healthier from a statistical POV because it has loads of supplier prices (and a handful of buyer prices). The median supplier price for 25mg is 0.0084 which is 9x less than the median buyer price for 10mg in 2015. Having records with limited data is quite typical of the MSH. If one cares about statistics, only the 25mg record is respectable. The 10mg record is clearly prone to sampling error due to limited data (and it has no supplier prices at all, which suggests it really isn't generally available). I hope you can see that doing original research on the MSH dataset is exquisitely sensitive to whatever arbitrary tablet size you use, and extremely prone to misuse (such as using both buyer and suppliers and presenting highs and lows which can be outliers).

  But Amitriptyline is interesting for a second reason. My UK GP explained to me that this is an old tricyclic antidepressant that they no longer use much for that, because although effective it is really nasty in overdose, and sedating. Instead it has found a new use, in low dose, for neuropathic pain, and there are newer safer drugs for depression preferred in the UK/US. The Defined Daily Dose on the MSH site is 75mg which corresponds approximately to the maintenance dose for depression listed in Drugs.com. In the article we give a cost [with many mistakes] "per dose", but we don't tell the reader "for what?". You and I might be able to reverse-engineer the price by searching Drugs.com/BNF for doses for various indications and working out that this dose is for depression, and not for neuropathic pain. I have not found any source that states which indication a DDD was calculated for. So I'm not sure there is any general way of using MSH as a source and giving a price "per dose" with the reader informed about what indication (illness) that dose is actually for. The "most common indication" may well vary from country to country. In the UK, amitriptyline is licensed for neuropathic pain. In the US, this is off-label. In the developing world, its very cheap price might encourage greater use for depression vs newer drugs than is the case in the UK. -- Colin°^Talk 10:03, 9 December 2019 (UTC)

  Colin, all excellent points. But probably for the RfC. Why don't you and Doc James come up with a list of arguments for and against the three elements of Q1? This would certainly be an argument against any inclusion absent significant coverage in multiple RS. Since a Pro/Con list would not need you to agree on the underlying question, I would imagine you could do that without too much friction. Guy (help!) 10:22, 10 December 2019 (UTC)

  Guy, to be honest, even when I looked at Lenalidomide, which already meets NOTPRICES in having significant interest in its high price, the article conducts original research and gets the NHS wholesale price totally wrong and then juxtaposes that with a US retail price, which Sandy explained, is not actually price most people pay. Asking "do you want pricing" rather assumes it can be done, can be offered simply, and the choice is merely a matter of preference. James below seems to be willing to change his approach to how prices are presented. I suggest we tackle that, and then we can have examples to go with the questions. -- Colin°^Talk 10:54, 10 December 2019 (UTC)

• Threshold questions? Q1 as written above appears to be clearly answered by NOTPRICES already, so I'd suggest two substitutes for Q1:
  • Q1.1: "Does NOTPRICES require one source or more than one source?"
  • Q1.2: "Is the WHO source a sufficient source for NOTPRICES?".

  If the answer to Q1.1 was "more than one" and the answer to Q1.2 was "no", would that kill this entire dispute, rendering the other questions moot? Or would there still be stuff relating to prices in dispute? It seems like the answers to Q4–Q7 above are all going to be "follow the sources", and so the answers will entirely depend on what we deem are acceptable sources (Q1 and Q2), and Q3 (infoboxes, etc.) would have to wait until all the others were answered. – Levivich 20:12, 7 December 2019 (UTC)

I'm not sure I understand either of your questions. Number of sources isn't particularly relevant. What seems to be a problem is the incorrect interpretation of NOTPRICES that interest in drug prices in general is sufficient to warrant detailed drug price information in every article. It wouldn't matter if there are 2 or 100 articles on drug-pricing-in-general. Nor sure what WHO source you mean. I think we all agree that drug pricing is a matter of public interest, but that can be covered in, well, an article on drug pricing. -- Colin°^Talk 21:08, 7 December 2019 (UTC)

James wrote above Being mentioned in a single reliable source is sufficient IMO. and pointed to the WHO essential medicines list. – Levivich 21:48, 7 December 2019 (UTC)

Yes, but we shouldn't necessarily frame questions round one idiosyncratic interpretation of policy, or read too much into terse replies. WP:WEIGHT of course requires editors to be familiar with the body of literature on the subject. This is something that requires one to study a subject, rather than flit from one article to another adding factoids. But WP:WEIGHT doesn't require editors to cite all that literature, indeed when editors do find themselves adding multiple sources it is usually a bad sign of contentious editing. To use the US insulin price increase example, this is well known to anyone who follows medical issues, esp. in the US, and only needs one to cite one great secondary source. However, if someone questioned whether this was indeed a story worthy of XX amount of article prose, one could, on the talk page, justify that with more references. And if none were found, then perhaps indeed, it was just one journalist and not notable enough. Using a primary source for insulin, as with all sources, depends on what you do with it. In a simple case, then giving a price "per 1,000 iu of NPH insulin" might be straightforward from primary sources. But it isn't simple, because the source has 10iu, 40iu and 100iu vials in different records, and buyer and supplier prices, and multiple prices that must somehow be extracted to something simpler. That requires a degreee of medical, statistical and economic literacy that is clearly lacking, and why we forbid original research. The more important point appears to be the incorrect thinking NOTPRICES is simply sastisfied by a general interest in drug prices, and never once citing a single source discussing the price, because database records cannot discuss anything. -- Colin°^Talk 10:50, 8 December 2019 (UTC)

• Type of page. I just came here from the ANI discussion, and I've read through the discussions so far. One additional issue that occurs to me is the relative desirability of including prices on individual drug pages versus on pages about drug classes. For example, it seems likely that there are some specific drug products, such as some of the high-priced specific-use biopharmaceuticals, where there has been a lot of independent coverage in terms of pricing. On the other hand, discussion of comparative pricing differences on drug class pages (example: H1 antagonist) might, perhaps, be more encyclopedic that listing individual prices on every drug page (examples: Diphenhydramine or Loratadine). I'd be interested in finding out what the community thinks about that. --Tryptofish (talk) 20:32, 7 December 2019 (UTC)
  • I don't think there has ever been any question that inclusion of issues that have been discussed widely and independently in sources can be encyclopaedic, whether at the drug level or drug class level. The contentious issue (and the discussion continues at WT:MED not just here) has been the egregious original research from raw database records in order to supply drug pricing in every single article, irresepective of whether our sources are actually finding the price of that drug a matter of interest or not. -- Colin°^Talk 21:08, 7 December 2019 (UTC)
    • I was suggesting something that could be included in the RfC, as opposed to taking a position on what the conclusions should be. --Tryptofish (talk) 21:15, 7 December 2019 (UTC)
  • I think it's an interesting idea to explore – maybe it would indeed make our life easier to list prices in articles on drug classes if feasible. But this still won't get us past the issue that in real world, prices are set for commercial products whereas Wikipedia articles are predominantly about chemical molecules (that's also why we use INN as article title). I can't fathom Wikipedia listing all doses and formulations (tablet, ointment, inhaler, IV, etc.) and corresponsing prices across the world. That's not only impossible to achieve but also utterly pointless. Oh, and what about drugs that contain more than one active ingredient, like for example Fourderm [36]? Where do we stick in its price? — kashmīrī ^TALK 11:01, 8 December 2019 (UTC)

• I agree with you entirely. I think there needs to be some decision-making about the most plausible options for those things, and then have the community decide among those in the RfC. And without that, it's an exercise in futility. --Tryptofish (talk) 20:22, 8 December 2019 (UTC)

• Agree it is an interesting idea to explore, and one of the reason wiki prefers discussion to polling -- polling should be to confirm the consensus you feel is already established and based on clear examples. If people want "prices on drug class" or perhaps "prices on disease treatment section", then before we ask, I think there should be some valid examples to look at. For example, Wikipedia policy requires that if folk are comparing X's then we need sources that are comparing X's -- it isn't acceptable for us to do original resarch from raw data and present prices for readers to compare or even worse, for us to declare which is cheapest or best value [that's would require we take efficacy, tolerance, etc into account too]. Take the DDD, currently being used for original research dose prices. WHO explains how to use DDD and really it is not intended for this purpose and they explicitly state it should not be used to compare drugs within a class. This is why such comparisons should be left to experts and we cite the experts, not database records. So, if we have good secondary sources for drug-class price comparison, and someone can make a few typical-case examples, then it is a good question to ask. If we can't make valid examples, we shouldn't ask. -- Colin°^Talk 12:11, 8 December 2019 (UTC)

  The problem with that approach, which usually is based on comparing drugs in a similar ATC class and is usually called internal reference pricing, is that it can lead to some biases whether we use per dose or per gram price, and is always unfit for international comparisons (only unfit when using DDD it seems) [37]. I'm not saying we should not explore such approaches, it could be very interesting, but we should first ensure we use the proper metrics for the proper purpose, I plan to post what I find after the discussion on prices encyclopedic pertinence. Signimu (talk) 22:07, 9 December 2019 (UTC)

• For Q3 "Where pricing is included..." I'd like a question option something like 'Should the price in the lead be detailed specific dollars and cents for a variety of countries/areas or be simplified to e.g. "cheap", "inexpensive", "very expensive", "affordable"'. IMO this is more in keeping with WP:LEAD and is what a professional article on a drug might say, and I suspect we are only including the detailed figures because we only have primary data records for sources and and to simplify translation efforts that are of no concern to this encyclopeadia. I suspect the community, if asked, would much prefer or even insist on the simplified lead text. Can this actually be achieved in the general case? -- Colin°^Talk 12:44, 8 December 2019 (UTC)

  Colin, that would be good, but where do we source it? Guy (help!) 08:54, 9 December 2019 (UTC)

  Guy I suspect a simplified statement like "wonderpam is a cheap antidepressant ..." is only valid for one country. So a US or UK publication could well say this. Whether we can say that is harder to say, particularly as our concept of "cheap" is distorted by people in the world earning less than a dollar a day. I think there are likely tertiary country-specific sources that say this sort of "knowledge", but may be hard to find a source that does this routinely. Our biggest problem with prices IMO is not just what we would love to say to our readers, but the extremely limited things we can say to our readers based on very primitive raw datasets. Already all the examples in articles require horrendous amount of original research. That's why I throw down the gauntlet on those who want to push for a "prices in all our drug articles" to be up front about what kind of prices, use a meaningful and consistent definition, and demonstrate you can do this routinely while obeying policy. As with WP:V, it is up to those wanting to insert text to justify their inclusion. -- Colin°^Talk 10:21, 9 December 2019 (UTC)

  Such a concise qualificative summary would be nice, but I don't think possible. This looks like trying to qualify 'accessibility', which is not a directly observable factor, as it is influenced by prices, volume (availability) and socio economic factors (such as wage and general 'richness' of a country and its citizens), and probably others. The closest thing that I've seen is the 'affordability' metric by the WHO/HAI[38], which is the number of wage days the lowest paid unskilled government worker has to pay to "purchase 7 days’ supply of a medicine to treat an acute condition, and 30 days for a chronic condition, based on standard treatment regimens". Eg, for Metformin 500 mg cap/tab (sorry they don't provide direct links to specific entries), the affordability ranges from 0.2 day of wage to 14.6 days. This can be used for comparisons, either between generics and originators, or between countries (since affordability is always based on a local wage, no adjustment is necessary). However, I did not find any source that defined what is unaffordable, although the WHO/HAI manual says that 2 days is already considered unaffordable "for many low-income patients" [39]. Signimu (talk) 21:57, 10 December 2019 (UTC)

  Signimu, Tryptofish See post 22:07, 9 December 2019 by James: Example text. The link opens at "valproate", which is a drug we discussed. This is a UK book and the book says "Valproate is inexpensive (around £9 for a 100-pack of 500mg tablets). However, cost increases with the complexity of the formulation. So the book qualifies "inexpensive" with a sample price and gives a caveat. In a global encyclopaedia, we can't just say "Valproate is an inexpensive anticonvulsant" and cite a UK-centric book that has caveats. If you can't read the whole book, google the title and you may find a handy PDF of the 2015 edition, which I'm not going to link. Looking at the other "cost" examples in the book, it reads very much like a doctor->doctor advice on saving the NHS a few bob and, in the 2015 book, I felt the commentary on the cost at "warfarin" vs newer drugs was outdated (the authors are pharmacology registrars, not cardiologists or neurologists) This, cough, is one reason why I keep banging on about there being a problem with MSH database ending in 2015. Basically, if you look at the book (which only covers 100 most subscribed drugs in the UK), there are only a small number of cases where the "cost" is described neatly in one word. It is never as simple as claimed.... -- Colin°^Talk 22:36, 10 December 2019 (UTC)

  If we explicitly mention this qualification is for UK only, then I think why not, but if not the problem I would have with that kind of qualification is that it is 1) not systematic since there is no explicit definition, 2) not generalizable to other drugs or countries (because of lack of explicit definition). The WHO/HAI database is certainly not perfect as you say, it's not complete and not updated as frequently as we would like, but at least they explicit (and validated) a standard method for each metric they provide. I'm not saying we should use it, I'm just describing possibilities Signimu (talk) 23:58, 10 December 2019 (UTC)

• What kind of price stage and measure There are multiple stages of prices in the supply chain that can be used, and not all have the same availability, see external reference pricing which is the first article of a serie we can make about pharmaceutical price monitoring and policies. Then there are also multiple price measures, as pointed out elsewhere by Collin. There are unfortunately statistical intricacies, such as the cost per DDD being suitable for estimating consumption but NOT for estimating price nor doing cost comparisons, except maybe for internal reference pricing, see this excellent literature review[40]. And not all of these price measures, and stages in the supply chain, are available publicly. If we are to accept prices, we need to have a consensus on what stage and what measure we should use. We can choose multiple stages, but then it should be clarified in text. I think we still should research sources and discuss further about what measures of price we should use (the ERP article is a first attempt at that), so I think doing a RfC at this point would be premature. --Signimu (talk) 13:02, 9 December 2019 (UTC)

• Current summary needed It feels like it might be helpful for someone to summarize current thinking on what the RfC questions are. If need be I will do this but I am hopeful given the talent of the editors involved that this won't be needed and we can see if there is consensus to launch the RfC. Best, Barkeep49 (talk) 02:53, 10 December 2019 (UTC)
  • Barkeep49, several editors have suggested an RFC is premature if the encyclopaedic text offered by the question cannot actually be achieved in the general case without original research and essentially arbitrary choices. The price for a drug can change by a factor of 15x or more depending on which database record is cited and what method used. There is even the more basic question that we are claiming a drug has a price (in US, developing nations, UK, etc) and, other than exceptional cases, it clearly very much doesn't. So really, what is the point of asking the community "should pricing be included ..." when the examples already discussed fail fundamental non-contentious policies. it really doesn't seem unreasonable to ask those wanting prices to demonstrate it can be done, and then we can have the community-consensus-opinion of whether we want them. If necessary, I can create a section that lists the sources currently being used, and how they are being misused in articles. Perhaps then the community can come to some consensus on a legal valid source/wording of prices that would be permissible if we are convinced they should be included. -- Colin°^Talk 10:16, 10 December 2019 (UTC)

    Colin, I understand that is the contention of several editors. The idea that prices are original research has been a contention of those opposed to to pricing information all along. If editors who have been in favor of including pricing information and editors who have been opposed feel like headway is being made towards consensus, well great. Don't let me stand in the way. However, that's not quite how I'm reading recent discussions, and I don't think it's how Guy, who is another uninvovled sysop trying to help this along, is reading it either. I really do mean it that if editors on both sides feel like progress is being made and not just old arguments rehashed, then great. But absent that - and so far I admit I don't see this feeling from both sides - I think a focus on coming to agreement on how to structure what will be a fairly complicated RfC could be the best use of editor efforts and energy. Best, Barkeep49 (talk) 14:48, 10 December 2019 (UTC)

    Although I do not agree with Colin's behavior focusing on specific editors, I do agree that the RfC is premature as it would be too vague if we don't discuss/prepare first some concrete options of what kind of prices/pricing can be used, in other words to discuss a bit more about the methodology. I think we made a tiny bit of progress with Ronz's price/pricing concise distinction below, with pricing seemingly being more acceptable. Now more specifically I think we should discuss what kind of pricing is acceptable, I'm expecting everyone will have a different opinion ofc, but I plan to provide some concrete examples I hope will further our discussion and give us material to properly formulate an adequate RfC. Signimu (talk) 14:59, 10 December 2019 (UTC)

• We would need to define "excluding trade press". I imagine you would be referring to medical sources here, like textbooks, review articles, and government websites? Doc James (talk · contribs · email) 19:21, 10 December 2019 (UTC)
  • Guy is your Q1.1 a reformulation of the existing WP:NOTPRICES? If not, what are you trying to achieve with any difference? Immediately above is a query on "excluding trade press". I should point out that at 18:37, 6 December User:Levivich quoted WP:NOTPRICES, to which the response was "And I agree with that. All the sources being used to discuss prices are independent. And we have sources that justify the importance of pricing information". Multiple editors disagree that WP:NOTPRICES permits the insertion of prices across all drug articles, though their opinion on the matter is dismissed. Your Q1.2 and Q1.3 rather assume your Q1.1 restricts the insertion of prices to a smaller or distinct set of articles. Yet Q1.1 does not specify whether the "cost" is the cost of that drug or drugs in general or that class or grouping of drugs. This matters because "a justified reason for the mention" has been argued to be satisfied merely by global interest in drug pricing. My point is that WP:NOTPRICES has been ineffective at preventing the addition of prices to all drug articles, and that if one's attempt is merely to restate it, then frankly what is the point? If you are trying to improve or clarify WP:NOTPRICES, then I think you really need to give some examples of what you think it would include/exclude, and get agreement from those have so far dismissed that WP:NOTPRICES is a restrictive policy wrt drug pricing, to agree that it will make a significant difference and agree with your examples of in/out articles. Without that agreement, we are simply wasting all our time. -- Colin°^Talk 21:57, 10 December 2019 (UTC)

    Colin, I am trying to forestall attempts to override consensus by claiming that articles in Chemist and Druggist discussing the price of Pharmaco's new generic wonderpam don't get used to try to crowbar pricing in when it's not demonstrably a matter of significant public discussion. Guy (help!) 22:36, 10 December 2019 (UTC)

    • I have not seen trade press (by which I presume you mean pharmacy magazines) used to justify costs in drug articles. I have however seen the general interest in drug prices by WHO. I have seen the suppression of drug prices in US commercials. I have seen the claim big pharma want to suppress and censor drug prices. All these are used as reasons to justify NOTPRICES is satisfied. I don't think merely excluding "Chemist and Druggist " is going to work. -- Colin°^Talk 22:43, 10 December 2019 (UTC)

Sample article please ?

This discussion could go nowhere fast if we don't have concrete examples to focus on. We will get dizzy from different dosages, generic and brand names, prices in different currencies and countries, and so on. And for example, in Trypto's post just above this, I did not realize he was proposing a question rather than asking for discussion, so the format here could get muddled.

If we had a mockup to work from, editor commentary here could be more focused and specific. I have suggested that an article like Epipen could be worked up to include pricing information that is fully compliant with WP:OR, WP:WEIGHT, WP:LEAD, WP:NOT, and WP:RS. Why do we not work that article up as one example, and then pick a drug from those that are currently contested (that is, based only on a database source), and work up the best possible mockup of what is proposed? If we have two samples, we could then analyze, side-by-side, what each of us suggests as questions based on those mockups, and it would be much more clear to the wider community when launched.

The idea of coming up with a question list is revealing that we are essentially planning to ask the community whether pricing data should be exempt from policy (NOT, WEIGHT, OR) and guideline (LEAD, MEDRS). Should we just ask that question outright? Putting forward an RFC to ask the community to basically exempt pricing from policy (or change policy to accommodate pricing) is going to be complex: we could at least present a well-thought out sample, showing our best effort. My impression from the text I've read so far is that we don't have a good sample that doesn't involve OR. Could we see the best shot at what is proposed? I believe that will make discussion much easier ... particularly when we take this to the wider community. SandyGeorgia (Talk) 03:30, 8 December 2019 (UTC)

Sorry that it was confusing. I was going by what Guy posted at the top of that discussion section. --Tryptofish (talk) 20:23, 8 December 2019 (UTC)

Tryptofish, the problem I am seeing is that it is unproductive for us to be suggesting questions until/unless we thoroughly discuss each issue and have sample sourced text to reference-- that is, I am concerned that Guy's format is not going to work here. Your suggestion provides a great example: we should explore this issue before proposing it as a question to the wider community. Again, I think we need a working sample. Colin has done utterly boatloads of work to uncover all of the problems here, and yet we still do not have answers to those issues, or a GOOD sample of a best practice proposal from which to work that uses sources correctly. JzG, I fear we will need to reboot the reboot, and set up individual sections to discuss each question, but do that only after those wanting to add prices can show us a sample where they overcome the voluminous issues Colin has uncovered. As of now, I fear we are not working towards something that can be presented to the community with a straight face. We are asking to change policy, but we haven't produced one good example to support that, and our sources do not support the text added in apparently any case (at least, one has not been offered as a sample).

I am also very concerned about the number of RFCs that fail because they ask too many questions. At some point, we have to gel this down. I suggest a new format is needed for this page, the facilitates discussion of each issue, and only when it can be based on side-by-side examples showing the goals and the issues. How, with a straight face, are we going to the community to ask for a policy change or interpretation without even presenting one article where cost is well added ? SandyGeorgia (Talk) 17:05, 9 December 2019 (UTC)

I think those are all good points. As a newcomer here, I was just trying to follow what Guy said, but I also did not suggest actual wording for an RfC question, because I'm unsure about those things too. I agree with you that we need to better understand what it is that we should ask, before we will be able to ask it. I also agree with Guy that, as we begin to workshop possible questions, we need to focus on the questions themselves, rather than on editors' opinions about what would be the right answers. We might, perhaps, want to consider a simpler RfC to come first, evaluating community sentiment on whether we should include pricing at all – sort of like just Q1 and Q2 above at #Opening proposal, probably with further revisions. (Although I'm sensitive to the issue that editors won't be able to answer broad questions about pricing until they understand what is being proposed about how the pricing would look on the page, I think that those two basic questions can be answered from the start.) Then, if the community puts boundaries on what is or is not acceptable, we could work on a follow-up RfC that explores the more detailed options. --Tryptofish (talk) 20:45, 9 December 2019 (UTC)

I would be strongly opposed to running "a simpler RfC to come first, evaluating community sentiment on whether we should include pricing at all" because essentially we had that in 2016 and it contained the same emotive pleas that "pricing is important so our drug articles should contain prices" and the same unfounded claims "we have great sources". Voters were asked to give an opinion on prices without a clear idea of what price. I mean, a lot of us might support "XXX is a cheap anticonvulsant" or "YYY is an expensive cancer drug" [assuming sources for that, and worldwide applicability of that claim] but have a big problem with the dollars and cents approach adopted so far. There seems to be an emerging consensus on this page that the pricing => prices logic is not justified and that we should encourage the improvement and expansion of articles on the issues of drug pricing. I hope I have managed to convince folk here that the sources are unable to supply the sort of prices some want to offer our readers, and that the concept of "a" cost or price for "a drug" is, in the general case, a mirage. So let's not repeat past mistakes. Our policy and guidelines are best built upon a foundation of best practice. Where is the "best practice" for drug prices in articles? I see plenty "awful practice; don't do this folks" examples. Learn from the mistake of Brexit: there was a slim majority for "Yes" but then it turned out there was no consensus of what kind of "Yes" folk actually wanted or could be achieved in practice.

Since the suggestion of an RFC was raised, I have only seen further tangential pleas about the general importance of "drug pricing", and no effort whatsoever to solve the problems I've uncovered already with "drug prices". I suspect strongly it is because the whole thing is busted and they know it. -- Colin°^Talk 22:08, 9 December 2019 (UTC)

I agree we should not ask questions that appear to offer a get-out-of-jail-free card to exempt editors from established non-contentious policy. ALL the current examples looked at fail abysmally in terms of WP:OR, WP:V, WP:WEIGHT, WP:LEAD. There was an attempt at WT:MED to get editors to give examples for valproate but framed as "How should we summarize this source" and cited a database record of one particular pill size that conveniently happened to have only one supplier (a reasonably international one at that). Instead how should the cost of valproate be described to our readers, and how would you achieve while obeying policy? The problem with Epipen is that we can probably find lots of discussion on the price and secondary sources for the price and it already meeds WP:NOPRICE in terms of the notability of its price (in the US anyway). The article economic details are also a bit of a mess and I think that will only complicate any discussion about it. The contentious issue is whether we can do this for all/most drugs. The examples looked at so far are:

• Ethosuximide -- the leading treatment for childhood absence epilepsy.
• Diazepam (aka Valium) -- leading benzodiazepine for anxiety, alcohol withdrawal syndrome, muscle spasms, seizures, trouble sleeping, and restless legs syndrome.
• Diclofenac (aka Voltaren, etc) -- NSAID for pain and inflammatory disease. Also has controversial veterinary use.
• Carbamazepine (aka Tegretol) -- anticonvulsant and neuropathic pain.
• Valproate (aka Depakote, Epilim) -- for epilepsy and bipolar disorder and to prevent migraine headaches. One of the most broadly effective anticonvulsants.
• Mebendazole -- parasitic worm infestations. This one does have potential for notability on its US price and only its US price. This is simpler than epipen. If, for example, we wish to compare US vs UK prices, how would we do it under policy rules.

Perhaps other editors have other examples they'd like to see. -- Colin°^Talk 12:38, 8 December 2019 (UTC)

Another example is Amphotericin B. Signimu (talk) 23:41, 8 December 2019 (UTC)

I will try to express myself more clearly. We cannot expect and should not ask RFC or talk page participants to scroll through long discussions and sort out a multitude of issues. We need to present concrete examples demonstrating the issues and summarizing them in a digestible format so that a new reader to this page can easily view, review and check vs. sources.

When we discuss OR, WEIGHT, etc, we have to discuss them relative to sources, what the text says, what the source says, etc, in a way the new participant can easily engage. Colin has thoroughly analyzed a number of articles and found numerous problems that are already detailed on this page: one of those as a sample needs to be summarized in a visually digestible format such as a table. We have already established there are many: we just need one that is typical and can be summarized.

Colin, could you pick an example and do something like what I did in sandbox so that any reader coming to this page can clearly see the issues? Could Doc James or QuackGuru pick a typical drug that you believe is not affected by the issues Colin mentions and do the same? Once you settle on a table or format, then we can give maybe six examples or something like that. Then, for example, Trypto could give an example of what typical text for a drug class would look like in the same visual format. And we could see a sample of what a policy-compliant drug cost (eg insulin or epipen, for which the sources do the work for us) would look like in the same format. SandyGeorgia (Talk) 04:56, 9 December 2019 (UTC)

Moving my sandbox content to here, so it can be deleted from my userspace. SandyGeorgia (Talk) 20:02, 22 December 2019 (UTC)

Extended content

Valproate [41]—(Depakote, Epilim)—for epilepsy and bipolar disorder and to prevent migraine headaches. One of the most broadly effective anticonvulsants. Article text Source Source text Notes The wholesale cost in the developing world is about US$0.40 per defined daily dose as of 2015. "Sodium Valproate | International Medical Products Price Guide". Put here exactly what the source says Put explanation here of concerns about OR, WEIGHT, etc. and limitations such as those discussed on talk (as in, only one data point supplied for one country, and so on. In the United States, the National Average Drug Acquisition Cost is roughly US$1.30 per day for the short acting formulation as of 2019. "NADAC as of 2019-11-27 | Data.Medicaid.gov". Centers for Medicare and Medicaid Services. Put here limitations related to US insurance, etc. The price to a consumer in the United States for this amount is about $US2.30 per day as of 2019. "Valproic acid Prices, Coupons & Patient Assistance Programs". Drugs.com.</ref> The long acting formulations are more expensive. "NADAC as of 2019-11-27 | Data.Medicaid.gov". Centers for Medicare and Medicaid Services. In the European Union, end-user costs are less than 0.60 EUR for an average daily dose in Germany. Regular pharmacy price, including all taxes, et cetera: less than 34,43 EUR for 200 controlled release pills with 500mg each; date: 2016-11-30 No source is given.

Sandy, wrt your sandbox, I think you underestimate how much can be wrong with so few words and numbers in article text. I've taken apart several examples and examined several sources and explained their limitations, and I'm not sure there is a way to do that in just a few words. At this point I feel strongly that every single example of drug pricing we have examined breaks so many rules, whether wiki policy or elementary statistics, that I think the ball is in the court of those who want to include prices. Over to you guys. Mock up these example of how you think it should and can be done, and let the community judge. If you can do this, and satisfy the community, then we can vote on whether we want it or not. I have serious doubts that MSH, Drugs.com, BNF, US NADAC databases of raw price data can be used to present prices in a way that has currently been done, or that people want, without breaking WP:NOR and without being plain incorrect to our readers about our claims. If people really do want a single place to see the criticism of a source and how it is used in an article, then I could do that in a sub-page, say. But if e.g. we start to agree that MSH is a non-starter, then I don't want to waste my and your time discussing how badly it was used. If we still don't agree that MSH is causing OR on a massive scale, then sure, I could write up something. -- Colin°^Talk 13:43, 9 December 2019 (UTC)

One of the questions that could be usefully asked is what an encyclopedic presentation of pricing/prices/costs would look like. Imagine that the sourcing question is solved (which it isn't, but that's a separable question). Which of these sounds most like a Wikipedia article?

• Compared to other treatments for the same condition, Wonderpam is inexpensive.
• Wonderpam's high price makes it unaffordable to about half the people in the world.
• The worldwide median National Average Drug Acquisition Cost (or some other specifically named price) of Wonderpam was approximately US$5 for a 30-day supply in 2015.
• Wonderpam's manufacturer shipped 30 million pills and reported a total net revenue of US$5 billion in 2015.

(Yes, these statements are contradictory. They're meant as examples of writing style, not as examples of facts.)

Assuming that you had the perfect sources in hand, and they were saying exactly this (only in slightly different words, because copyvio), and that nobody disputed the idea of including something about the financial aspects of Wonderpam in a Wikipedia article, which of these is the kind of thing that you would want to write?

I think that editors who aren't willing to go through the complicated stuff about how the pharma world works could still share a valid opinion about whether the best practice (assuming ideal sources, etc.) focuses on relative costs (it's inexpensive), access (it's unaffordable), wholesale costs ($5), manufacturer's revenue, or something else. WhatamIdoing (talk) 23:43, 14 December 2019 (UTC)

WhatamIdoing, I don't think your first, second or third sample sentences are problematic wrt should-we-shouldn't-we, although none of us can override WP:WEIGHT (though there has been an attempt to wrt global-interest in pricing => prices in each article). If our sources think the relative cheapness of wonderpam compared to its peers is relevant, of course we can include that. But that doesn't justify for example, us additionally doing OR and inserting random numbers into our articles based on our own made-up-and-usually-wrong treatment costs. The first three samples all make global statements, which can be problematic. While a relative-cost statement might be ok, I think any absolute claim that X is inexpensive is likely to be country-specific. There is plenty evidence from WHO that drugs we might consider inexpensive simply aren't available though national health services in the developing world, leading to extremely high private retail price. This makes a nonsense of the claim that a portion of the world are paying out-of-pocket => we should include wholesale list prices in our articles. I don't think we can have all discussion purely on an "assuming perfect sources, saying exactly this" basis because we do live in the real world. The fourth bullet point is likely to come from a press release so we must be cautious not to repeat self-reported claims. I assume figures that come from a company annual report to shareholders have a legal obligation to be correct, so that might be better. I haven't really looked to see if that information is generally available. It probably isn't for drugs that don't affect company bottom line, and likely to be a commercial secret. The third point... well that's the problem because we simply don't have sources doing that because it doesn't make sense to do it (in the general case) without being much more specific. I think there has been a naivety / deliberate over-simplification to suggest that in the general case drugs have prices. They really don't. -- Colin°^Talk 13:05, 15 December 2019 (UTC)

Disputed

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

---

What is currently in WP:MEDPRICE[42] is disputed. The current text was edit warred in, and does not reflect consensus. I don't edit war, and don't want to revert to older consensus version, but we should not leave the impression that this text has consensus. It seems the best interim solution, as we wait to formulate an RFC, is to add a disputed tag to that section. Do we now have neutral admins following this page, and is it acceptable for me to tag that section? We cannot leave the impression that this is consensus text during RFC formulation and execution. SandyGeorgia (Talk) 19:44, 8 December 2019 (UTC)

I think the best option is to blank it. It was added in this edit in October by QuackGuru, a diff that I very much encourage others to read as it explains the agenda being pushed. It was blank before the article edit-warring that sparked this renewed debate began, and in the interests of neutrality, should be blank again. -- Colin°^Talk 21:58, 8 December 2019 (UTC)

The text was: When including content about product pricing an independent source should be used. The pharmaceutical industry has tried to conceal medication prices because of their continuing legal cases in the United States. There is a lack of transparency regarding medication and vaccine prices among non-governmental organizations. SandyGeorgia (Talk) 02:24, 9 December 2019 (UTC)

So there's no discussion, and a clear BATTLE behind it? Unless there's some non-disputed version along the way (I'm not seeing one), blank it. --Ronz (talk) 22:08, 8 December 2019 (UTC)

Yep, there has been no discussion beforehand. QuackGuru amended the manual first and only then posted on Talk. That's already bad. Another issue is the text itself: suggesting that patients' "lost wages" are a component of drug price (sic!). Man, this guy should start reading before writing for an encyclopaedia. — kashmīrī ^TALK 22:45, 8 December 2019 (UTC)

There was a discussion back in October. Changes made without a RfC may be against the closure at AN/I. There was a consensus among editors back in October for a rewrite. QuackGuru (talk) 22:49, 8 December 2019 (UTC)

It wasn't WP:CONSENSUS as you can see - the discussion is right above and there were several voices who did not agree with it. That's why we are having this RfC by the way - because there was no consensus. The closure at ANI prohibited adding drug prices to articles. — kashmīrī ^TALK 22:54, 8 December 2019 (UTC)

The main discussion, which this is a subsection, was started about adding pricing information to this guideline. (Wow, we've been here a while.) The RfC we're now forming is to determine what will be included. Starting with a clearly disputed section is probably a very bad idea. --Ronz (talk) 23:17, 8 December 2019 (UTC)

Noting that I have fully protected this MOS page pending the completion of the RfC. All are also reminded that the community, including I believe nearly every editor in this subsection, expressed broad consensus for The above debates will be subject to civility restrictions with strict enforcement of WP:AGF, WP:CIV, no WP:BLUDGEONing and no rehashing of grievances. (emphasis added). Best, Barkeep49 (talk) 23:32, 8 December 2019 (UTC)

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Proposal for resolution

These discussions have been going on for several months, and although we are now actually discussing and trying to develop consensus, the issues revealed are increasingly complex, and we still don't have a best practice working example. So, I want to float an alternate proposal in the hopes we can find a less burdersome way to resolve the entire matter.

In some cases (samples Epipen and insulin), we have no problem discussing costs and pricing without breaching policy and guideline, as the MEDRS-compliant sources do the work for us.

In other cases, we are relying on database-type information, and having to do the math which involves a multitude of problematic factors, discussed at length above and which will be quite cumbersome for the wider community to digest.

What if we were to simply provide an External link template for each of the data sources that have been employed so far, and agree to add them to External links on drug articles, and let the reader do their own original research? Then we would also need to agree that only compliant text would be added to the article, that is, subject to WP:V, WP:MEDRS, WP:WEIGHT, WP:LEAD, WP:NOT, etc.

And if we were to come to consensus to do this instead, would that satisfy the ANI closers although avoiding an RFC? We are short on medical editors, and all of us could be more "gainfully employed" here improving content, so my hope is that we can find an alternate solution that gives our readers more access to price information, while not breaching Wikipedia policy. SandyGeorgia (Talk) 17:40, 9 December 2019 (UTC)

SandyGeorgia, that's one of my options above. I am OK with proposing just this (include where primary subject of multiple sources specifically about that drug; external link for WHO essentials) rather than multiple options. Guy (help!) 18:37, 9 December 2019 (UTC)

For the databases, the HAI did the work for us: multi-countries databases list[43] (+ the HAI one: [44]) and national databases list[45]. But I don't think this will fix the issues. There are real methodological issues, even if we just take the prices straight from the databases, they are not necessarily comparable because they may originate from different stages in the supply chain (see external reference pricing where this is clarified). The HAI somewhat simplify this issue as they specify the supply chain stage for each price database. I think we should further study prices before launching a RfC, the creation of the external reference pricing article is a first attempt at doing that, but we can do more with the references I've lately found. Signimu (talk) 19:05, 9 December 2019 (UTC)

I think we need to have an RfC in any case (in other words, to not avoid having one). There really needs to be a community consensus about some of the most basic questions about whether or not to include pricing on all drug pages. --Tryptofish (talk) 20:50, 9 December 2019 (UTC)

With respect to prices for LMIC, I consider the sources we are currently using to comply with WP:V, WP:MEDRS (which they do not necessarily have to as this is not a health claim), WP:NOT, and WP:WEIGHT. With respect to having these details covered both in the body and in the lead that is easy to comply with, and I am happy to do so.

With respect to the wording around how to summarize this source I am more happy to discuss.[46]

Currently we have "The wholesale cost in the developing world is about US$0.40 per day as of 2015."

If people feel this is overly summarized we could go with

A external reference pricing in the developing world is about US$0.40 for a typical maintenance dose in adults as of 2015."

Would be interested in hearing other peoples suggestions.

We also have sources such as this one we list it as "inexpensive".[47] Would also be happy with simple that in the lead with the rest of the details going in the body.

Not supportive of simple adding an external link.

Agree we have a lot of other stuff that needs doing such as updating epidemiology in a few hundred articles in 2017 data.[48] Doc James (talk · contribs · email) 22:07, 9 December 2019 (UTC)

Tryptofish, yes, absolutely. That's the consensus from the ANI thread and is also the best way to definitively resolve the underlying dispute. Guy (help!) 10:18, 10 December 2019 (UTC)

Well this is the first time James has accepted that his approach to listing detailed prices in the lead should end. Yay for WP:LEAD compliance. He cites a UK textbook that interprets "inexpensive" in terms of NHS drug pricing and socialist healthcare, which, as we have seen many times, is not reflected in US let alone developing nation pricing -- we have quotes above from WHO that drugs with low wholesale prices may be unavailable through public health and only on private prescription charges, which can be many many multiples of the wholesale price and most certainly not "inexpensive". Unfortunately the various suggested article texts above are simply (a) not supported by the source and (b) still require original research. And currently I'm not seeing suggested texts by anyone other than one editor, yet several people have previously expressed support for prices. This should ring alarm bells. We only have text offered by one editor that has already multiple times shown to break even the most basic non-contentious policy. Wikipedia is a collaborative editing project, which means I expect to see a consensus of editors offering examples of their wonderful drug prices in articles they wrote. I don't. At the moment, a premature RFC would effectively giving carte blanche to one editor to do as they wish. That doesn't sound like how a collaborative editing project works to me. -- Colin°^Talk 11:57, 10 December 2019 (UTC)

"And currently I'm not seeing suggested texts by anyone other than one editor, yet several people have previously expressed support for prices." That's not entirely true, other editors (such as me) may be working on the issue in other ways, such as by reviewing the literature and expanding articles such as external reference pricing and median price ratio. For an example of a possible basis of prices, see my reply above [49], although please note that I plan to provide more content for other editors to evaluate as I wrote elsewhere. Signimu (talk) 13:52, 10 December 2019 (UTC)

Colin, As noted by Barkeep49, please restrict your comments to the specific issue of what questions should be asked at RfC and do not comment on other editors or you may be sanctioned. Guy (help!) 14:17, 10 December 2019 (UTC)

Guy, Barkeep49, Any questions in the RFC must contain examples of what the proposer means: article text + sources, or a worked example. And if you want to ask a question about "Authoritative primary sources" or "Secondary sources only" then you'll need to give examples of those because editors may disagree on what you mean. And of course we can't discuss sources without an example of what you propose is drawn from them. We know quite well in MEDRS and elsewhere that whether a source is permitted is not just the subject-domain but what exactly you are saying about it. So it isn't fair to ask us to say primary/secondary without an example of what you are using the primary source to say or what you are using the secondary source to say. Same for product cost and treatment cost. I don't really understand how, in good faith, an admin on Wikipedia can propose to ask the community if they want to see treatment costs in articles, say, without a working example of an article with treatment cost and an agreement that this was achieved without breaking other policy that it is not within your power or remit to change. I see it as a bit futile to run an RFC with a question + example that will instantly be rejected because of obvious shortcomings with the example. So I really really would like to see more than one editor participate in doing that. Perhaps you want to post a notice at WP:MED asking for further contributions to examples of great article price text. -- Colin°^Talk 16:37, 10 December 2019 (UTC)

I can give a few examples of primary/secondary sources issues:

• The most primary level of sources for prices would be databases, but there are some particularities. The MSH International Medicines Price Guide for instance allows to write things like 'the median supplier price is $XXX'[50]. The WHO/HAI database[51] builds on the MSH guide and provide higher-level price metrics, such as the median price ratio or affordability. Of note, some of these databases are accompanied with manuals that also provide some analyses (eg, the WHO/HAI methodology and database manual[52], see figure 1.1 and table 1.1).
• Papers analyzing the databases, such as those using the WHO/HAI methodology[53] or MSH price guide[54]. We then get higher level infos, ranging from "prices of originator medicines and generics are respectively 21.33 and 11.53 times the international reference price in Boston area"[55] to the availability and prices of cardiovascular medicines in 36 countries[56][57].
• Reviews such as [58][59][60][61], but at this level they usually only pertain to the effect of policies on prices, but sometimes we get some specific infos when it's used to illustrate a point such as "Public sector data from 13 LMICs showed that the mean public sector price of human insulin (100 IU isophane/regular 70/30) represented 0·7 to 6·2 days’ minimum public sector wages for 1 month of treatment" or "Private pharmacy prices for four commonly used cardiovascular medicines in 18 countries showed that they were potentially unaffordable for at least some patients in every country [...] 0·14% of households in HICs, 25% in upper-MICs, 33% in lower-MICs, as much as 60% in LICs (excluding India), and 59% in India"[62].

So can we consider papers studying databases as enough for citing prices? If not, we are left with review sources, which are mostly focused on a high, policy level view.

Worth mentioning is this excerpt from a chapter from a specialist on how to choose an appropriate source for medicines prices: "In general, literature does not qualify as appropriate data source for price surveys, analyses, and comparisons. As the price studies in Part 1 also indicated, primary price data were published in only few papers, and they were usually focused on a low number of countries and medicines. Despite its limitation as data provider, scientific literature can prove to be useful in offering additional (background) data for the analysis and interpretation of prices (e.g., a global price survey on sofosbuvir and ledipasvir/sofosbuvir was based on the collection of medicine price data from national sources, but a discount mentioned in literature was considered for the calculation of discounted prices)." In other words, if we use the databases directly, we have a more uniform and standardized data source, whereas with literature we get more interpretation but less data. That's a trade-off we see with pretty much any other metric in fact. --Signimu (talk) 18:15, 10 December 2019 (UTC)

Remember that advice for academic research and publishing does not necessarily align with Wikipedia sourcing best practice. Academics are taught to cite the original research paper whereas we are taught to cite the review. Of the sources used so far, MSH, BNF, Drugs.com all operate at the level of low-level data. Whether they are strictly "primary sources" is an academic debate (clearly they aren't the drug companies or suppliers own list prices, so their data is second hand). I'm only just now looking at HAI and so tried searching for some of the drugs that we've discussed. For example "valproic acid" and "amitriptyline". For the former I get 500mg "Latin America, lower-middle income December 2009" or for 200mg I get "Morocco April 2004". For the latter I get 25mg in loads of regions. I tried "Sudan March 2012" and got nothing for "Public sector procurement prices ", nothing for "Public sector patient prices " but data for "Private sector patient prices" And then lots more numbers and percentages I don't yet understand. I'm seeing two things about this (and the MSH it builds on) that (a) the data is often very very limited and (b) doesn't go past 2015 and (c) is still for a particular tablet at a particular mg. So while the HAI may have some additional metrics, it still doesn't solve the problem of original research to pick which pill-size. Neither let us write "the median supplier price is", we have to say "of a 500mg tablet is.." And then we might need to be honest with the reader if that isn't a median of 5+ records, say, but is one record from one supplier who is only supplying one country that was in civil war in 2014. All the pricing so far suffers from gross over-extension in terms of claim vs what the source actually honestly tell us.

"Papers analyzing the databases" have names like "A Survey of Medicine Prices Availability, Affordability and Price Components in Shanghai, China Using the WHO/HAI Methodology". WP:MED readers will immediately recognise such as primary research papers. Looking at a couple gives me the shivers. I don't think they are useful on drug articles or even on drug-pricing articles. But let's pause a second and consider that this "primary research" is exactly what has been happening with all our drug articles to date. We have been doing original research on the raw data.

The reviews you cite seem to deal mostly with public health and I agree that any specific information they contain is likely to be a side-issue wrt the conclusions/focus of the paper, and filled with caveats.

The more I see of this sort of thing, the more respect I have for a field of researchers trying to make sense of an extremely complex topic and using their university degrees and PhDs and extensive professional experience in this domain to produce good research that moves health onwards. It is unfortunate that none of them seem to be editing wikipedia and could advise us the errors of our ways.

I don't think we can use the databases nor the research papers that investigate those databases. That's just too low level and requires original research. Maybe you could search more to see if you can find higher level sources that are closer to the sort of thing some people want to include in our articles. While there may be academic advice of methodologies, that doesn't mean editors here are allowed to do it. -- Colin°^Talk 19:17, 10 December 2019 (UTC)

A focused RfC is certainly preferable to an open ended RfC. An RfC where people clearly understand the options is obviously preferable to one which can be misinterpreted by participants and the closer. That's all great. However, people opposed to this change cannot say "find me an example I agree with before we offer it as a choice". The RfC is about what should be allowed. It is complex and the people who wish to include the information have the burden of convincing the whole community that what they propose is either compliant with other policies and guidelines or else that as part of the change they wish to amend those policies and guidelines. However, the people that need to be convinced of this is the community as a whole, not to the editors who have thoughtfully considered this issue and decided it's not appropriate. Best, Barkeep49 (talk) 19:25, 10 December 2019 (UTC)

Yes Barkeep49, that's exactly what I'm trying to do, to document the issue, so that others can evaluate and decide with more info. I agree for the need of a RfC, and as you say, with adequate and precise questions (and possibly examples), it would be much better I think. I'll try to summarize how the recent progress may pertain to the formulation of the future RfC. Colin: yes for the methodology, it's not our job, and the documents mostly pertain to conducting price studies, but it can help us navigate in the various methods available and avoid us from making bad choices (I hope). Signimu (talk) 21:22, 10 December 2019 (UTC)

I just saw this as a result of Signimu's subsequent post: However, people opposed to this change cannot say "find me an example I agree with before we offer it as a choice". Barkeep49, I fear there is a serious misunderstanding. I continue to ask for a sample not because I expect to agree with it (the methodology), rather to be fair in what sample we use to discuss the problems once the RFC is advanced !!! We should not pick apart one of the weakest drug article's pricing text; I want to know which is a drug article for which the sources used don't have as many of the problems as have been highlighted, in the interest of fairness. Do all of them have all of the problems Colin has highlighted? Do any of them have only a few of the problems? Is there any one that is better than the norm that will serve as a best-practice example?. We should use the best example the editors advocating for those sources have to put forward, so as not to be accused of having cherry-picked one of the worst articles as an example to highlight, and to avoid going through four, six or eight examples of all the problems. Bracket the problem by showing one of the best, and one of the worst when arguments are advanced in the RFC.

Had you not made that statement above, I would not realize that perhaps intent was being misunderstood here. (I also note that it was Signimu's closed/archived section below to led to a breakthrough in price/pricing, so again, let discussion run … we are making progress.) SandyGeorgia (Talk) 21:33, 10 December 2019 (UTC)

I am catching up here at the bottom of the last post re several items earlier on the page. I have to attend an unexpected and heartbreaking wake and funeral, and may be tied up for a few days, but will continue to weigh in as much as I can.

First, I do believe we are making progress, albeit of the "slowly but surely" type, perhaps not as quickly as some may want. While the adminning is greatly needed and very helpful, please consider not shutting down discussion that is finally happening after so many years of seeing !voting rather than discussing on this page. It is a complex issue, and that people are finally really engaging is, IMO, a very positive sign. A cultural switch to discussing rather than !voting is unlikely to happen in just a few days.

As can be seen on this page in the last few days, not all of the very involved parties have even had a chance to weigh in here, so we should remember that an RFC does not have to happen tomorrow. Hearing all voices, and getting all issues out on the table to give the RFC the best possible chance of reaching a useful conclusion is much more important than being in a hurry to put something up, that could be a debacle because of unexplored issues. While it seems like this conflict has been going on for a very long time, I suggest the real timeline started only a week ago, when everyone started engaging.

Another thing that can be seen on this page is that there is voluminous information to digest, and different editors are digesting that information (or not) at different rates. Yes, it is frustrating that this results in lengthy repeat posts, but that is part of the problem we must deal with. So again, no hurry. That we are seeing some progress in getting editors (not yet all, as I am concerned about those that have gone silent) to actually read the page, and digest and respond to concerns, is a very good sign. A better sign would be to begin to see less terse answers to the concerns being raised, to help avoid repeatedly asking the same question(s).

As to getting back to organizing the actual RFC, I am more concerned than ever that we not launch an RFC fraught with the kinds of issues that often crater RFCs. The idea of advancing specific question proposals was a good one, but the organization of that on this page is less than optimal. We need a name for the RFC and to begin to draft it in a subpage, with the idea of eventually moving it to a Wikipedia page. Next, we are risking having too many questions, which could be a killer; once we get to work, we can try to gel them gelled down. Also, the questions we are asking is only recently becoming clear even to many of us-- without an example of the problem that brought us here, it seems to me that the average editor coming to the RFC will miss the point, and respond with, well of course you can discuss prices following the restrictions and guidelines in NOPRICE. I continue to think we need an example that facilitates discussion of WHY we are asking what may appear to be a no-brainer question. We don't want to end up with a conclusion based on people not understanding what the dilemma is or what we are asking.

And finally, MANY of the questions we've advanced above are, considering the complexities here, oversimplified. All of the questions have been put forward in single sections, with lots of mixed-up discussion in each section. If we were to begin work on the actual text in a sub-page, we will begin to see what concerns remain and how the wording is coming. But I do not believe we are to the point of doing that yet: I do believe we are making progress. SandyGeorgia (Talk) 20:17, 10 December 2019 (UTC)

SandyGeorgia, my deepest condolences to you and your relatives, please take the time necessary for your mourning, that's certainly more important than whatever happens in Wikipedia... (and I agree about the slow but sure move forward) Signimu (talk) 21:25, 10 December 2019 (UTC)

Sandy, likewise from me. My deepest condolences and my best wishes. --Tryptofish (talk) 21:52, 10 December 2019 (UTC)

Thank you both very much. Not a family member, buta very close family friend; there is a bit more detail at Barkeep's talk. Thanks again, I will be following best I can here, SandyGeorgia (Talk) 22:01, 10 December 2019 (UTC)

References

Likewise, sending lots of good thoughts your way, @SandyGeorgia. — kashmīrī ^TALK 01:09, 11 December 2019 (UTC)

Thanks, Kashmiri; much appreciated. SandyGeorgia (Talk) 02:48, 11 December 2019 (UTC)

An alternative suggestion

I think it's very clear that, at some point in the future when editors here are comfortable with it, an RfC is going to be necessary. It's not optional.

After following this discussion for several days, I'm frankly very pessimistic that editors who are active here are ever going to agree on a series of questions (ie, should we use primary or secondary sources?) for the RfC structure. It's hard for me to envision a scenario in which there will be wide agreement about what any given choice, among multiple-choice options for a single question, even means.

So I want to suggest an alternative approach. Present the RfC respondents with a series of proposals, and ask them to choose among those proposals. For example, Proposal 1 might be to not list prices on every page, but instead discuss prices only when there has been a lot of independent discussion in sources – accompanied by reasons why editors think alternative proposals are no good. Proposal 2 could be to have prices in the infoboxes of every drug page, sourced to the WHO – accompanied by reasons why editors think this is a good idea. Or something roughly like that. And Proposal 3 would be something else, and Proposal 4 would be something else, and so on.

I think that most editors here either are able now, or expect to be able later on after more discussion, to say what they think the right solution is, and why. But they might not agree with what some other editors want to do, although they will be able to explain why they oppose those alternatives. Some editors might support Proposal 1, and oppose all other proposals; others will support multiple proposals while opposing just a few. I think it's feasible to develop such proposals, in whatever amount of time that will take, and maybe agree on some basic points about fairness regarding how each proposal is presented.

So the RfC would have the proposals listed by numbers, and there would be a !vote section below where each responding member of the community could indicate what they support or oppose and why, followed by a discussion section for more extended discussion. I think it would be good to have 3 uninvolved administrators determine the consensus at the end. Something like this worked surprisingly well at WP:GMORFC, in a subject area that was actually a lot more contentious than this one is. --Tryptofish (talk) 22:12, 10 December 2019 (UTC)

That sounds like a very reasonable and concrete way to formulate this RfC, plus it would allow to avoid the issue with presenting a combination of options that is impossible to follow (eg, lacking data). Please allow me to document a bit more what I read so that we can maybe weed out some more impossible choices. Signimu (talk) 22:32, 10 December 2019 (UTC)

And this is another of the reasons I am after a good sample drug: we can show under each of the proposed scenarios for that best sample drug what would be allowable. The complexity here may to be too much for average RFC (non-pharm) respondents to digest. And we are risking getting a non-meaningful result with all the complex questions. The other thing I like about this proposal is that it won't make it look like we are asking the community to endorse or reject what is already in NOPRICE. SandyGeorgia (Talk) 22:53, 10 December 2019 (UTC)

Ok, for example I would argue that generally there isn't "one price" for "a drug". For some there is. For some drugs that are highly notable for price issues, our best tertiary sources may simplify pricing and give "one price" for "the drug" (often as an example or rule-of-thumb or ballpark price). We could use those prices from those sources, as long as we caveat them as much as the source does. Pretty much all the existing examples of prices in our drug articles are based on cherry-picking raw database records from inappropriate sources, conducting extensive original research involving statistically/economically/medically unsound methods, and then presenting a statement to the reader. That statement either over-simplifies the facts or adds artefacts that misrepresent or is simply just totally unsupported by the source. We then compound this by juxtaposing incompatible prices without even informing the reader.

I suspect the response to my argument might be that the sources are the WHO and are offical international reference prices or respectable sources like the UK's BNF, and any mistakes that have been made can be fixed and would be more than happy to do so. Which, you know, is just 100% BS but that never stopped 100% BS from winning votes. The proposal 2 "prices in the infoboxes of every drug page, sourced to the WHO", is frankly making me lose the will to live. -- Colin°^Talk 23:27, 10 December 2019 (UTC)

Sandy is right that another potentially useful aspect of this approach is that it pretty much forces everyone to make it specific. I had forgotten to mention that, but I agree. And I certainly don't want to make anyone lose the will to live – but of course I only wrote those things as examples, not as what an actual proposal would be. I was the filing party in the GMO ArbCom case, and I was worried that the RfC would succumb to BS arguments, but what actually happened was that all "sides" got discussed and a rational result (or at least the proposal that I had written!) won out after 30 days. There were quite a few individual editor comments that made me want to facepalm, but overall the three uninvolved admins were readily able to determine a proper consensus. (And the GMO dispute makes this one look like a love-fest! I've lost count of the number of users who ended up topic-banned or site-banned.) --Tryptofish (talk) 23:48, 10 December 2019 (UTC)

Need to check out that one, too. Must be lots of fun to edit there! — kashmīrī ^TALK 01:12, 11 December 2019 (UTC)

Colin here is why I see this proposal working. The proposed questions we were presenting earlier boil down to asking the responders to deal with policy questions as if NOPRICE, NOR, WEIGHT, LEAD, etc didn't already exist (making it unlikely we would get a meaningful result because of the confusion of what we are asking). Then, each question we came up with was oversimplified or leading, with a complexity of issues that would be very hard to present (without specific examples of how the sources were used, and how that relates to WEIGHT, V, NOT, etc). Even if we were to finally decide on questions, I could not imagine what kind of neutral preamble we would come up with to introduce the conflict.

In this proposal, I envision a neutral preamble that begins with "Here's what NOPRICE, WEIGHT, LEAD and NOR say, there is a conflict about how to use certain sources for text about drug pricing". How do you see using these kinds of sources, and then offer the various options for the sources we are conflicted about.

• Preamble … conflict over how to interpret NOTPRICE, V, OR, LEAD for drug pricing in articles
  • Can a source like A be used to (and then provide Trypto's various option idea):
    • Option 1
    • Option 2
    • Option 3
    • Option 4
      • with the options for a sample drug summarized in a table that shows what text looks like for a sample drug for each disputed source. This gives a more natural format for one source, one drug, for us to lay out pros and cons in our responses.
  • Can a source like B be used to:

… and so on.

The heart of the dispute here is not whether drug articles can include prices: they can, per NOPRICE and subject to NOPRICE, WEIGHT, LEAD, etc. The core question is what kinds of sources can be used for what kind of text.

I haven't put nearly the amount of work you and Signimu have in to looking at the sources and the problems, so please feel free to tell me I'm wrong. I just think this is a more productive direction to head than the list of questions we have above.

Separately, I don't think we should be asking about prices in the LEAD at all in this RFC … there is already a guideline on LEAD, the question here is complex enough, and if WP:MEDLEAD persists in being a non-consensus section extending beyond and breaching the Wikipedia-wide WP:LEAD, we are likely to end up with another community-wide RFC to resolve that similar problem. I am hoping it will not have to come that, as this RFC will become a model for better consensus discussions occurring at WPMED. SandyGeorgia (Talk) 02:46, 11 December 2019 (UTC)

The thing is the "Can a source like A be used for ----insert article text here ----" is really a simple question of the sort Wikipedians deal with every day. If folk want, I can write up a very simple question of the "Can ... source be used to say ...." variety and we encourage some community discussion on it. It might resolve things much quicker and with less pain than an opinionated RFC about the emotive subject of pricing and Big Pharma censorship. -- Colin°^Talk 11:36, 11 December 2019 (UTC)

I think I understand the proponents argument for using databases directly is akin to using WHO ICD codes, we don't need a reliable source to mention each and every one of the ICD codes to be able to add them to entries. So in theory, if we had such a well established and complete database of medicines prices, I think there would be no argument against using that. The problem here is that we do have well established and wide databases of prices such as WHO/HAI or the MSH guide, but they aren't complete, as they are particularly missing high income countries (of course, their purpose is to improve accessibility, and accessibility is most often, but not always, an issue in low to moderate income countries). So are these databases enough to be used systematically for pricing, akin to WHO ICD codes? I'm not sure, and that may be a question for the RfC. It could be implicitly formulated through an example (and explicited in a description of this option). Signimu (talk) 13:03, 11 December 2019 (UTC)

No, Signimu; there is no original research in using an ICD code with citing it, and that practice is fully compliant with all policy, practically akin to not having to cite "the sky is blue". Not only do we not need a citation for common information, if we wanted one for ICD codes, we can find them in legions of MEDRS-compliant sources that present none of the kinds of issues these database-type sources do. And there is only one DSM and one ICD directory-- not multiple different interpretations using different rubrics. The OR that is being used to price drugs in contested articles is not available-- at least in any case presented so far-- in any source other than database types. (If that data were available in a source that doesn't require OR, we could use that source and avoid these problems.)

I think this discussion has also shown that there can be no such thing/never will be any such thing as a "well established and complete database of medicine prices", because that is not how medicine pricing works (see my description of drug prices in the US). I think the comparison of WHO/HAI/MSH to WHO ICD codes will get us off on the wrong track-- ICD codes have very little variability (allowing for the simple descriptions of where they differ from DSM).

I agree that we should be asking if WHO/HAI/MSH can be used as they are being used: I disagree that there is any similarity to ICD codes, and am reluctantly seeing that Colin may be correct in asserting that we are not engaging the core dilemma here.

Rephrasing: basically the approach that Tryptofish and I are advocating asks the simple question about whether these kinds of sources can be used as they are, by giving the RFC respondent a choice of how to use the source. That is the correct question. Thus far into the RFC formulation, we may not have yet fully engaged the core issue, and I fear that equating ICD codes to the dilemma with drug prices may mean that we may still have more work to do to make sure we all understand the problems. SandyGeorgia (Talk) 14:57, 11 December 2019 (UTC)

SandyGeorgia, please allow me to have a slightly different perspective, and opinion. Until recently, I did not know about WHO ICD codes. Maybe for you this is the de facto, and unique, standard, but that's not the case, there are several other diseases codification schemes. Diseases codification is of course less dynamic than prices since there is less impact from the market (the only impact probably being R&D allowing to discover and treat new diseases), but it's still dynamic and flexible, it's a human mind construct. There are different codifications depending on the interpretation and purposes, and even controversies as are well known for the DSM. That said, I'm not saying that OR is not an issue, obviously it is and we should provide examples that do not fall in this issue. I can give you an example of a pricing example straight out from the WHO/HAI database involving 0 OR, in the two figures of: median price ratio. Again, it's not to say we should do that, but although OR is of course a concern to have, using databases does not necessarily involves doing any level of OR. --Signimu (talk) 15:01, 11 December 2019 (UTC)

PS: to be clear, I'm not saying I agree with this view, as I said before I see merits in a lot of arguments from the various sides. I'm trying to understand all points so I can provide adequate documentation to help us make sense of this issue. My point is that yes OR is a concern, and databases may be at higher risk of misusage, but databases != OR if used properly. --Signimu (talk) 15:15, 11 December 2019 (UTC)

My sincerest apologies, Signimu for overstepping in my response. I believe your perspective and hard work here has been invaluable, and I do see your broader point about how/when we can apply database info. I suppose I reacted strongly because even though I dislike infoboxes because they often oversimplify nuanced information, even I am not opposed to having ICD codes in infoboxes, even without citation, as they are so standard and widely accepted (controversies you mention notwithstanding).

Also, because discussion on this alternate proposal is advancing, I am hoping @WhatamIdoing: will provide feedback. SandyGeorgia (Talk) 15:09, 11 December 2019 (UTC)

No need for apologies, you did nothing wrong Ah the whole issue around infoboxes is... a separate issue to say the least. I'm not even sure that's possible. That's why I agree Tryptofish's suggestion is great, it's better to provide an example of a 'good' or at least 'possible' option rather than propose something that may be in fact impossible for us to do. --Signimu (talk) 15:20, 11 December 2019 (UTC)

Thanks everyone for this useful feedback. I feel cautiously optimistic about how this part of the discussion is going. I like the idea of asking about whether "this source can be used for this text" or not. One option (certainly not the only one), if we ask RfC respondents to support or oppose various proposals, is to have something like Proposal 6 be "Source X may be used for example text", and editors could support or oppose that statement along with the rest. Such "proposals" would be very brief, so they wouldn't be much to read.

I do encourage editors who have not yet expressed their opinions of the alternative suggestion to do so. And if we continue to have support for it, I think it would be fine for anyone who wants to (and no hurry at all about this!) to start a new section of this talk page called "Proposal 1" and start workshopping possible text for it. I think the fairest, most neutral, way to number the proposals is that the first one to start composition here is #1, the next one #2, and so on, with the short statements about source acceptability coming after. --Tryptofish (talk) 18:44, 11 December 2019 (UTC)

I think that this might be an interesting source to consider, because it doesn't have many of the usual complexities:

At a price of more than $600,000 for a 25kg patient on a WAC basis and with 8% of all DMD patients amenable to exon 53 skipping, Vyondys has an annual U.S. total addressable market opportunity north of $500 million, Credit Suisse analyst Martin Auster said in a note.

The drug only exists in one country. It (so far) only has one price (per amount of drug, not per patient). What, if anything, could you use that for? WhatamIdoing (talk) 02:18, 15 December 2019 (UTC)

Firstly, the article is I suspect little more than a reformatted press release, and as such, somewhat dodgy wrt reliable sources. It wasn't entirely intelligible, and I don't know whether that's because the subject is very complex, the writer didn't actually understand the subject, or I haven't had my second cup of tea yet today. This kind of article (assuming we had one written by a business editor of a professional business publication, rather than a press release) seems more appropriate for the economic area of the body of a drug, and if this is a key drug for the company, then the company article. Hard to know what if any of that belongs in the lead, other than perhaps something about approval. I found Understanding Drug Pricing interesting wrt "on a WAC basis" and their comment "The wholesale acquisition cost (WAC) is an estimate of the manufacturer’s list price for a drug to wholesalers or direct purchasers, but does not include discounts or rebates.3 Without including rebates and other incentives provided by manufacturers, it is hard to estimate the actual cost of the drug.". We see the same in the UK with some new cancer treatments that are very expensive having an official price but then a "patient access scheme" to discount. This scheme can be very complex and can also at times be confidential. So is the "At a price of more than $600,000 for a 25kg patient" little more than a boast to encourage the market to buy shares of the company (I assume this is completing treatment, though drug cost will only be part of the treatment cost). As such, the press release does not need to care that the $$ figure is inaccurate and merely suggestive of "extremely large". We would, for example, be very wrong to write "It costs $600,000 to treat a child with DMD using the drug Vyondys. But then, being "very wrong" hasn't seemed to matter wrt drug prices. -- Colin°^Talk 11:05, 15 December 2019 (UTC)

This is not a press release; it's a business article written by a professional journalist, who quotes a business analyst to estimate potential future revenues for the manufacturer. The price given is for the (alleged) drug costs, not the full cost of the treatment (e.g., the cost of doing the IV infusion).

Would you accept a statement like "One business analyst estimated the manufacturer's initial list price at approximately US$600,000 per 25 kg patient"? WhatamIdoing (talk) 03:12, 18 December 2019 (UTC)

WhatamIdoing, here is the press release. I can find only tiny portions that do not come directly from the press release, with mixed competence in paraphrasing vs straightforward copy/paste. This is Churnalism: reformatted press release. I have no idea if "a note" from Martin Auster is considered a reliable source. Seems he gives ephemeral buy/sell market tips. I'm actually struggling to get the $600,000 number independently. The press release states it is "priced at parity to Exondys 51, the price of which has not increased since its launch in 2016". This article claims that net annual cost is "$300,000 per patient, less than Wall Street analysts had thought". Presumably those "Wall Street analysts" include the likes of Auster. The article also notes that 25kg is a 7-8 year-old boy, and the price will increase with age, perhaps to $450,000. And this is per year (for life?) whereas I read the statement as a total treatment cost. Further, the efficacy of both drugs and lack of good quality trial data seems to be highly questionable, raising concerns about whether we could claim "to treat" while meeting MEDRS (probably not, yet). It seems the jury is out on that matter, and that they are going ahead with approval on basis of parent's request, and to seek evidence as it becomes used. So while I can find sources suggesting $300,000 parity price, the only sources for $600,000 cite Auster. Perhaps we should give stock analyst market tips as miss when it comes to this. It seems terribly hard to get it right. -- Colin°^Talk 13:19, 19 December 2019 (UTC)

IMO "getting it right" can be accomplished by making the claim small enough. One analyst said $600K for the specified size of child, but perhaps we should give more WP:WEIGHT to the $300K source. (Of course it's cost per year; that's typical for discussing the cost of DMD drugs.) WhatamIdoing (talk) 16:27, 23 December 2019 (UTC)

NOR sample discussion

User:Signimu your investigations and research into pricing sources, etc is very helpful and progressing things, but I think you are looking at this from a global-health researcher POV rather than a Wikipedian POV. Btw, MSH guide is incomplete for more than just high income countries but also most modern drugs and even for the drugs it does include, it very often has few, one or no suppliers for pill sizes, and of course, has no prices since 2015. Our opinion on how complete/good/etc the MSH data is and what WHO recommended methods are used to extract this into meaningful results for public health research is one thing and perhaps lets us judge if tertiary sources are really doing a sound job. But from the Wikipedian point of view, it is really quite simple. Let me cite some policy from NOR.

• "Wikipedia articles must not contain original research...facts...for which no reliable, published sources exist. This includes any analysis or synthesis of published material that serves to reach or imply a conclusion not stated by the sources. To demonstrate that you are not adding OR, you must be able to cite reliable, published sources that are directly related to the topic of the article, and directly support the material being presented.
• "Take care not to go beyond what is expressed in the sources....The only way you can show your edit is not original research is to cite a reliable published source that contains the same material....Drawing conclusions not evident in the reference is original research regardless of the type of source.
• "All analyses and interpretive or synthetic claims about primary sources must be referenced to a secondary or tertiary source, and must not be an original analysis of the primary-source material by Wikipedia editors"
• "A primary source may be used on Wikipedia only to make straightforward, descriptive statements of facts that can be verified by any educated person with access to the primary source but without further, specialized knowledge."
• "Do not analyze, evaluate, interpret, or synthesize material found in a primary source yourself; instead, refer to reliable secondary sources that do so"
• "Do not combine material from multiple sources to reach or imply a conclusion not explicitly stated by any of the sources. Similarly, do not combine different parts of one source to reach or imply a conclusion not explicitly stated by the source. If one reliable source says A, and another reliable source says B, do not join A and B together to imply a conclusion C that is not mentioned by either of the sources"
• "Routine calculations do not count as original research, provided there is consensus among editors that the result of the calculation is obvious, correct, and a meaningful reflection of the sources. Basic arithmetic, such as adding numbers, converting units, or calculating a person's age are some examples of routine calculations"

Take that and examine the following article texts:

• Ethosuximide: The wholesale cost in the developing world is about US$27.77 per month as of 2014[63]
  • The number $27.77 does not appear in the source. This particular drug has one pill size in MSH for 2014 and only one supplier. However, the supplier ASRAMES only supplies the north of Democratic Republic of Congo using their own small fleet of vehicles [i.e. not "the developing world"]. The maths done to reach the price in the article is
    • Use Defined Daily Dose as the "daily dose". This is contentious as the DDD is an artificial metric created for research into population usage of drugs, and not drug costing. It is explicitly not a therapeutic dose and may in fact not be related to any therapeutic dose used in practice. This drug is mainly used in paediatrics.
    • Use 30 as a month supply. In the UK the package size is 56 (for 28-day months). In the source, the price is for a bottle of 100 tablets. In the UK and US oral syrup is also available.
    • ( Defined Daily Dose [1250mg] / Strength [250mg] ) * Unit Price [$0.1845] * Month [30] = $27.675. Round to 2 decimal places.
• Carbamazepine: The wholesale cost in the developing world is about US$0.07 to US$0.24 per day as of 2015[64]
  • The numbers $0.07 and $0.24 do not appear in the source. Searching for Carbamazepine in 2015 produces 5 results though 2 are to the same URL (just different therapeutic categories). The 100mg/5ml syrups has no suppliers and five buyers.The 200mg sustained release pill has no suppliers and two buyers. The 400mg sustained release pill has no suppliers and one buyer. The 200mg standard tablet has 12 suppliers and 4 buyers. The maths done in the article is
    • Pick the 200mg standard tablet. The database does not indicate which to pick, nor is there an established/official method for picking one in any literature.
    • Take the lowest unit price [0.0138 from the Suppliers] and the highest unit price [0.0480 from the Buyers].
    • ( Defined Daily Dose [1000mg] / Strength [200] ) * Unit Price [$0.0138] = 0.069. Round to 2 decimal places.
    • ( Defined Daily Dose [1000mg] / Strength [200] ) * Unit Price [$0.0480] = 0.24
    • Although we give a min and max price, we have taken the min and max from different kinds of price (supplier/buyer) but not from different strengths or formulations, which may be cheaper or dearer.
• Mebendazole: The wholesale cost in the developing world is between USD 0.004 and 0.04 per dose.[65]
  • The numbers 0.004 and 0.04 do not appear in the source. Searching for Mebendazole 2014 produces 5 results. The 100mg standard tablet has 6 suppliers and 5 buyers. The 100mg chewable tablets have 4 suppliers and no buyers. The 100mg/5ml syrup has 6 suppliers and 3 buyers. The 500mg standard tablet has one supplier and one buyer. The 500mg chewable tablet has 3 suppliers and no buyers. The maths done in the article is
    • Pick the 100mg standard tablet. The drug is used for lots of different parasitic infections. Drugs.com generally indicates the chewable tablets and either the 100mg as a single dose with another perhaps in two weeks, 100mg twice a day for 3 consecutive days, or the 500mg tablet once as a single dose. Some require 200-400 three times a day for three days, then 400-500mg orally 3 times a day for 10 days. This is not a drug taken long term, but in short treatment duration that is quite variable in strength/length depending on indication.
    • Assume the "dose" is 100mg. The source does not indicate what a "dose" is. It does give a "defined daily dose" of 200mg but see above for that.
    • Round the lowest price from the suppliers [$0.0035] to three decimal places [$0.004] and round the highest price from the buyers [$0.0393] to two decimal places [$0.04]. As above, the min and max come from different kinds of price (supplier/buyer) but not from different strengths or formulations (e.g. chewable or 500mg).
    • The article does not give a treatment cost even though the length of treatment is knowable depending on indication (from other sources). The article does not state what treatment that dose cost is for.

I don't think it is rocket science to see this is original research of the most obvious kind. The NOR point "can be verified by any educated person with access to the primary source but without further, specialized knowledge" is particularly useful here. There is nothing in the source to suggest which record to choose. There is some advice on the help pages about buyer and supplier prices and about perhaps using a median, but nothing concrete enough to say what our educated person might actually do, and our articles have generally ignored the advice, and statistical common sense about avoiding outliers, buy choosing the min and max. A careful reader might question if buyer and supplier prices can really be combined, as they have in our articles. The help pages also have some advice about DDD, which is generally off-putting in terms of thinking it should be used for pricing a daily treatment cost (e.g. " the ATC/DDD system by itself is not suitable for guiding decisions about reimbursement, pricing, and therapeutic substitution."'). There is nothing in the source about whether a daily dose is divided (usually in two) or single. Nothing to indicate starting dose and maximum dose and titration increments which strongly influence a choice of pill size the doctor prescribes. Nothing to explain for example that an enteric-coated pill or a capsule cannot be split in two half doses, but that dose variation is simple with a syrup (though watch out for the per 5ml gotcha). Nothing to suggest picking chewable or enteric-coated vs standard. Nothing to suggest e.g. that paediatric usage might suggest a syrup. Nothing about treatment duration (continuous, pulsed, short-term, one-off). Nothing to say what indication the DDD was calculated for. Nothing to suggest additional costs such as blood level monitoring or follow up appointments. Nothing to explain what kind of wholesale cost a supplier price is, or the many caveats to use when considering the relevance (if any) of a buyer price. -- Colin°^Talk 16:23, 11 December 2019 (UTC)

Here's a very brief summary of the other database-level sources we have used:

• NADAC for "US wholesale cost ... per month". This starts off bad with a database result for a historical week but no selection on drug at all (25,194 records returned). If we enter a drug name in the "Find in this dataset" box we again get a mix of syrup and tablets of various kinds and mg sizes. Might also get variants on the drug itself such as different salts or in combination with other drugs. The NADAC does have an average price, so we don't have the problem of multiple prices per supplier/manufacturer. There is no dose information or notes on what therapeutic purpose the drug might be used. So no way to use this site to calculate a "per dose" or "per day" or "per month".
• Drugs.com for "In the United States, it costs about US____ per dose". Website return various sizes and formulations (e.g. 10mg, 25mg, 50mg, 75mg, 100mg, 150mg), syrups, capsules, etc. Webpage does not explain e.g. low dose is for neuropathic pain off label in US, higher doses for mental illnesses licenced indication. No information on this page about typical dose or how many tablets to take per day or how to divide doses. Price-per-pill reduces considerably the more pills you buy in one go.
• BNF for UK prices (might not be accessible outside of UK). Again multiple size tablets, syrups and capsules, and different package sizes. Multiple manufacturers, some with different "NHS Indicative Price". Most records have "Drug tariff price". No indication on website which price to pick or what they mean in terms a normal "educated person" might grasp. No indication if drug is subject to special discounting (due to e.g. patient assistance program) or currently has a higher price (due to shortage). Although other pages in the BNF give dosage advice (for professionals) this of course varies per indication or patient age and health. So no way to determine one "daily dose" or "individual dose".

All these are highly respected sources of raw data about prices. Nobody is questioning the accuracy of the data. -- Colin°^Talk 18:42, 11 December 2019 (UTC)

As several of us have now pointed out that this is the core issue affecting a potential RFC, I have posted a query for further feedback to the NOR noticeboard. I suggest we cannot put together an RFC that is likely to yield a useful conclusion without getting to the heart of the NOR matter. SandyGeorgia (Talk) 17:58, 11 December 2019 (UTC)

Well, I see that noticeboard is just hopping with activity. Perhaps by next week we need to start putting up some concrete proposals of format for the RFC. I still think we need a name for the RFC, and a draft page for coordinating work. SandyGeorgia (Talk) 15:11, 13 December 2019 (UTC)

I will do, hopefully today but more likely on sunday, hard to manage such a big issue along with work :-/ Signimu (talk) 19:45, 13 December 2019 (UTC)

• This is going to end up at ArbCom. The task is to formulate a question for a central RfC. Nobody's doing that, all that's happening is endless rehashing of the answer you think the RfC should return. Guy (help!) 23:22, 15 December 2019 (UTC)

• JzG, then I must be misreading or missing something, because my impression is that both WAID and Signimu have indicated that they are working on it. From my end, it's holiday season, and hashing out wording isn't the spirit I'm in. I have said multiple times that we first need a name and a draft space: it is my understanding that Signimu has said they are on it. SandyGeorgia (Talk) 00:13, 16 December 2019 (UTC)

• See my comment below -- Colin°^Talk 12:31, 16 December 2019 (UTC)

• It's difficult to create an RfC when the content policy seems clear, and those asserting exceptions for policy are almost totally silent.
• For the RfC, we need to distinguish questions about pricing from questions about specific prices. --Ronz (talk) 16:00, 16 December 2019 (UTC)

• About the problem of some editors seemingly being silent, I have a suggestion on how to jump-start it. Just go ahead and start formulating RfC questions that represent the proposals that you would likely support. At some point, it will become pretty obvious that the RfC is going to happen, and when previously silent editors realize that their perspective might not be presented, that will have a tendency to focus the mind. And if that doesn't work, then propose something for them, and make them realize that they would be better off revising it so that it doesn't look bad. And if that doesn't work, then let the RfC turn out overwhelmingly to support what you are hoping for. --Tryptofish (talk) 20:03, 16 December 2019 (UTC)

Are pharmaceutical prices encyclopedic information?

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

---

Now that I caught up with the latest developments (sorry I could barely connect the last week due to IRL things), I would like to suggest to discuss what I think is one of the most essential issues: whether pharmaceutical prices are encyclopedic information. There are other crucial questions, such as the methodology (what prices to use?), editorial (where to place prices - lede, infobox, article's body, wikidata? what sources are acceptable?) and others, but I think this issue precedes the others, so I'll start here with that and later write about the methodology in a separate section. Since finding sources is something I like to do, and I like to learn new things, I will here show the results of my findings on this question, which may help further our collective understanding and discussion. Before going further, please note I had no in-depth knowledge of the whole issue around medicines prices before the new discussion started at WT:MED#MEDPRICE, and I here applied the same WP:MEDRS methodology we use for finding reliable sources.

• WHO stance on prices (emphasis mine): "Access to essential medicines is part of the fulfilment of the right to the highest attainable standard of health (in short: the right to health). So why do millions of people across the globe go without the treatments they need? The reasons are now becoming clearer – and the price and availability of medicines to those who need them are crucial factors."

• United Nations stance: "Greater transparency, fairer prices for medicines ‘a global human rights issue’, says UN health agency"

• The latest World Health Assembly approved a resolution on price transparency: "Seriously concerned about high prices for some health products, and inequitable access to such products within and among Member States, as well as the financial hardships associated with high prices which impede progress towards achieving universal health coverage; Recognizing that the types of information publicly available on data across the value chain of health products, including prices effectively paid by different actors and costs, vary among Member States and that the availability of comparable price information may facilitate efforts towards affordable and equitable access to health products; Seeking to enhance the publicly available information on the prices applied in different sectors, in different countries and the access to and use of this information, while recognizing different national and regional legal frameworks and contexts and acknowledging the importance of differential pricing;"

• Vogler, researcher at WHO (emphasis mine): "Ensuring affordable access to essential medicines (i.e. those medicines that satisfy the priority health needs of the population) is a major policy objective globally. One-third of the world's population; however, is estimated to have limited or no access to essential medicines. One of the key barriers is the high price of medicines. [...] Pharmaceutical pricing policies are key because medicines are no normal goods and health care, including pharmaceutical, systems are not normal competitive consumer markets."

• A Lancet commissioned review paper: "Governments, national health systems, and the pharmaceutical industry must promote transparency by sharing health and medicines information."

To summarize: the sources above, and virtually all sources I have, have noted the specificity of pharmaceutical prices, which directly impacts medicines affordability and thus efficiency. All sources agree price is not a side parameter, but a "crucial" or "key" factor. It's not like the prices of other consumables, where the price is mostly/only an economic factor: medicines prices are also and foremost a health issue.

Now let's go back to Wikipedia's policies and RfCs:

• WP:NOPRICES states: "Sales catalogues. An article should not include product pricing or availability information unless there is an independent source and a justified reason for the mention. Encyclopedic significance may be indicated if mainstream media sources (not just product reviews) provide commentary on these details instead of just passing mention. Prices and product availability can vary widely from place to place and over time. Wikipedia is not a price comparison service to compare the prices of competing products, or the prices and availability of a single product from different vendors or retailers."

• 2016 RfC Wikipedia_talk:WikiProject_Medicine/Archive_84#Price_of_medications (emphasis not mine): "Except in the cases where the sources note the significance of the pricing (which did have consensus), there is no consensus to add the pricing to the articles. There were several ways to present this information proposed in the discussion (such as wikidata). Please feel free to start a follow-up discussion regarding that."

We have several sources above that independently assert the importance of prices, for all medications. I have yet to find any reliable source that would qualify pharmaceutical prices as irrelevant. Not to mention that a wide array of countries are actively implicated and/or implementing medicines price policies, external reference pricing being just one of them. Since virtually all sources seem to agree on the importance of medicines prices as a significant (even "crucial") health factor (contrary to prices of other consumables), and the worldwide consideration of various entities such as health organizations, national and transnational regulatory bodies, and researchers, it looks to me that medicines prices may be considered encyclopedic information per se, just like we consider epidemiological information as encyclopedic without requiring each source to specify why epidemiology is important for a particular disease, the epidemiology is simply a factual bit of knowledge that stands by itself, and furthers our knowledge on the topic (eg, the prevalence of an illness is important by itself, we don't need to put this in perspective with other diseases).

What do you guys think? --Signimu (talk) 18:46, 9 December 2019 (UTC)

PS: to be clear, even if we reach here an agreement that prices may be encyclopedic in theory, we would still need to resolve other issues as stated above by me and others, such as the methodology (eg, if we can't find a common ground with a good methodology to base the pricing, we can't add prices anyway). Signimu (talk) 18:53, 9 December 2019 (UTC)

References

• Thank you Signimu for your thoughts. Just a minor although I think important comment from my side. When the sources you quoted mention prices, they talk about them in relative terms: prices matter only to the extent as they hinder access to medicines. Prices are discussed as relative to the wealth of a given society, to its per capita GDP, etc. But here on WP the entire discussion has been about absolute pricing – i.e., disconnected from any socioeconomic context and simply presented in crude monetary units.

In my view, it absolutely is encyclopaedic to add information that a cost of a specific therapy has prevented access to it in specific countries. I even have no problem with listing countries where a given drug is largely not available due to cost.

But dumping raw pricing data into all the drug articles on Wikipedia - no, I do not think this is the type of information that any encyclopaedia should provide. — kashmīrī ^TALK 19:08, 9 December 2019 (UTC)

Thank you Kashmiri, to be accurate, yes and no, let me explain: prices are mostly studied in relation to essential medicines accessibility, because that's the goal the UN and WHO were tasked (or tasked themselves) to do. Usually, pricing and accessibility issues arise in low and moderate income countries, but they also happen in wealthy countries: it was reported that hepatitis C medicines, which are now part of the essential medicines list, are too expensive even for wealthy countries, and following the 2008 global crisis, european union countries which were affected reported having restricted access to essential medicines. That's why this same source writes: "Affordability of medicines has become a key issue for governments, as well as public and private payers for health care, regardless of a country’s income level." I however agree with you that a dump of raw prices is not encyclopedic, but we should keep that for the discussion around the methodology of what kind of measure we can use (which I am still researching). A preliminary example of what is possible can be seen at median price ratio, although I don't think this can be generalized to multiple articles as this is limited to LMICs countries. Signimu (talk) 19:17, 9 December 2019 (UTC)

• (after edit conflict) Signimu, I suspect the answers to your well-positioned query (thanks for the analysis) are buried in the lengthy discussions above, so at the risk of repetition (no bludgeoning intended :), my summary response is that:

• Yes, MEDRS-compliant sources address the importance of pricing, and that information would be relevant to a broad, general article on drug pricing; but,

• No, we don't have any indication that the sources put forward so far are adequate to add the content that has been added to many/most drug articles, and we don't even have a best-case example to go on. As we've discussed, we can specifically discuss pricing in some cases (sample, insulin and epipen) because the sources do the work for us, but we are using database-style sources that carry a multitude of problems discussed above to add pricing info as of now.

Overall, while we possibly all agree that prices are important per MEDRS, the specifics applied so far to most drug articles are non-compliant and even misleading. That is the dilemma. I feel we are spinning our wheels until we get one, good, compliant example that does not involve breaching NOT, WEIGHT, V, LEAD, NOR and MEDRS. Short answer: for specific drugs the importance and relevance of pricing/cost must be based on good sources for and about that specific drug, which is not the case that is being advanced now. Pricing content for an individual drug is encyclopedic, when specific sources, compliant with policy, are used to add the relevant information, subject to policy and guideline. SandyGeorgia (Talk) 19:14, 9 December 2019 (UTC)

• Signimu I think your initial quotes about the importance of drug pricing is sufficient to justify a set of articles on the issues and methodologies and projects worldwide. But they don't seem convincing, as Kashmiri notes, to dump raw dollars and cents prices into our articles. Think of the basics for life: a home, heating and electricity, food, water, education, health (not just drugs), etc. These are all important and countless words have been written about affordability of food [In the UK, we have a scandal of being one of the richest nations on earth and yet those working in the poorest jobs, which can include nursing and teaching, can find themselves reliant on food banks]. But the affordability of food as a world issue, is not being used to justify prices of mince pies and others into every single food article. That's what we see here: a plea to emotion.

See other posts I've made today. Even for drugs that are notable for their high price like Lenalidomide, Sandy I'm sorry but we don't even get those right. The article juxtaposes a retail US price (without saying it is retail) with a wholesale UK list price and claims that is the price it costs the NHS. But the BNF book does note that this eyewateringly expensive drug is subject to a "patient access scheme" that heavily discounts the price to the NHS. The discount for some of those schemes is public and very complex and for some confidential.

Read also my post above about "out of pocket" prices paid by those in developing worlds. The WHO document cited basically busts to smithereens the idea that prices in the MSH price guide bear any resemblance to the often private prices paid by patients in such countries, if they have access at all. The misuse of MSH data is basically an example of how many 1001 ways can you completely screw up when doing original research. Guy's "price sharply increases" observation is another good example. Perhaps, when this is over, someone can use it as a textbook example of why original research is a bad idea. -- Colin°^Talk 19:40, 9 December 2019 (UTC)

Please let's keep methodological issues, which are real, for another discussion, else if it's mingled with the issue of pertinence I fear we will go in circle without reaching any agreement. Your point and analogy with food is very interesting, thank you. Signimu (talk) 19:59, 9 December 2019 (UTC)

Pricing in general is clearly important, and I earlier pointed out Medication costs, Prescription drug prices in the United States, Prescription charges [66].

We generally don't use references that don't mention a specific item to determine weight within an article about a specific item. (That's a horribly convoluted sentence, but I hope it makes sense.) So, while prices are important, that doesn't mean we've met NOTPRICE criteria for an article about a specific medicine unless we have a source clearly showing the importance of prices for that specific medicine.

Prices may not exist in a form that is suitable for encyclopedia articles at all, as all the discussions are indicating.

To answer the question of this subsection: No. Encyclopedic prices don't appear to exist.

Meanwhile, I think a good case could be made for giving extra weight to pricing information when it's available: Information about the ranges in pricing and the many factors involved. --Ronz (talk) 20:09, 9 December 2019 (UTC)

I agree very much with Ronz's distinction between prices and pricing. --Tryptofish (talk) 20:52, 9 December 2019 (UTC)

Yep, Ronz gets the Succinct Prize for the day, with the price/pricing distinction. SandyGeorgia (Talk) 20:58, 9 December 2019 (UTC)

Your pricing/price distinction captures what I was qualifying as "prices may be encyclopedic in theory", I should have written "pricing". Signimu (talk) 21:49, 9 December 2019 (UTC)

I like your pricing/price distinction, but pricing needs price metrics, which complicates things (because it requires discussing the methodology of 'what price metrics to use'). I agree however that the sources could be interpreted that way (the pricing, or affordability/accessibility, of medications is a health issue, not the price per se). --Signimu (talk) 21:56, 9 December 2019 (UTC)

And that seems to me to be a good reason to treat the matter as something to be discussed in the text, based on independent secondary sources, as opposed to just listing numbers. (Of course, other editors will disagree.) --Tryptofish (talk) 22:03, 9 December 2019 (UTC)

References

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

While I agree that we don't need to have the discussion right now, this section heading is a potentially important question. If editors believe that some kind of information about money is normally appropriate to include (subject to the limits of the available sources, just like everything else), then we need to know that. WhatamIdoing (talk) 01:06, 15 December 2019 (UTC)

Wikipedia is full of prices of products

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

---

We have an entire article on the price of oil, with multiple sub pages. We have a section on Gasoline#Use_and_pricing in various parts of the world. And than an entire subpage on the topic.

For other commodities we see the price of wheat, milk, and rice. For consumer products we have the price of the Nintendo_Switch (both in the infobox and body of the article in various currencies), Nintendo_Switch_Lite, Apple Watch Series 3, Honda Legend, Acura RL, and Pentax K20D.

Now if we just want to look at just featured articles, we include prices in Sega Genesis, Wii (in the infobox and body), Sega Saturn (in the lead and body), NeXT, Macintosh Classic, and PowerBook 100 (in the lead, infobox, and body).

We include prices for paintings The Triumph of Cleopatra and Streatham portrait. We include prices for elements such as Xenon and Germanium. We include prices for universities such as University of Michigan and Shimer College#Tuition and fees. We include prices of Durian and Maple syrup. We include prices for water fluoridation.

I am happy for us to continue including all of these things. Including the prices of medications is not out of line with current practice both generally and for featured articles. Doc James (talk · contribs · email) 01:41, 10 December 2019 (UTC)

Given the discussions we've had, it's clear that determining a price for a medication that we can use in an encyclopedia article is nothing like any of these examples. --Ronz (talk) 01:46, 10 December 2019 (UTC)

Doc James, have you read this section? You are repeating arguments that have already been addressed in several sections on this page. WP:OTHERSTUFFEXISTS, and yes, there are many instances of pricing that comply with the restrictions of WP:NOTPRICE.

What is needed next for us to be able to advance towards an RFC is for you, Quack Guru, or anyone else to provide a sample of your strongest, tightest, most accurate drug article on pricing, that does not have the multitude of problems that have been outlined on this page, so that we can advance RFC questions based on what you consider to be your best practice example. Many people here have expended considerable effort in analyzing and discussing the issues to be resolved: please give us what you consider to be the best-sourced drug article from the sources you want to use, so that we can begin to address things with specificity. Reading this section for a list of drug articles that have already been well discussed here may help. SandyGeorgia (Talk) 02:16, 10 December 2019 (UTC)

Sure I have tried here.[67]

And happy to continue if others are interested. This at Xenon "approximate prices for the purchase of small quantities in Europe in 1999 were 10 €/L for xenon, 1 €/L for krypton, and 0.20 €/L for neon" is what we do sometimes do for medications. Here we have a very specialized source used to support it.[68] One could summarize this table Wii_launch#Release_dates_and_pricing by just giving the highest and lowest numbers. But meh.

Including prices of commodities and products is common practice. Just noticed you went through a number of them above. Doc James (talk · contribs · email) 02:33, 10 December 2019 (UTC)

I'm not sure any of you are going to be able to convince each other of anything at the moment. Not for lack of knowledge or skill, but simply because you have discussed this so much and still have a good faith disagreement. I would suggest that your energy might be better spent coming to consensus on the RfC questions as that, in my estimation, is your best way out of this morass. However, if you all feel like you're making progress please ignore me and continue your discussion :). Best, Barkeep49 (talk) 02:48, 10 December 2019 (UTC)

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

WhatamIdoing's summary

I've been taking notes, and I've left comments elsewhere, but some themes come up in multiple sections, so I'm skipping down here to tell you what this looks like to me at this point.

First, I agree with User:SandyGeorgia that we're really making progress. This section has some 300 comments and about 50,000 words on how to address these questions. I know that long discussions on complicated subjects aren't everyone's cup of tea, but I like them. Hang in there if you can, and don't worry if you can't.

Some things that have really been helpful include:

• User:Signimu and User:Colin in particular have done a really amazing job in identifying the complexity of "the price" for "a drug". I hope that drug pricing will be improved while people still have this information in mind and sources at hand, because I strongly suspect that some editors will be relying on our articles to figure out their answers to the RFC.
• SandyGeorgia proposed providing ==External links== to a drug price database. This kind of out-of-the-box thinking is really helpful, and the particular idea is something that we might want to do no matter what else is decided.
• User:Ronz had a particularly good comment on the difference between "prices" and "pricing". We need to be clear about which one(s) we want in which circumstances.
• Several people have identified specific details that future editors will want help with, such as User:Tryptofish's point about whether prices/pricing is better suited to articles about individual drugs (e.g., Diphenhydramine) or articles about a class of drugs (e.g., H1 antagonist or Antihistamine). The only point that I didn't see mentioned recently was the effect of patent status on drug pricing, but that was touched on in the pre-ANI part of the discussion.

Next, it seems to me that there's no way to write this RFC until we have settled the question about how to present the decision to editors. That means that we have to decide whether we're asking questions that identify relevant principles or proposing guideline text. I can go either way with this. If people answer our questions, then there are several people in this discussion who can write guideline text to align with their answers; if editors instead pick a proposed addition to the guideline, then we'll know the answers to our questions. This is not a decision that will change what editors say, but it is necessary to decide this now. (Feel free to actually vote, straw poll-style, for your preference. I can work with any outcome on that point.)

Among the open questions, the main one seems to be whether (other) editors think that drug prices should normally be included or normally be excluded. Yes, NOTPRICE says to normally exclude, but some editors are uncertain whether that applies as strictly to drugs as it does to T-shirts. One way to put this is, are we talking about a subject that is mostly a business matter or a healthcare matter? If it's mostly a business matter, then we'd mostly leave them out, because prices aren't encyclopedic except when there is significant discussion (e.g., a price-gouging scandal). If it's mostly a healthcare matter, then we'd mostly include something (as much as we could based upon sources, which might not be much).

Among the other open questions, my notes include:

• What constitutes "discussion" of prices in a source for NOTPRICE? Is it enough to have a source that says Drug Prices Always Matter, and then we add prices to all drugs? Or do we need something saying that the price of a specific drug, or a specific class of drug, matters? Is it enough to have solid numbers (which are hard to come by for many – but not all – drugs), or do we need solid numbers plus an explanation of whether and why the number matters for that particular drug?
• Is it the price, or the affordability? Price matters for business reasons. Affordability matters for medical reasons.
• Do we actually care about anything other than drugs? Does anyone care about the cost of surgeries, hospitalizations, time-consuming therapies, durable medical equipment, or consumable supplies?

Finally, there was a point when it seemed like people were saying that we can't get a good source for all drugs, so we should all give up and not do anything. I don't think we should write this guideline on that basis. We have, in the past, had people tell WPMED folks that they were going to write a review article, and was there anything in particular that we needed a decent source for? It's likely that drug pricing will get more transparent over time, and the ideal sources will emerge. That time might be years from now, but it will likely happen. I think that we should take User:Bluerasberry's advice above and talk about that magical future, and what we consider to be the right way to address that question. A worldwide average? A range? A number for wealthy countries, and another number for low and middle income countries? A description (such as "affordable" or "expensive")? Historical prices? List prices? Something else? Would it be one sentence or one paragraph or one whole section? There are a lot of options, and I think that people ought to be considering the options. WhatamIdoing (talk) 02:14, 15 December 2019 (UTC)

Wrt healthcare vs business matter. Hmm. I think for most drugs, they don't have significant business WP:WEIGHT. Any sales figures are likely to be commercial secret unless they hugely affect share price. None of us dispute that there are drugs that are important for company share price or important for high-cost scandals where there is a business aspect. Again this doesn't justify the original research that's been going on. There is also a political aspect and a retail aspect. I'm really not convinced that a healthcare issue suggests we include a price [there are so many to choose from, only only a few available]. While some drugs have become notable for their healthcare burden, is the price of valproate relevant to healthcare -- that drug is chosen for its effectiveness in treating epilepsy. Another newer on-patent drug might be notable wrt healthcare, if it was for example, not recommended by NICE as a first-line therapy largely because it is more expensive than existing treatments which work much of the time. And we'd have sources that say that information and our reader cares that "expensive -> not-first-line choice" rather than $45.04 median price per month.

I strongly disagree with the magical wishful thinking approach. It doesn't in fact seem that "drug pricing will get more transparent over time". Already there are doubts if MSH price guide is being maintained since 2015 (I really would like an answer to that). The UK seems to be doing more secret patient access pricing for expensive new drugs, leaving behind the BNF to show a mythical wholesale list price that is misleading. Brexit and a US trade deal is likely to lead to the UK being pushed towards US drug pricing. But I think the fact that the sources we have used so far all give prices down at a basic level "A 100-pill bottle of 50mg tablets of enteric-coated sodium valproate" is important. The BNF even gives separate prices per brand for some drugs, the NADAC gives an averaged price, MSH gives prices for each supplier and likely-to-be-misleading buyer prices. Drugs.com gives retail list prices that very much depend on how many pills you order, and that Sandy has explained nobody pays. None of these give us a "therapeutic dose" and none of them give a price per day or month -- doing that is ALWAYS original research. In a regulated drug price market like the UK, I've seen prices per MG relatively constant for different MG size tablets. But in the US, it seems many drugs are the same price per tablet regardless of how many MG, though there are exceptions. The MSH prices are even more variable for pricing which size tablet. Our sources operate at this very low level because the real world is complicated and can't be simplified. Sure, a high-level source could suggest a rule-of-thumb price (like one of the pocket pharmacy sources did with their $$ symbol system, but we abused that source too.

We are surrounded by experienced editors here. If this was simple to do, we'd have done it. -- Colin°^Talk 13:28, 15 December 2019 (UTC)

Appreciate the thoughtful consideration. Worried about work progressing during busy holidays; my IRL calendar is jam packed. SandyGeorgia (Talk) 13:20, 15 December 2019 (UTC)

I am not sure we need an RfC at this time. I see other relatively more simple directions we can develop to make progress.

1. Develop Wikipedia:Experiment to present guidance on how to run and interpret live experiments. We have thousands of drug articles and some subset of them should be experimental. We can get information from including anything new, prices or side effects or any new data, in these. We still do not have standards for running experiments. In the case of prices being yes/no, we need a safe and sane path to do limited experiments with measurable outcomes and to collect feedback on live articles.
2. Develop WP:Infobox for the sake of prices and structured data. Suppose that we had perfect price information on all drugs, and we had consensus to include it in some form. What would our pipeline be and how would we manage it?
3. Develop wiki process of original research for aggregating data to present scales. Right now I have a draft for this at Wikipedia:Defining data. If we included price information, it may not be exact numbers, as what we actually want to communicate is low/medium/high price. I asset that it is absolutely fundamental to knowing about a drug if it is inaccessible by price. I think for any product shown in Wikipedia, communicating whether a product is luxury or affordable is also general reference encyclopedic content. Wikipedia claims to some interpretation, but actually we have a backlog of undocumented examples of original research which we designate as acceptable. It is taboo to talk about this because of respect for the simplicity of WP:OR, but we need an evaluation method for identifying and considering when some original research and primary data interpretation is necessary context for understanding a topic.
4. We need to develop English Wikipedia's relationship with Wikidata. The Wikidata community is light on its editorial policy and ethical standards and right now, I expect that project would draw heavily from the precedents of various language Wikipedias, if Wikipedias wanted to recommend policy. It seems apparent to me that Wikidata is attracting government, nonprofit, and corporate investment at a pace and scale which Wikipedias never have. It does not take too much money to massively disrupt community projects, and if English Wikipedia does not build out the social infrastructure to have discussions about its relationship to general reference data in Wikidata, then the cost of integrating Wikipedia and Wikidata will rise as Wikidata develops with corporate capture. Some imagination about the future of Wikidata and new technology would be helpful here. English Wikipedia could draft the first Wikidata policy on drug prices, which is a privilege which someday companies might pay money to do. English Wikipedia does not have to do this but doing something seems worth considering.

If we had more developed positions on these things then answering the question about what to do with prices on English Wikipedia would be easier.

Blue Rasberry (talk) 20:33, 15 December 2019 (UTC)

Those all sound like interesting projects, but in the meantime, we have to deal with original research in articles. Since there has been no response from the NOR noticeboard, it looks like an RFC is unavoidable. Like Colin, I am not a fan of the idea that these problems will go away with wish lists; we need to deal with problems in articles now, based on current policy, not as the Wikipedia wheels turn on new ideas. SandyGeorgia (Talk) 20:40, 15 December 2019 (UTC)

Break on WAID's summary

Two points:

• The consensus coming out of the ANI discussion means that there must be an RfC. It isn't optional. I see nothing wrong with thinking things over until after the holidays, but by then, we will have to go forward with it.
• I still think that the best approach to the organization of the RfC is to have "proposals" in the form of how prices should or should not be included, and how they should be sourced, and have editors who respond either support or oppose each of the various proposals. But I don't think we need to propose guideline language for adoption. Please remember that editors said at ANI that there needs to be a widely-announced RfC, and the basic issues at hand are whether or not we should include prices, and if so, how they should be presented and formatted. I agree with Guy's comment higher up on the talk page, that this will end up at ArbCom if editors don't come up with a workable RfC. --Tryptofish (talk) 00:34, 16 December 2019 (UTC)
  • Tryptofish, Guy. I have an idea for a question then. There are some here who believe that a drug has a price that can be expressed in dollars and cents. It has been claimed this is "defining data" and thus appropriate for leads, info boxes and wikidata. None of our sources used to-date have "a price" for "a drug". Those who believe in this one-drug-one-price approach have stated that either the raw database sources and OR required to produce a price is just fine (e.g., James), or that we should run an RFC based on a "magical future" that at some point sources will emerge that directly state a price for a drug (e.g., WAID, BR). There are a bunch of proposed questions about leads and info boxes and the old primary/secondary chestnut, but many of them presume that a price for a drug is a sensible and encyclopaedic concept and all we need is an opinion about whether or where to put it. And this is before we even get into thinking about which kind of price we might present (many different wholesale and retail prices, most of which are "list" prices that fail to present a true cost/price anyone or any organisation actually pays). And btw, simplifying a price to an adjective has been suggested, but to date no reliable comprehensive source of such drug-adjectives has been offered, and any source is likely to be very country-specific -- so unless someone finds such sources, that's a non-starter.

So perhaps the question we first need to ask is

"Do you think that a pharmaceutical drug has a one price that can be expressed in dollars and cents, for each region such as US, UK, developing world, and this is information that Wikipedia drug articles should include and can routinely be sourced by editors without original research."

Something like that. This question could be presented with example texts, example sources, and perhaps opening statements by both sides. Perhaps we need to have little *clarification that we are talking about a drug, such as ibuprofen and not a packet of 16 Nurofen Express 200mg Liquid Capsules from Boots the Chemists in London Liverpool St Station. -- Colin°^Talk 13:15, 16 December 2019 (UTC)

Pinging previous participants in this conversation SandyGeorgia Bluerasberry, Signimu, WhatamIdoing, QuackGuru, Doc James, Kashmiri, Levivich. Barkeep49 (talk) 16:11, 16 December 2019 (UTC)

​Speaking for myself that is really two seperate questions and I think RfCs are generally more successful (as in process not outcome) when there is not a compound ask in the initial question. If people feel that's generally on the right track I would simply recommend making it a two question RfC. Best, Barkeep49 (talk) 16:13, 16 December 2019 (UTC)

I like the idea of working on a series of questions that incrementally address the various aspects of the content disputes. Colin's question would be a good start.

Yes, this will make for a problematic RfC, but that seems to be a result of the complexity of the topic itself. --Ronz (talk) 16:17, 16 December 2019 (UTC)

Barkeep49, there are sort of three questions, so I'm not clear which you think is "two". I don't see any of them as independent, so unclear how a two or three question RFC would make sense. If people reject the idea there is a price for a drug, then really it is back to square one for those promoting prices to rethink some alternative approaches to the whole matter: no point then in asking the other two. And if people think it is a valid concept and want it on wiki, then we must absolutely require them to examine carefully whether their wishes can be achieved with the suggested sources and without OR. The policy requirements are clear: it is up to those adding information to Wikipedia to demonstrate that this can be verified by reliable sources and does not require original research to calculate, synthesise or select records from a database. I wish that clarity, in my mind, was a no brainer, but it seems we have to ask -- nobody responded to the noticeboard request to examine the OR. --Colin°^Talk 17:10, 16 December 2019 (UTC)

Of course there are subsequent questions for a later RFC if that is required, such as what should go in the lead, or infobox or body, and which prices we might show. Again the latter is in danger of offering choices we simply don't have. It seems we have different sorts of hard-to-explain "wholesale list price" for each region, never mind the abomination of juxtaposing wholesale and retail list prices. We only seem to have retail prices for the US, and those are "list prices" of dubious worth. So plenty meaty questions could follow on, assuming the first one is accepted. --Colin°^Talk 17:14, 16 December 2019 (UTC)

Colin, in taking another look you're right it's really three questions (expressing in dollars and cents, should include, sourcing). That only amplifies my concern about it - such questions are much harder to get clear community consensus and input on. Best, Barkeep49 (talk) 20:09, 16 December 2019 (UTC)

• I endorse that Colin's basic question or questions must come first. I don't mind that we then go on to ask other questions, even if the answers may be obvious per the first. (I have seen this done in an 'if-then format on other RFCs.) I also don't mind if we do one RFC, starting with Colin's question or questions, and later do another, if that will lend greater clarity.

  I reject pressure to get the wording finalized right away: I am hosting a holiday party this Friday with 200 invitees, and I have done no Christmas shopping. As others advance on the wording, I am following every post, but I cannot initiate work during this stage of December. SandyGeorgia (Talk) 17:25, 16 December 2019 (UTC)

• I'm not convinced this dispute requires any changes to policy pages, or even necessarily a full-blown RfC. On the "pricing v. prices" issue, I'd venture to say that nobody would disagree that Aspirin should say that aspirin is an inexpensive drug; the question is whether Aspirin should say, as it does now, that aspirin costs two- to two-and-a-half cents "per dose" in "the developing world", sourced to a pricing catalog (a primary source). Why not go to an article like Aspirin and just have a talk page thread about whether the price should included or excluded? And then start a thread here about whether a sentence should be added saying "default include" or "default exclude". I wonder if those threads would end up with quick and clear consensus without even needing an RfC tag. If opinion is divided, then an RfC tag can be applied. So I guess those are the "two questions" I see: a specific example, in or out; and a "default" to apply that specific example to all articles, probably on the MOS page (or maybe elsewhere). To the extent this has already been tried in the past, I'd say try it again, given there are more editors paying attention now. – Levivich 19:42, 16 December 2019 (UTC)
  • Again, there must be an RfC, and it's not optional. (But I see nothing wrong with waiting until after the holidays.) I think that Colin's question(s) do/does get at the basic issue at hand. I can imagine that editors who come to the RfC without being familiar with the discussions here would initially react to that question by wondering how they could answer it: do I think that such prices exist in that way? I'm not sure, and that question sounds like a loaded question that is challenging me to conclude that there must be some kind of problem with sourcing, but I'm not sure what that problem is. For that reason, I'm leaning towards an RfC format that presents editors with (1) an assertion that such prices can be included, and specific reasons why, and (2) an assertion that such prices should not be included, and specific reasons why. And then, editors can support one of those and oppose the other. In that way, the reasons for and against will be easy to see, and editors can weigh which argument is the strongest. Also, a proposal or assertion of this type is less restrictive in terms of having a single question or two questions. --Tryptofish (talk) 20:18, 16 December 2019 (UTC)

Sure we could start with "do you think the approximate price of a medication can be estimated for various regions of the world?" And "Should Wikipedia articles contain information about the cost of medications?" Doc James (talk · contribs · email) 23:50, 16 December 2019 (UTC)

I think that the first of those two questions is problematic, because it fails to capture all of the issues that play into the disagreements here. If reworded, the second question does seem more apt, although I don't think that "cost" is the correct word to use. But in any case, the more I think about it, the less I like the idea of asking yes/no questions. First of all, there must be some explanation of why a particular editorial approach is either good or a problem. It's not nearly enough to frame things in terms of "approximate" and "estimated", because one editor's "estimate" might be another editor's "original research". For those reasons and more, I think it would be much better to present responding editors with specific examples of how one might write about drug prices, with explanations, and ask them to support or oppose those proposed approaches. If we don't have specificity, there will be no agreement about how to implement the RfC results. And if there is going to be specificity, there might as well be one example that reflects the position of some of the editors here, another example that reflects a different position held by other editors, and so on – and then ask the community to endorse one or more of those concrete formulations and oppose one or more of the others. --Tryptofish (talk) 21:22, 17 December 2019 (UTC)

Tryptofish, I would like to see one or two such proposals from you. If you write one, others might be able to write similar ones, to show their idea of the best approach. WhatamIdoing (talk) 05:27, 18 December 2019 (UTC)

That's a good suggestion, thanks. I had actually been starting to think the same thing. On the other hand, I think there are several other editors who have been working on these problems a lot longer than I have, and understand the issues a lot better than I do. So I'm thinking that I can create some examples of the formatting, but leave it to others to fill in the details once they can see what I'm talking about. I was traveling all day yesterday and I'll have a very unreliable internet connection through the weekend, but I'll try to get on it soon. --Tryptofish (talk) 16:11, 19 December 2019 (UTC)

James's questions are problematic because a "medication" could be viewed as a packet of pills with a bar code: we need to be clear the scope is the article scope, because that's the scope that has been done several hundred times already. "the wholesale cost is" is article-scope. We claim this entire drug has a price in dollars and cents, and also apparently, has a dose. One. Not many. The "can be estimated" is asking people to speculate and is answered true if you can do it for even one drug. And when people think of "estimate" they think of rounded numbers or even adjectives, not of "$43.77" or even "from $38.34 to $99.55 per dose" which isn't estimating. We mustn't ask for permission for something that is ambiguous or tangential to the problem with the current approach to drug prices. The second question is clearly broken. Clearly some of our articles should (though we still need to avoid OR and incorrect use of sources), and that's permitted by WP:NOTPRICES. Below, I see an argument between two editors where one wants every sentence and sub-clause and fact in the lead to be followed by at least one inline citation number. And the other wants far fewer. The claim is "No guideline prohibits addition citations to leads", which is used to suggest that citation proliferation is permitted. We have the same problem with this second question: it would be taken as permission to add them to every drug, to add them to (often only) the lead, and to do it in a way that misleads our users and breaks OR. I don't think either of those questions are acceptable or will resolve the problem. I think the one I proposed (clarified to avoid the misunderstanding below) is necessary. -- Colin°^Talk 15:20, 18 December 2019 (UTC)

User:Colin, the question that you wrote needs some tweaking, or perhaps should just be replaced by a statement of the facts. Do I think that a pharmaceutical drug has a price that can be expressed in dollars and cents, for regions such as US? Yes. In fact, I think that a drug has multiple prices that can be expressed in dollars and cents in the US: AMP and WAC and retail price so forth. I think it is possible for us to write, without violating WP:NOR, a sentence or two about one or more of these prices for most drugs. It would have to be specific, e.g., "Shortly after its original approval in 2010, Medicaid's US average wholesale price for a 60-mg prefilled syringe of denosumab was reported at US$990 per dose, with two doses expected each year to treat osteoporosis.[[[PMC (identifier)|PMC]] 2957751]". Is that "the only" price? No. But it is "a price", and I do not believe that it violates WP:NOR. WhatamIdoing (talk) 06:01, 18 December 2019 (UTC)

WhatamIdoing, where is a difference between "a" and "multiple" that is key to my question, and I agree we need to be extremely specific: but being specific is not what has been done in hundreds of articles where we have been misleadingly vague, and is impractical for most drugs. The stubbornness that we can simplify basic database record queries down to single prices for the US, UK & developing world, is really why we are here facing an RFC. So perhaps the question needs rephrased and I think the question would also be clearer with examples and/or opening statement that summarises the problem.

Most medicines are available in multiple formulations, strengths, package size and brand. There are a minority of typically new and expensive medicines where their cost is well documented and not really for debate here (other than to consider OR or cherry picking examples, etc). Many medicines are used for multiple indications. The dose and duration of treatment varies with indication and patient characteristics and simple trial-end-error. There are some medicines with one simple dose regime (e.g. contraceptives) but even then there are brand choices and price choices. Look at ethinylestradiol with desogestrel. A pack of 63 20mg/150mg formulation costs £5.04 (Bimizza) or £5.08 (Gedarel) or £8.44 (Mercilon) or £5.07 (Munalea). The 30mg/150mg formulation costs £6.13 (Alenvona) or £3.80 (Cimizt) or £4.19 (Gedarel) or £7.10 (Marvelon) or £4.18 (Munalea). And of course, that cost is only borne by the NHS, because contraceptive prescriptions are free in the UK. Which specific price should our article include?

You tried to pick an example where the price was fairly simple. But according to the BNF, denosumab has several dose regimes. There is "60 mg every 6 months" for osteoporosis. Then there is "120 mg every 4 weeks" for bone metastases. Or there is "120 mg every 4 weeks, give additional dose on days 8 and 15 of the first month of treatment only" for giant cell tumour of bone. Try sticking that in an infobox. There are also associated-costs because the treatment has risks that suggest dental examination and treatment are done prior to taking them, and plasma-calcium concentration monitoring required ongoing. While the 60 mg vial costs £183, the 120mg vial only costs £309.86 (a £56 saving on two vials). Those are list prices that the NHS "pays". The people getting this probably pay nothing in the UK (due to age or cancer exemption). But here's the problem: the drug was only approved on a "patient access scheme" with confidential discount. You can read the proposal here -- look at page 6 and you will find black boxes over the discount offered. So the "list price" is simply a fantasy "recommended retail price" that the NHS has agreed to display as part of negotiations. In the US, according to the official website there are discounts on what anyone pays, all they way to $0 for those who are uninsured and unable.

If you do think that being very specific about our prices is an approach to consider, then I don't think even denosumab has information we could explain in the lead, in an infobox, and for the UK we'd have to admit that we really have no idea what price the NHS pays. See how that approach works for valproate -- the most commonly prescribed anticonvulsant and available as generic worldwide. -- Colin°^Talk 15:05, 18 December 2019 (UTC)

"Prices vary according to dose and other factors, but Medicaid's US average wholesale price for a <size> pill of valproate was reported in <source> as being US$0.xx on <date>". It might be a WP:DUE violation, but it would not be a NOR violation.

But do you know what I might rather do? I might rather say that PMID 29564159 in the BMJ has estimated the production costs for the active pharmaceutical ingredient to be about US$26/kg. This is likely to be fairly stable and to apply wherever the drugs are made.

People who care about the public-policy-and-patients side more than the chemicals-and-manufacturer's side of that drug will note that the same lines of that table give an estimated generic price of 1.5¢ each for a 100mg pill, 1.9¢ for a 200mg pill, and 2.9¢ for a 500mg pill. I don't think that would violate WP:NOR, although we'd probably need to explain exactly what their phrase "estimated generic price" means. WhatamIdoing (talk) 17:50, 18 December 2019 (UTC)

I think we are getting a little off track, and you missed my joke at WT:MED: "Maybe we should list the wholesale price per kilogramme of the pharmaceutical intermediate [of valproate] in 200kg drums at Alibaba". The paper you cite is primary research, and I'm not sure a bunch of medics are qualified to estimate production costs of pharmaceuticals, and if doing this at a 10% profit was good business, it begs the question why it isn't being done. One thing I did find interesting in the paper was

"Most of the high-priced medicines in India were found only in the private market price source, and not the (Tamil Nadu) government tender list, suggesting a lack of availability in public facilities. Over 75% of health expenditure is out-of-pocket in India, of which the majority is spent on medicines. While we found Indian prices to be below our estimated generic price in many cases, Indian prices were mostly government tender prices, which are likely to be significantly lower than the private market prices more often experienced by those needing medicines in India. Further analysis of the Indian market would be necessary to determine prices available to various facilities, provinces and patient groups"

There has been an argument that we are doing our readers a favour by informing them of the price that Big Pharma want to hide, or because patients in some countries pay out of pocket. The irony is, by quoting wholesale list prices or retail list prices, we are in fact publishing the mythical prices that Big Pharma is happy to display, but which neither government, pharmacies nor patients actually pay. Government is usually paying much less than list. In the US the patients are paying much less than list. In the UK they may often be paying nothing at all or a fixed £9. In India they are paying through the nose because of lack of government-supplied medicine availability, and we have no data on what is actually paid in the developing world, assuming the drug is even reliably available. So we are actually doing Big Pharma a favour, and not actually educating our readers of anything meaningful at all. -- Colin°^Talk 19:11, 18 December 2019 (UTC)

• I've made a tweak to the above proposed question to try to address WhatamIdoing's concern that there may be ambiguity about the scope of "a" meaning "one for the whole article (in a region)" or "one of many possible". Open to other ideas about how to word that. More extensive tweaks or suggestions will probably require to be stated again, rather than done in-place. I think sometimes here (and below) the discussion is a bit deviating towards imagining/thinking about various potential prices (e.g. industrial) or other articles being in scope for discussion at this RFC. I don't think we should expand that far right now. We aren't here because of that. We need to remain grounded in the edits that have been made to hundreds of drug articles, and to MEDMOS, and the intractable nature of the discussions about those edits (article/guideline text + sources). That's why we reached ANI and prompted an RFC, so that is the article text + sources that should be considered. We can all imagine other kinds of price-related edits or subjects to discuss another day. And as Sandy warns below, we need to guard the RFC against permitting all sorts of tangents to be created during the RFC. -- Colin°^Talk 08:34, 19 December 2019 (UTC)
  • Colin, if editors actually voted that yes, they did think that there was one single price for common drugs in each area, and that it both could and should be included routinely, what would you do with that (I mean, after the part in which we all groan and wonder whether humanity is doomed)? I am not convinced that this question is going to produce actionable results. WhatamIdoing (talk) 16:33, 23 December 2019 (UTC)
    • WhatamIdoing, what would I do with that? I would have to conclude I was wrong. Other people clearly have different value systems when it comes to Wikipedia. I haven't seen any indication from you or indeed mostly anyone else, that this would be a groan-worthy and humanity-doomed scenario. We've even had the opinion expressed and codified in proposed RFC text below that we might go further and include these "simple" prices in our info boxes, sourced to wikidata. WhatamIdoing, we advertised for comments about whether the source->text broke NOR and not a single person responded. So I'm totally expecting James to be vindicated and if folk think the text in those articles is just dandy, then that's me put in my place. I'm quite used to election results defying rational sense, so it would not be at all unexpected. -- Colin°^Talk 18:44, 23 December 2019 (UTC)

Timing

Does anyone urgently need this RFC to start in the next week or two? I propose that – unless there are significant objections – we plan to start the RFC no sooner than Thursday, January 2nd. Does anyone object to that? WhatamIdoing (talk) 05:30, 18 December 2019 (UTC)

That should work for me. SandyGeorgia (Talk) 14:29, 18 December 2019 (UTC)

Certainly not during the holiday period. -- Colin°^Talk 08:36, 19 December 2019 (UTC)

Yes, perfect for me. I am sorry for falling silent recently. Had a couple of conferences, international travels, etc., over the last weeks which taken together consumed most of my energy. Now let's have a peaceful Christmas break and get back to work on the 2nd. — kashmīrī ^TALK 14:35, 19 December 2019 (UTC)

I, too, agree. I think it should definitely not be any earlier than that, and may even need to be a bit later. --Tryptofish (talk) 16:24, 19 December 2019 (UTC)

"A bit later" is perfectly fine with me, too. WhatamIdoing (talk) 16:35, 23 December 2019 (UTC)

Name the RFC

It does not appear that anyone is taking the lead on this; it was clearly stated on the ANI that Colin should not be the one to formulate the RFC, but no one else has. I have repetitiously suggested ad nauseum above that we need a name and a draft space to start work. Please pick a sub-page so that we can use it, with an Under construction tag until ready for launch. The situation is complex, so how to name it is a problem; the name should not mislead. Let's not obsess on this; the name can be changed as we work (via a page move), but we need a starting place, that can be labeled as "under construction". SandyGeorgia (Talk) 17:23, 17 December 2019 (UTC)

1. RFC on NOPRICE and pharmaceutical drugs
2. RFC on pharmaceutical drug content
3. RFC on pharmaceutical drug pricing and sources
4. RFC on pharmaceutical drug prices and sources

Please pick one, or add suggestions. SandyGeorgia (Talk) 17:27, 17 December 2019 (UTC)

• All of them work for me. I prefer 3, because application of NOPRICE is not at the core of this dispute-- the proper use of sources relative to NOR is. We also have the Ronz notion of prices/pricing. I believe 3 to be a correct formulation of the essence of the dispute. SandyGeorgia (Talk) 17:34, 17 December 2019 (UTC)

  Option 4 works for me as well; in either case, we have to define our terms. SandyGeorgia (Talk)

• Support 3. I agree with you, and I think that 3 is simply the most informative and neutral of those options. And thank you for moving this along. --Tryptofish (talk) 21:07, 17 December 2019 (UTC)
  • Update: Support 4 as first choice and 3 as a close second choice.. Either of those is OK with me, but I think that 4 puts the spotlight on the center of the disagreement that needs to be resolved. --Tryptofish (talk) 16:15, 19 December 2019 (UTC)
• Support 3. The issues are broader than NOPRICE, while certainly more specific than all pharmaceutical drug content. --Ronz (talk) 23:26, 17 December 2019 (UTC)
• Do we only care about pharmaceutical drug prices? Should this RFC cover other types of prices/pricing/costs? WhatamIdoing (talk) 05:26, 18 December 2019 (UTC)
  • I remain very concerned that if we try to do too much in this RFC, we will not get a usable result. The disputed matter right now really boils down to how a few sources are being used, by not very many editors, in very defined ways. I suspect that by addressing that issue first, we will solve any disputes not already addressed by NOPRICE. I really appreciate the work you've done to show that we have similar potential problems with medical devices, etc, but I feel those are bridges we should cross separately, so as not to muddy an already complex issue. SandyGeorgia (Talk) 14:33, 18 December 2019 (UTC)
    • I think any single RFC is going to be difficult. The first question could probably be more generic, however. There's a line of argument that says money matters more in healthcare industry than it does in, say, the tech industry, so NOTPRICES applies to mobile phones but not to drugs. That could be addressed in the general case, inclusive of all medicine-related financial content. NB that I'm not saying that the first question needs to be more generic; if you all would prefer to talk only about drugs, that's okay with me. WhatamIdoing (talk) 17:24, 18 December 2019 (UTC)
    • I'm a bit confused that #3 says "pricing" especially given Ronz distinction of pricing being unarguably important but "Prices may not exist in a form that is suitable for encyclopedia articles at all, as all the discussions are indicating". So I think the question is "drug prices" not "drug pricing". -- Colin°^Talk 19:15, 18 December 2019 (UTC)
      • I thought the preamble of the RFC would define the Ronz concept of price/pricing early on. If we use "prices" in the RFC title, people may automatically go straight to NOPRICE and not engage the distinction. IF we use "pricing", I am hoping they will be more likely to digest the distinction we are making in the discussions here. SandyGeorgia (Talk) 19:46, 18 December 2019 (UTC)
        • We really aren't having an RFC about drug pricing. The last thing wiki need is "Support. Drug pricing is important" votes. -- Colin°^Talk 20:38, 18 December 2019 (UTC)
          • Are you rejecting the third option totally, or just wanting to change "pricing" to "prices"? Either way, we are going to have define the distinction in the preamble to the RFC. SandyGeorgia (Talk) 03:44, 19 December 2019 (UTC)
            • SandyGeorgia the third option, with "prices", is perfect. And agree we need to clarify these things, because repeatedly "drug pricing important => drug prices in articles" is a point made. We need to make it clear that is one of the contentious issues the RFC is clarifying. -- Colin°^Talk 07:54, 19 December 2019 (UTC)
              • @Tryptofish and Ronz:, I have added an option 4 per Colin's suggestion above; please revisit so we can flesh this out. I will change my !vote as needed. SandyGeorgia (Talk) 14:20, 19 December 2019 (UTC)

User:Barkeep49, do you have some examples of RFCs (in the same approximate area of content) you think worked well, and perhaps some you think were bad because of the question(s) asked. -- Colin°^Talk 20:38, 18 December 2019 (UTC)

Colin, good question. Give me a couple days to see what I can dig up. Best, Barkeep49 (talk) 20:53, 18 December 2019 (UTC)

Well, here's an epic fail, for which I am (only partly) responsible, and providing examples of all the things we are trying to avoid here: Wikipedia:WikiProject Medicine/RFC on medical disclaimer. It started out with a few choices, to show how it might be done, and then as the RFC was running, other editors kept tacking on other sample ideas, to the point that we ended up with nothing. Tryptofish, have a look at this in terms of what can happen when you offer samples ... that the participants will grow to their own idea of what can be done. The RFC became so confusing and diluted that it became a nothing-burger. It should have been a yes-no question: does Wikipedia need a medical dislaimer. There's lots more wrong with that RFC, so it should serve as a good example of what to avoid. SandyGeorgia (Talk) 03:44, 19 December 2019 (UTC)

Thanks Sandy. As I noted above, I'm traveling and will have an antique internet connection for the next few days, but I will definitely look into that. I agree that we must ensure that there will be a usable outcome. (Perhaps one solution would be to disallow adding new options after the RfC has opened, or perhaps not. I think there are significant dangers of inconclusive results with almost any RfC format, and it's essential that we figure that out before going live with the RfC, even if it takes a little longer than expected. Better to get it right than to get it fast.) --Tryptofish (talk) 16:21, 19 December 2019 (UTC)

Thanks for the link to that earlier RfC. I've studied it, and I think I need to take a drug now. I fully agree with you that we need to avoid that. I don't claim to have all the answers, but something that stood out to me right away is that editors were presented with a list of choices for the disclaimer, of which to approve or disapprove, and then were presented with views by individual editors about which choices were good or bad, and ultimately had to support or oppose those secondary views. That clearly did not work. Also, there was a lot of background information to work through. I think it might have been better to ask editors to support or oppose each of the disclaimer versions, as opposed to supporting or opposing what other editors thought about those disclaimer versions. Again, I don't have all the answers, but I'm leaning towards giving editors a limited number of specific choices, giving them succinct reasons for and against each of those choices, and asking each editor to support or oppose the various choices directly (as opposed to supporting what somebody else said about it). (At WP:GMORFC, there were a whopping 23 different options, and yet it was quite clear at the end of the month what the community consensus was.) --Tryptofish (talk) 16:33, 22 December 2019 (UTC)

Re, "I think I need to take a drug now", pass the prozac. I think I can afford it (although I notice the a price is mentioned in the lead, with no mention whatsoever in the body of the article, WP:LEAD problem). I was partly responsible for the dismal structure, but did not foresee it would get so out of hand with alternate options tacked on. SandyGeorgia (Talk) 16:52, 22 December 2019 (UTC)

Structure of the RfC

I've created User:Tryptofish/Drug prices RfC draft. Some editors here suggested that I should show what it might look like if we structured the RfC around propositions to support or oppose, as opposed to questions to answer yes/no. So that's what I've tried to do there. I don't think it's anywhere near to being ready for prime time. It's just an example to start some discussion. What I'm trying to do is to (1) pin down what actually might appear on pages (as opposed to generalized descriptions that might mean different things to different people), and (2) get responding editors to say clearly that some options are OK and others are not. --Tryptofish (talk) 16:41, 22 December 2019 (UTC)

Considering my epic fail discussed just above this, I will wait for others to weigh in. SandyGeorgia (Talk) 16:53, 22 December 2019 (UTC)

I really really don't like the 6 propositions options (individually and the idea of having 6). Can't see how that wouldn't be a total disaster. Can we try to be consistently clear about our terminology. "Drug pricing" is not the same as "drug prices". The former is the whole social/economic/policical/health issue of how to set the price of drugs, and the consequences of those decisions or methods. Drug prices are numbers. We really aren't discussing whether articles should mention "drug pricing". The question "a simple listing of a price or price range may be cited to a primary source" is loaded unfairly -- if there is one continuing theme about the discussion over all these pages is that (a) the person putting up the price thought and still thinks it was simple but (b) it turned out to be anything but. I don't think we should ask questions that subvert WP:NOR, but instead ask people to judge whether it can be done while obeying policy. For example, here's what WHO say in their essential medicines review report: "The MSH (Management Sciences for Health) International Medical Products Price Guide 2015 reports a median supplier unit price for tramadol hydrochloride 50-mg tablet/capsule of US$ 0.0427.". Not once has any price in our articles even approached that level of direct reporting of a primary source, and unlike the WHO researchers, we can't arbitrarily pick the 50mg tablet vs the 100 mg/ml solution, 100 mg/ml ampoule, 50 mg/ml ampoule, or the controlled-release capsules in 50, 100, 150, 200, 300, 400mg sizes.

I think we need to stick to a simpler one question design, and forget for now the fanciful ideas about info boxes and external links. We are here because of the seriously problematic prices in hundreds of articles leads, let's stick to discussing and resolving that problem first. Tryptofish, if you think, after all this discussion about the problems, that this will be resolved by someone voting "There's nothing wrong with sourcing a drug price to a primary source, and readers need to know this information" then we really have all been wasting our time and have failed to explain the issue.

I'll repeat my suggestion below:

"Do you think that a pharmaceutical drug has one price that can be expressed in dollars and cents, for each region such as US, UK, developing world, and this is information that Wikipedia drug articles should include and can routinely be sourced by editors without original research."

I think this works because it requires the participant to explain why (or at least, agree with someone else's explanation). And the above should include examples from articles, and the good-idea/problematic sections like your proposal so people get some info up-front. -- Colin°^Talk 09:27, 23 December 2019 (UTC)

I'm fine with using whatever format for the RfC that gets consensus from editors here. But I do want to emphasize that the examples that I gave are not actual proposals for how the RfC would be worded. I'm only trying to show a possible format. And I've said explicitly that the number of questions could be very different than what I put there. And I certainly wasn't trying to illustrate the best possible RfC responses: those were just some rather crude examples and nothing more. So let's please not get bogged down in the way that sentence was written on Tryptofish's draft page is bad, so we have to reject the entire approach shown there.

Since you do appear to be proposing the specific wording for your single question, I will repeat, in turn, my opinion that the question as you wrote it is not really neutrally worded, because it subtly implies that anyone who would answer "yes" must be making a mistake. Also, it's hard for me to envision a way that your single-question format would allow us to show responding editors what it would actually look like on a given page. Let's see if we can combine the better parts of your approach with the better parts of mine.

I do agree with you that the issue contained in your proposed question is the most essential one here. And I'm happy to agree that other stuff (infoboxes, external links, and so forth) can be omitted, assuming that other editors here will agree with that. And it sounds like we both believe that it's useful to provide "for" and "against" reasons.

So, how about this:

Proposition: A pharmaceutical drug has one price that can be expressed in dollars and cents, for each region such as US, UK, developing world, and this is information that Wikipedia drug articles should include and can routinely be sourced by editors without original research.

Supporting arguments:

1. reason
2. reason
3. etc.

Example of specific page content that is written assuming the proposition is true.

Opposing arguments:

1. reason
2. reason
3. etc.

Example of specific page content that is written assuming the proposition is false, and showing how to write it correctly based on that assumption.

References

That would be in effect a single question, and all I did was omit the first few words of your question. But it presents the issue as two alternatives, one of which could be supported and the other opposed. If that works for you, I can happily support it, although I anticipate that other editors will want to discuss the exact wording. --Tryptofish (talk) 15:12, 23 December 2019 (UTC)

Tryptofish, there isn't an "example" of text that assumes the proposition is false. The conclusion would be the current approach to listing prices is wrong and needs removed. If people then want to finesse the cases where prices might appear and what sort of sources to use, that's a question for another day. The "blockage" the RFC is trying to resolve is whether the approach on hundreds of existing articles is acceptable at all, not to merely tweak at the wording. -- Colin°^Talk 19:14, 23 December 2019 (UTC)

The problem with that proposition is that we're asking editors to vote on whether certain verifiable facts (e.g., it's usually more complicated than "one price") are actually facts, which is not an appropriate subject for voting. What if they all vote in favor of the non-factual side? WhatamIdoing (talk) 16:50, 23 December 2019 (UTC)

WhatamIdoing, it may be your opinion that it is usually more complicated than "one price", and I agree with you, but we have hundreds of articles clearly stating one price for "developing countries", one price for US, etc. And despite me complaining about that for a month, virtually nobody has backed me up on this. I've repeatedly pointed out multiple pill sizes and formulations and nobody cared. I've repeatedly pointed out multiple indications and other factors mean there is not one "dose" but nobody seemed concerned. I've repeatedly pointed out that we are doing original research and again nobody seems to mind. I've found mistakes in nearly every price and nobody even shrugged. We advertised for people to comment on whether the current text broke NOR and nobody came. If they vote in favour, then Wiki gets to keep these prices in the lead of hundreds of articles. The "facts" haven't mattered so far. -- Colin°^Talk 19:02, 23 December 2019 (UTC)

What is the reasoning behind adding to articles prices of drugs in developing vs developed nations? Is it meant to imply or inform readers that drug companies are overcharging people in developed nations?--Literaturegeek | T@1k? 20:05, 23 December 2019 (UTC)

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Lead

I have removed some additions to the lead section. MEDMOS needs to stick to dealing with medicine/health article issues and not become some fork of standard guidelines. We already have guidelines on lead sections and on making technical articles accessible, so no need to add more. Given that "people don't read the manual", the shorter and more to-to-point this guideline can be, the better. As an aside, there is more skill involved in making an article accessible and engaging than just replacing words with more basic simple words. -- Colin°^Talk 10:16, 12 November 2019 (UTC)

Disagree. Having these basics here is still important. Doc James (talk · contribs · email) 18:19, 21 November 2019 (UTC)

Colin, do you think the other guidelines on lead sections and making technical articles accessible do a good job of making it clear that those guidelines also apply to medical articles? If so, can you point me toward the guidelines you referred to, please? I would like to have a look. Thanks! I do think that is important. Many analyses of Wikipedia medical articles have found them to have too much jargon, which can be confusing to our readers. I do agree that the lead should be kept fairly simple (target is ~8th grade reading level) and we can go into more depth in the later parts of the article. TylerDurden8823 (talk) 20:13, 21 November 2019 (UTC)

A level of grade 12 is more realistic and what we appear to be managing for the leads.[69] Doc James (talk · contribs · email) 22:35, 21 November 2019 (UTC)

Tyler, general guidelines apply to all articles; they need not explicitly say so.

Doc James, you have cited an off en-wiki document to justify your addition, rather than an on-wiki policy or guideline. SandyGeorgia (Talk) 21:10, 8 December 2019 (UTC)

Doc James, you restored the lead text you wrote prior to attempting to achieve consensus and prior to posting your above "Disagree" comment. You are simply edit warring. As I made clear in my above comment, there is nothing James has written that is specific to medical articles. The same is true of any potentially complex topic on Wikipedia. We do not fork such guidelines. I really don't think James is in any position to lecture others about good writing technique. If it isn't specific and relevant to health/medicine, it does not belong here. If someone is including a "Reader Native Language by Language" chart in MEDMOS, then you can tell they are desperately trying to make point to meet an agenda, rather than stating something that has Wikipedia-wide consensus. Let's leave the advice on writing leads to the whole Wikipedia community, where those who are actually competent writers can craft competent guidelines. -- Colin°^Talk 21:49, 23 November 2019 (UTC)

I support Doc's changes with regard to the lead material. I don't see an issue without spelling out those aspects in this guideline. Flyer22 Reborn (talk) 03:38, 26 November 2019 (UTC)

This is what I refer to as a "me, too" support, which offers no analysis of how this page conforms with/differs from WP:LEAD. In many areas of Wikipedia discussion, the closing admin or coordinator is empowered to ignore reasoning that does not offer a rationale. This text specifically departs from or extends beyond WP:LEAD, the relevant guideline page; it is up to the editors supporting this addition to explain how they believe it interprets the Wikipedia-wide guideline for medical content. If consensus is to be found in medical guideline pages, it behooves us all to discuss rather than !vote with "me, too" and WP:ILIKEIT supports. We can't come to consensus when given nothing to understand upon what !voting is based. SandyGeorgia (Talk) 21:07, 8 December 2019 (UTC)

Another disputed section

Regarding this addition, which is also not based on consensus, we have had this discussion many times, and yet here we are again. O one of the reasons the Medicine Project guidelines were widely accepted years ago is because they did not contradict or extend beyond Wikipedia-wide policy or guideline, rather explain how to interpret policy or general guidelines for medical content.

This addition goes beyond WP:LEAD, and because we have had this conversation many times and in many places, I am concerned about why it was again added as if it had consensus. This is a sample of the broader WPMED disputes mentioned at ANI, and should also be tagged disputed and considered part of the same issue, where we see personal preferences being written into guidelines and being applied to broad swatches of articles (even FAs that comply with Wikipedia-wide policy).

Almost every piece of this non-consensus version of WP:MEDLEAD either extends beyond what WP:LEAD says, or is at variance with what LEAD says, or repeats what LEAD says-- the biggest problem is where it extends beyond what LEAD says and is used to impose a structure on leads that is at variance with what LEAD calls for and what is called for in Featured articles.

• The lead of an article, if not the entire article, should be written as simply as possible without introducing errors or ambiguity. Around a third of readers of English Wikipedia, have English as a second language. Language can often be simplified by using shorter sentences, having one idea per sentence, and using common rather than technical terms. The British National Formulary for example often uses "by mouth" rather than "oral". It is also reasonable to have the lead introduce content in the same order as the body of the text. Avoid cluttering the very beginning of the article with pronunciations or unusual alternative names; infoboxes are useful for storing this data. Most readers access Wikipedia on mobile devices and want swift access to the subject matter without undue scrolling. It is useful to include citations in the lead, but they are not obligatory. Two reasons for using them are:

  Medical statements are much more likely than the average statement to be challenged, thus making citation mandatory.

  To facilitate broad coverage of our medical content in other languages, the translation task force often translates only the lead, which then requires citations.

As can be seen by the concluding sentence, these extensions to Wikipedia-wide guideline are being imposed apparently to facilitate a different project, that is, translation of leads only to other languages, which has been a focus of WPMED for several years now, as opposed to focusing on having English-language articles comply with English-language policies and guidelines.

As in article editing, the burden to explain an addition should be on the editor adding the addition, so rather than have me go line-by-line to explain why all of this text is disputed, I believe it would be helpful for the editor(s) who want to add this text to go line-by-line and explain why they believe this text is supported by broader, Wikipedia-wide guidelines. It is not, and the application of this personal preference has caused FAs to be out of compliance with English-language standards only to make translation easier.

This text is disputed, and is part of the same problem discussed at the pricing ANI. A disputed tag should be added. SandyGeorgia (Talk) 20:12, 8 December 2019 (UTC)

I have struck portions that may be construed as "rehashing of past grievances", although my intent was to provide history. I can see that this text could be problematic. SandyGeorgia (Talk) 01:58, 9 December 2019 (UTC)

There is a fair bit of guidance recommending we use easier to understand language such as "Make your article accessible and understandable for as many readers as possible." and Wikipedia:Make technical articles understandable Doc James (talk · contribs · email) 03:01, 9 December 2019 (UTC)

Thanks, Doc, but there is nothing in those pages that is at odds with WP:LEAD, nothing that justifies the proposed additions to MEDLEAD, and even some portions of those guideline and supplement pages that the proposed addition to MEDLEAD is at odds with (eg, a lot of what is in medical articles now is oversimplified because of this trend). I have done the work of reading a page that doesn't respond to my concerns, and I don't see that you have answered my query. Could you please justify the proposed additions, point-by-point, so we can all understand why you think this proposed text complies with the broader and widely accepted WP:LEAD guideline? SandyGeorgia (Talk) 03:41, 9 December 2019 (UTC)

The reading level of our leads has improved, in my opinion, from a reading level close to grade 16 to just over grade 12.[70] I would call that neither over nor necessarily under simplified.

Here is the discussion in 2015 when the main sentence in question was added.[71]

IMO "The lead of an article, if not the entire article, should be written as simply as possible without introducing errors or ambiguity." is another way of saying

"The content in articles in Wikipedia should be written as far as possible for the widest possible general audience."

Doc James (talk · contribs · email) 22:13, 9 December 2019 (UTC)

Doc James, it would help if you would engage the entire issue, as we have discussed many times over the years. Citations are being added unnecessarily to the leads of every medical article, including FAs. No guideline requires this, and it is important that FAs conform with WP:LEAD, as from the lead the mainpage blurb is written. Overciting a lead can prevent FAs from providing a compelling summary. Similarly, a specific structure is being imposed on leads, which presents the same problems. Well-written leads that pass FAC are being altered in ways that are not compelling to read only so that a specific structure can be imposed, based on no Wikipedia-wide guideline, and taking FAs out of compliance with WP:WIAFA. And language is being dumbed-down in many cases to the point of lost clarity. None of this has consensus, none is based on guideline, all jeopardized FA status, and all is being done not for en.wikipedia, but for the an off-en.wiki translation project. If this were being done only to B- or C-class articles, I would be willing to be silent, but adding a non-consensus issue to a guideline, that takes this guideline out of compliance with general guidelines, while altering FAs to comply with a non-consensus guideline is a problem. SandyGeorgia (Talk) 03:29, 10 December 2019 (UTC)

No guideline prohibits addition citations to leads. It makes them much easier to maintain / verify. Doc James (talk · contribs · email) 03:42, 15 December 2019 (UTC)

There are various guidelines on citations in leads. But, as with all of MEDMOS's disputed section on leads, there is no evidence that there is a uniquely medical aspect to the content or citation guidance, vs personal opinion more appropriate for an essay. Firstly, "Citations are often omitted from the lead section of an article, insofar as the lead summarizes information for which sources are given later in the article, although quotations and controversial statements, particularly if about living persons, should be supported by citations even in the lead" emphasises that the norm is to omit citations from the lead, where the text summarises the article. This is something that editors who write article body content, and then summarise that content in the lead, will find more natural, than editors who add factoids mostly in the lead. Secondly, "The necessity for citations in a lead should be determined on a case-by-case basis by editorial consensus". Can anyone please give examples where editors working on an article reached a consensus wrt the need for citations in the lead for a given sentence or claim? I can find reverts and edit wars, but have been unable to find examples of collaborative editing or editors respecting each other and working towards consensus. This suggests to me, that perhaps a dogmatic approach at odds with general policy and guideline has taken dominance. Further reason that we should not have a MEDMOS fork of community guidelines. I would, of course, be overjoyed to read of examples of an article-consensus approach to lead citations. -- Colin°^Talk 15:29, 15 December 2019 (UTC)

Doc James, I am unsure why you are entering a comment that "No guideline prohibits addition citations to leads" on 15 December as if we had not already discussed this at 20:38 9 December and 21:47 at 11 December at the Schizophrenia talk page. No one has said a guideline prohibits adding citations to leads; the problem is what a preference for this style (not supported by WP:RS or WP:V) is doing to prose in leads. We have had this discussion already, so I am confused why you keep raising the same point, without addressing the other points. SandyGeorgia (Talk) 00:34, 16 December 2019 (UTC)

CFCF, I have reverted your addition of a separate section heading here on talk, which separated the section I started from the section it was directly responding to, and mischaracterized the nature of the dispute.

As you are a very involved party in these discussions, I request that you refrain from closing discussions, archiving discussions, or altering other editors' posts. I also ask you to please take greater care to read the case being discussed. The disputed text involved much more than what you call language. Since we have a neutral admin following the page now, if you feel it necessary to alter, close or archive something, you might find it useful to query Barkeep49 first.

Barkeep49, this section (LEAD) of the guideline is also disputed, and is part and parcel of everything discussed at the ANI (an issue that keeps being added in spite of no consensus). Because the article is protected, I am unable to add a disputed tag. What would it take to make that happen? SandyGeorgia (Talk) 16:05, 9 December 2019 (UTC)

Also, Barkeep49, I should explain the significance: I rarely edit drug articles, so while the pricing issue is quite important, it does not affect my daily editing. But these extensions of LEAD are affecting/have affected every single FA in the medical suite, so directly impact the articles I edit and WP:MED's top content. It has been hard to keep FAs up to standard when their leads are being edited in non-compliant ways, so IMO it is important that this section be tagged as disputed, not a consensus version.

Here is a sample from this week; it is the first FA I checked, the only one I have checked so far, and it is concerning that the first FA I checked after a not-so-lengthy absence from medical editing shows the very issues of concern (leads edited only, so that the body of the article is out of sync with the lead, and language in the lead oversimplified to the point of losing clarity, with the structure of leads altered in ways that do not lend the prose quality required of FAs). This kind of editing takes FAs out of compliance with WP:WIAFA, and valuable editor time (eg Casliber) is then needed for repair to avoid a WP:FAR. SandyGeorgia (Talk) 16:30, 9 December 2019 (UTC)

SandyGeorgia, I didn't receive your ping and this page has so much discussion that I had not noticed it until now. What is it that you're asking me to do? What I am reading is an argument over content (in this case the composition of the MOS) and would be inappropriate for me to weigh in without becoming INVOLVED. Best, Barkeep49 (talk) 02:45, 10 December 2019 (UTC)

Sorry, Barkeep49, I am not sure about the pinging problem. I will also, then, separately post this to your talk page to make sure you see it.

What I am asking is, considering the page is protected, how can we have an {{disputedtag}} added to the WP:MEDLEAD section? I was also pointing out, ala full disclosure, that although this is a separate dispute from the pricing issue, it is also related, as this is another of the ongoing disputes that was mentioned in the ANI you closed. Would it be appropriate for me to add an {{editrequest}} to ask another admin to add the disputed tag, or are you able to do it as part of the overall issue? We have a protected page because of the pricing edit warring, but there is a separate but related dispute in another section, which should be tagged as that section does not have consensus; it is an ongoing smaller matter that has been obscured by the larger pricing dispute. SandyGeorgia (Talk) 03:06, 10 December 2019 (UTC)

SandyGeorgia, ah now I got you (and I did get this ping - dunno what happened with the last one. Did you have to fix the ping or signature?). Let me look into this a little before responding on the substance of what you wrote now that I understand. Best, Barkeep49 (talk) 03:10, 10 December 2019 (UTC)

Barkeep49, When it comes to pinging, I am old school and never know what works or what I do right or wrong. Thanks for having a look, no hurry; the issue with leads has been unresolved for a very long time, and a day or a week changes little. Regards, SandyGeorgia (Talk) 03:16, 10 December 2019 (UTC)

SandyGeorgia, ok I think I have a grasp on this issue having looked through the edit history. The language of the lead has indeed been an ongoing disagreement for a longtime. As for moving forward I also don't think it's helpful for anyone if I make all sysop related decisions because I am uninvolved and am currently watching this page. So what I would ask is that you go ahead and make a formal edit request and that will hopefully draw the attention of someone else to decide what state to leave the LEAD in while this is protected. You, James, Colin and others are of course welcome to continue discussion and if consensus can be reached that could of course be implemented at any time. Best, Barkeep49 (talk) 03:36, 10 December 2019 (UTC)

Barkeep49, thanks so much, and seems like a wise course of action. I will compose the editrequest tomorrow-- past my bedtime and there may be further feedback when I check in tomorrow. Thanks again, SandyGeorgia (Talk) 03:41, 10 December 2019 (UTC)

Fully protected edit request for MEDLEAD

See discussion above, beginning at LEAD.

The WP:MEDLEAD section of this guideline page has been constantly disputed for several years, as can be seen in the page history. Multiple past local discussions have been used to claim local consensus, which is not apparent, (sample 1, and sample 2), or not enough to override broader Wikipedia-wide policy or guideline.

The version of MEDLEAD that has been alternately removed and re-instated for several years here has a particular impact on Featured articles; FAs must conform with WP:LEAD because the lead is used to write the mainpage blurb. MEDLEAD is at variance with LEAD in ways that have an extra impact of the project's top content, as medical articles must now answer to two different lead guidelines. A sample from this week only (but repeated across many other FAs) can be seen at FA Schizophrenia.

The wider community should be involved in a WP:CENT RFC when a WikiProject guideline is out of sync with Wikipedia-wide policy or guideline, and local consensus has not resolved the problem. The specific issues are:

1. "Language can often be simplified by using shorter sentences, having one idea per sentence, and using common rather than technical terms." No such restriction is in LEAD, and this leads to short, choppy sentences in leads that are not up to FA standards. It has also led to a loss of clarity and the precision required in medicine (see Schizophrenia example).
2. "It is also reasonable to have the lead introduce content in the same order as the body of the text." This is not true for every article, and forcing the lead to a specific flow causes prose deterioration in articles (particularly Featured articles with carefully written leads) where the flow of information may need to be presented differently than the set structure that has been imposed. There is no such requirement at LEAD.
3. "Avoid cluttering the very beginning of the article with pronunciations or unusual alternative names". This is distinctly at odds with the wider guideline, LEAD.
4. "Medical statements are much more likely than the average statement to be challenged, thus making citation mandatory." At odds with LEAD, and based upon an unproven (and often inaccurate) assumption. Again, see example at Schizophrenia.
5. The final sentence now in MEDLEAD indicates why these deviations are being introduced: "To facilitate broad coverage of our medical content in other languages, the translation task force often translates only the lead, which then requires citations." So, to facilitate a non-en.wiki project, restrictions in medical articles that go beyond WP:V and LEAD are being added to MEDMOS.

This page is fully protected because of a separate, but related, dispute. An independent admin, not involved in adminning the separate dispute, is requested to add

{{Disputed tag|section=yes|talk=Lead}}

to the WP:MEDLEAD section (using the "section =yes" option), as it will be some time before the separate dispute can be resolved via RFC and the page unprotected. It is likely that a community-wide RFC will also be needed to resolve this conflict. SandyGeorgia (Talk) 19:54, 11 December 2019 (UTC)

PS, Barkeep49, who is adminning the pricing discussion, agreed that I should submit a separate edit request, see the section just above this one. SandyGeorgia (Talk) 19:57, 11 December 2019 (UTC)

SandyGeorgia, it would be helpful if you could make clear exactly what you would like the section to read as (maybe throw it in a collapsed section if it's long?) Best, Barkeep49 (talk) 20:33, 11 December 2019 (UTC)

Thanks, Barkeep49, but I was hoping to avoid asking an admin to revert, reinstate or otherwise get involved to help resolve the dispute-- just to tag the guideline to indicate that there is a dispute. If we were to revert to the last, undisputed wording, it would be to one where there was no section on LEAD at all; there is nothing in our current MEDLEAD that is not disputed and is in sync with LEAD. The table that is presented is related to the translation project, so not part of this guideline per se, and every addition there is disputed. The dynamic that led to the wording being retained was discussed at length in the ANI, which is why the disputes are related. SandyGeorgia (Talk) 20:48, 11 December 2019 (UTC)

SandyGeorgia, it is beyond doubt that this is disputed so I have added the tag. Guy (help!) 00:29, 12 December 2019 (UTC)

JzG, the over-worked admin corps appreciates your response, as do I. I added the "answered=yes" parameter to the template so another admin won't need to come by. SandyGeorgia (Talk) 00:55, 12 December 2019 (UTC)

SandyGeorgia, sure, I left it unanswered in case there was dispute but I don't think there is. Guy (help!) 09:36, 12 December 2019 (UTC)

Comment As Sandy indicated I thought it helpful for someone beyond me to respond to this request. I have looked into the matter and am happy to answer any questions you have or to serve as a second opinion as I remain UNINVOLVED. For reference I believe the key talk page discussion is here. In looking at the project page history, the dispute goes back to March of 2018 from best I can tell. Best, Barkeep49 (talk) 20:36, 11 December 2019 (UTC)

Barkeep49, the dispute actually dates to 2015, when the order of sections in all medical articles (even FAs) and their leads were altered. But ... I can't make WikiBlame (or the alternate) work today, so I can't find when the additions occurred ... I only know the LEAD section was first added in 2015, and that year corresponds with the end of FA production from a once very vibrant medical FA-writing community, because it became impossible to obey two masters (two different guidelines). Again, as mentioned at the ANI, it is an entrenched dynamic, that will probably require another community-wide RFC. SandyGeorgia (Talk) 20:54, 11 December 2019 (UTC)

I think that the specific edit that is being requested here is to add:

{{Disputed tag|section=yes|talk=Lead}}

just under the nutshell near the top of the page. Sandy, please correct me if I'm wrong.

For what it's worth, I think it's a reasonable request. --Tryptofish (talk) 21:31, 11 December 2019 (UTC)

Thanks, I fixed the template above (but not under the nutshell at the top of the page ... I am asking that it be added to the top of WP:MEDLEAD to single out that specific issue. I don't think we need to indicate the entire page is disputed, as we don't have editors adding prices now and that is under a general restriction until resolved. SandyGeorgia (Talk) 21:54, 11 December 2019 (UTC)

OK, with that clarification, the requested edit is as shown, but does not go where I said, instead going at the top of WP:MEDLEAD. (I confused one lead with another!) --Tryptofish (talk) 23:50, 11 December 2019 (UTC)

Further discussion

• Oppose change — I don't understand how the above has been construed as an edit request, and frankly find it bizarre the section was tagged disputed.
  I oppose points 1—5 above on the grounds that neither is anything in odds with WP:LEAD, and that which is stated as an "unproven and often inaccurate assumption" — is in fact reasonable and well-known. The example given is unconvincing, and the insistance on FA as the pinacle of our work is misleading. Having a corpus of 100 decent or good articles trumps 1 FA. Carl Fredrik ^talk 22:26, 15 December 2019 (UTC)

• You oppose that the section is disputed, yet your post indicates that there is a dispute?

  I understand that you could be confused about how a disputed tag is used, because the norm on this page/project has not been to come to consensus on disagreements, but this is the usual process for addressing a disputed guideline. It ended up at editrequest because I couldn't add the tag myself due to the protection, but there is no doubt there is a dispute; adding a tag is uncontroversial.

  The usual procedure is to discuss with each other to develop a consensus, while this page tends to devolve to "me, too" or "I don't like it" discussion, and then claim a consensus. I am willing to go forward with a community-wide RFC if we are unable to come to local consensus. I am intentionally not putting up the RFC (so far) while we are formulating a different RFC.

  I understand that some editors have said they see no problem with choppy prose and poorly worded leads in medical articles; many others have expressed that the prose damage is a problem, and that the deviation from Wikipedia-wide guideline (LEAD) is also a problem. We can let the community decide: I would be surprised if the community endorsed a WikiProject Guideline that deviates from Wikipedia's broader guideline. But I would much rather that WPMED not have to bring two issues from one page to the community in a short timeframe. SandyGeorgia (Talk) 00:21, 16 December 2019 (UTC)

I contest that what has been written here justifies the tag, yes. I do not contest that someone may find it disputed. The text as it stands is long-standing and there is nothing to indicate that an interpretation of "general consensus" as opposed to "local consensus" justifies calling for a general dispute.

The norm is not to tag with disputed: that which has been discussed by hundreds of editors and is long-standing — when one or two editors (even policy-knowledgeable editors) calls it into question. Carl Fredrik ^talk 06:54, 16 December 2019 (UTC)

@CFCF: There is nothing "long-standing" here. The disputed text is not even 4 weeks old.[72] It has also been added without establishing consensus first. Please read with more attention, including diffs. — kashmīrī ^TALK 09:33, 16 December 2019 (UTC)

Actually part of that text was added in 2015 and was based on this talk page discussion.[73] Doc James (talk · contribs · email) 23:55, 16 December 2019 (UTC)

Another discussion between five editors on a local guideline, discussing text that is at odds with a Wikipedia-wide guideline, and where one of the five dissented. Four editors are insufficient to install something that is not in accordance with a wider guideline. SandyGeorgia (Talk) 18:00, 17 December 2019 (UTC)

Regarding that which has been discussed by hundreds of editors and is long-standing, CFCF, could you produce an example of hundreds of editors supporting these accumulated additions to MEDLEAD? Alternately, could you produce a dozen? A Wikiproject guideline can not extend beyond a Wikipedia-wide guideline. Local "Me, too" and "I like it" supports are unhelpful in any case, but particularly insufficient to trump a Wikipedia-wide guideline page.

If you could please engage the five specific issues, we might be able to come to consensus without a centralized RFC to consult the broader community. SandyGeorgia (Talk) 18:00, 17 December 2019 (UTC)

This is not one of the many Wikipedia:WikiProject advice pages. MEDMOS is part of the site-wide MOS, and it is required to remain consistent with the rest of the MOS. It is also, as a part of the site-wide MOS, open to changes by people outside any WikiProject group. If WPMED wants to write its own advice pages, it can do so in the group's equivalent of a userspace essay. This particular page, however, belongs to the entire community. WhatamIdoing (talk) 06:11, 18 December 2019 (UTC)

RFC on MEDLEAD

• Wikipedia:Manual of Style/Medicine-related articles/RFC on lead guideline for medicine-related articles

  Limited engagement, stalled discussion, broader RFC launched. SandyGeorgia (Talk) 22:27, 22 December 2019 (UTC)

DMOZ/CURLIE

My post above in the lengthy pricing discussion got no response, so separating out here.

MEDMOS (for about a decade) recommended DMOZ specifically as an external link, but that text was removed in 2018 because DMOZ no longer existed. The new {{Curlie}} template, which replaced DMOZ, was never added back in. (Sample [74])

Because this page is fully protected, unless anyone objects, I will submit an edit request to reinstate our long-standing text, but corrected to CURLIE from DMOZ. SandyGeorgia (Talk) 19:04, 11 December 2019 (UTC)

I have no concerns with it being returned. Doc James (talk · contribs · email) 19:36, 11 December 2019 (UTC)

SandyGeorgia, overall I have a big "don't care" about this, but much of DMOZ was undermaintained before it officially closed. Are we sure that the transferred version is actually active enough to be worth recommending? WhatamIdoing (talk) 06:44, 12 December 2019 (UTC)

I suspect that most of us (active medical editors) don't much care one way or the other, but the links greatly simplified our editing. When novice editors add external links to support groups, we can easily point them to the guideline, and to the EL page, and suggest they add the link to Curlie instead. It saves a lot of editing time to provide the kind of information some readers are seeking, and some novice editors add, in one external link. I will wait to submit an editrequest until we have agreement on several items. SandyGeorgia (Talk) 15:09, 13 December 2019 (UTC)

My impression is that there just aren't that many new editors trying to add ==External links== to medical articles these days. I still follow WP:ELN, and I don't think we've had a question about DMOZ/Curlie there for multiple years. WhatamIdoing (talk) 19:09, 14 December 2019 (UTC)

@Doc James, Colin, and WhatamIdoing: please let me know if I should submit edit request (2) as below. We need to get some stuff cleared off of this 800KB talk page. SandyGeorgia (Talk) 14:10, 6 January 2020 (UTC)

My comment above is clear. I am happy for Curlie to be used rather than extensive ELs to charities here etc. Doc James (talk · contribs · email) 14:24, 6 January 2020 (UTC)

Proposal for edit request (2)

Please let me know if there is any disagreement, so we can submit the editrequest and get this section dealt with. SandyGeorgia (Talk) 15:21, 28 December 2019 (UTC)

In the External links section:

• Merge the existing sentence (Normally, however, it is better to link to an external web page that lists such charities, rather than try to provide such a list ourselves.) in to the previous paragraph (If the disease is very rare ... such as a detailed article on the specific topic.)

and re-word it to:

• strike the word very before rare (redundant), and add:
• ... such as a detailed article on the specific topic. It is usually better to link to an external web page that lists such charities, rather than try to provide such a list ourselves. The {{Curlie}} template links to a directory based on the Open Directory Project that contains many such links. For example, on the Tourette syndrome page:

  * {{Curlie|Health/Conditions_and_Diseases/Neurological_Disorders/Tourette_Syndrome/Organizations}}

gives:

• Template:Curlie

Archiving

WhatamIdoing I'm not sure it's a good time to speed up the archiving bot: I haven't submitted the edit requests yet to deal with this section and the next, and the page is protected. How about leaving the archiving time as before, but manually archiving any sections already addressed?

While I'm here, I view this DMOZ/Curlie thing as something that may not help, but doesn't hurt. There doesn't seem to be opposition if we re-instate it. Shall I go ahead and do the edit requests? SandyGeorgia (Talk) 19:28, 27 December 2019 (UTC)

The page is over 500KiB before processing, which is beyond what some people will be able to edit, or even read. This section wouldn't have been affected for another week (even before today's comments). WhatamIdoing (talk) 19:41, 27 December 2019 (UTC)

WhatamIdoing, how about if we do this instead? Leave the archiving bot and numbers as typical, but start a separately named archive for all of the RFC stuff? That's what I've seen done in other cases ... keep all of the RFC stuff in one separate archive, and then we can do that manually, and put a hatnote to it on the top of the new RFC sections. SandyGeorgia (Talk) 19:51, 27 December 2019 (UTC)

That is, leave the bot currently archiving at number 10 with 60 days, but move all of the RFC stuff to a separate Wikipedia talk:Manual of Style/Medicine-related articles/Archive 10b, which can then be a hatnote at the top of the new RFC section, to remind us to archive everything manually together. SandyGeorgia (Talk) 19:54, 27 December 2019 (UTC)

Whatever else, I would prefer not to archive anything that might feed into how we formulate the RfC, until we have the RfC finished. --Tryptofish (talk) 23:29, 27 December 2019 (UTC)

Should I submit the two editrequests now, to be done with these two sections? SandyGeorgia (Talk) 23:32, 27 December 2019 (UTC)

Treatment/management

What happened to Management as an alternate for Treatment in Wikipedia:Manual of Style/Medicine-related articles#Diseases or disorders or syndromes for those conditions where no treatment (in the conventional sense) is needed? It was long an alternate here and is now gone. WikiBlame is not working, so I can't tell why that occurred. SandyGeorgia (Talk) 01:33, 12 December 2019 (UTC)

It's still being used in articles. It has the particular virtue of not implying "permanent cure" for incurable diseases and chronic symptoms.

It looks like Doc James removed it in March 2017 because (according to the edit summary) some students typed ==Treatment or Management== as their section headings. WhatamIdoing (talk) 06:48, 12 December 2019 (UTC)

Yup lots of students added "==Treatment or Management==" rather than just picking one. This has decreased since that change. I do not care which is used. Doc James (talk · contribs · email) 04:28, 13 December 2019 (UTC)

I'm not sure it made sense to delete something useful from a guideline because students are misusing it; rather, the heading could have been clarified, using the same format that is used for other sections that have multiple possible names. I will wait til we have sufficient feedback on several items to submit an edit request. SandyGeorgia (Talk) 15:06, 13 December 2019 (UTC)

Agree that mentioning management would be nice, else articles like NAFLD may seem like using a wrong layout. Signimu (talk) 19:47, 13 December 2019 (UTC)

Damn, I always remembered I could use either. "Management" should come back IMHO, it i very useful for diseases for which we write that no treatment is known. — kashmīrī ^TALK 21:09, 13 December 2019 (UTC)

Perhaps "Treatment (or Management, especially for chronic conditions):" would be less confusing. I wonder if they made the same mistake with other section headings (like ==Prevention or Screening==). WhatamIdoing (talk) 19:05, 14 December 2019 (UTC)

Edit request pending, so archiving bot will leave this section. SandyGeorgia (Talk) 19:29, 27 December 2019 (UTC)

@WhatamIdoing, Doc James, Signimu, and Kashmiri: please respond to Xaosflux below so we can finish up this section; this talk page is sprawling. SandyGeorgia (Talk) 13:11, 4 January 2020 (UTC)

Proposal for editrequest (1)

Please let me know if there is any disagreement so I can submit the editrequest and we can get this section dealt with. SandyGeorgia (Talk) 14:58, 28 December 2019 (UTC)

• In "Content section", change Treatment: to Treatment (or Management, especially for chronic conditions):

 Administrator note Once a decision has been made, please reactivate the edit request if the page is still protected. — xaosflux ^Talk 15:38, 2 January 2020 (UTC)

Xaosflux the last comment on this aspect of the page was made on 14 December, and no one was in disagreement. I waited two weeks before adding the editrequest. Now the talk page is so large, we want to get a few of these things moved off the page. SandyGeorgia (Talk) 12:22, 4 January 2020 (UTC)

Reactivated. — xaosflux ^Talk 12:34, 4 January 2020 (UTC)

 Done — Martin (MSGJ · talk) 14:05, 6 January 2020 (UTC)

Thank you ever so much, MSGJ-- one thing we can now get off of this 800KB talk page! SandyGeorgia (Talk) 14:08, 6 January 2020 (UTC)

RfC Formulation (Clean Start)

Introduction

Barkeep's Background

A few days ago, Colin quite reasonably asked me for some examples of good RfCs to use as a model. I spent some time today looking into past MOS (and related) type RfCs and also asked a few people with experience closing Wikipedia related RfCs for examples. I am reluctant to share any because I don't know that they will actually help move the conversation forward. What seems to happen is that someone puts forth an idea and it gets criticized for either structure or wording. Much/all of the criticism is fair (and I've done it) but ultimately it means we aren't getting any closer to an RfC. Above Colin expressed frustration with the status quo of the information being included in hundreds of articles at the moment. A few others seem to be feeling this as well. Others are no doubt being frustrated at the inability to add the information in new places. The way past this is to get this RfC completed. Editors of all stripes need to have some faith in our process. The RfC question itself is not the right place to put forward compelling arguments about why pricing is/isn't appropriate. Instead that can happen during the RfC or in some sort of supporting material. If the goal is for the RfC is to solve every pricing related question the whole RfC is going to fail and nothing will be decided. Instead the decision should be made, by the people here who care most, about what's most important. To everyone I remind it is possible that not everything that's important will get decided by this RfC. So what's most important to decide? Some willingness to to accept that important issues won't be decided right away needs to be tolerated in order for there to be any chance of moving this forward and it seems clear that this needs to move forward. If a person's answer to "what weaknesses/compromises can I live with?" is nothing then that person is not going to be able to successfully participate in formulating this RfC. The good news is that no one seems to yet be at that point. The further good news is that multiple people are seeing progress being made. In rereading this talk page it seems like there is some level of agreement behind a single question (that can be answered with a support or an oppose). People seem OK with coming back to the details later. This is helpful because a single question also seems most likely to achieve consensus from the community. If we can't get consensus around a single question I think it could also give us insight into what the multiple questions have to be at this point and we can, if necessary (though I think it might not be) examine different formats for multiple questions. As such I am going to propose the following.

I am suggesting we focus, for now, on trying to perfect a single question about pricing (or whatever your preferred term is). I am also going to ask that people to follow my lead and collapse extended content, leaving only the most important information visible. Say as much as you want, but let's make it easy for people to navigate. Thoughts? Barkeep49 (talk) 21:22, 23 December 2019 (UTC)

Possible Questions

So far the following single questions have been proposed:

A

Do you think that a pharmaceutical drug has one price that can be expressed in dollars and cents, for each region such as US, UK, developing world, and this is information that Wikipedia drug articles should include and can routinely be sourced by editors without original research.
— User:Colin

B

Should Wikipedia articles contain information about the cost of medications?
— User:Doc James

Please take one (or both) of these and wordsmith them to your heart's content. And if you want to explain, at length, why your version is good, or issues you see with someone else's proposal, feel free but again please consider collapsing those comments. Barkeep49 (talk) 21:22, 23 December 2019 (UTC)

Question B does not resolve the issue with the existing content on hundreds of drug articles. We must avoid asking questions that are true occasionally or even quite often, but which then permit something universally. Prices have been added routinely and in a manner that claims a drug has one price and using sources that require original research to present the article text added. This is the practice that either wiki accepts or rejects. -- Colin°^Talk 21:39, 23 December 2019 (UTC)

Colin, do you have a tweak to that formulation? If not, it's hardly surprising that you would prefer Option A considering you were the one who originally crafted it :). Best, Barkeep49 (talk) 21:56, 23 December 2019 (UTC)

Nope. It is a bit like "Should we offer good healthcare to citizens" which, depending on ones politics, could be interpreted to allow anything from private health insurance to socialist state health. And perhaps there should be a bit more input from James. -- Colin°^Talk 22:02, 23 December 2019 (UTC)

• I don't like either A or B (for reasons that I already stated earlier). I'm not going to attempt any further to suggest what a good question would be, but I think that there are basically two "ideas" that need to be incorporated into the single question, and in a manner that community responses will lead to an unambiguous consensus:
  1. Whether drug prices should be widely presented on drug pages, and
  2. What kind of sourcing is needed to support such content.
• I also think that responding editors must be able to see specific examples of what the various options for doing this would look like. (If editors here cannot come up with a specific example of their preferred approach, then that approach does not merit inclusion.) I also think that (obviously) the wording must be absolutely neutral.
  I look forward to seeing specific proposals from other editors that will meet those criteria. --Tryptofish (talk) 22:19, 23 December 2019 (UTC)

I think Tryptofish is closer to the mark than the two formally stated questions. · · · Peter Southwood ^(talk): 17:21, 24 December 2019 (UTC)

Both of the formally stated questions do not have simple answers.
My answer to A would be No, but sometimes it could happen, and this is not the only situation in which a price might be of interest, and to B it would be Yes, but not always, and only when it is of encyclopedic value.· · · Peter Southwood ^(talk): 09:06, 25 December 2019 (UTC)

SG's Background

I fully reject B as a non-starter, because we already have policies that answer that question, and to ask it will be an insult and generate a d'oh, unuseful response. Also, it completely glosses over the heart of this dispute, which is NOR and sourcing, and glosses it over to such an extent as to be a misleading way to position an RFC. I partially support A, but it is (as attorneys say) "multiple in form". I would split it. On C, it does not have to be US$, and I tweaked the punctuation, etc. a tiny bit. On D, I took out "should include" because, again, we already have a policy on that and I don't think any of us are arguing that WP:NOPRICE is the problem here, or trying to change NOPRICE. I think we should include our price/pricing distinction in the preamble. And I think if we frame it like this, the responders can give samples in their responses. I believe we do need an example, but no good example has been provided by those who want to use the disputed sources, and if we keep trying to come up with one, we'll be stalled for another month. I would like to see, perhaps Ronz put forward one or two sentences on our price/pricing distinction. SandyGeorgia (Talk) 02:07, 24 December 2019 (UTC)

C

Do you think that any individual pharmaceutical drug has one price that can be expressed in a given currency, for any region such as the US, UK, or the developing world?
— User:SandyGeorgia

D

Followup: What kinds of sources for pharmaceutical drug prices can be cited by editors, without original research, for Wikipedia drug articles?
— User:SandyGeorgia

Thanks for all your hard work, Barkeep49; this amount of effort is beyond the call of duty. SandyGeorgia (Talk) 01:43, 24 December 2019 (UTC)

E

For articles discussing medications, do you think that Wikipedia can, reliably and without original research, source medication prices as used in various regions of the world; and if found, should convert them to a common currency and include in articles?

— kashmīrī ^TALK 02:22, 24 December 2019 (UTC)

I am convinced that the average editor is not capable of answering this question in any useful fashion. Can it be done? Yes, because it has been done in many articles (e.g., multiple articles with accusations about price gouging). Also no, because it can't be done for all of them. And then the poor closer has to tot up the answers, and gets stuck with telling people that !voted yes that they actually meant no, and vice versa. The real question isn't whether it can ever be done; we have done that, and nobody's trying to blank the prices in Martin Shkreli. The real question is under which circumstances we ought to do that. So one answer's clear: No making stuff up or misrepresenting a price just to get something crammed into an article. Another answer's clear: If the price is in headlines all over the English-speaking world for months, then it gets mentioned. But where's the line between those two extremes? WhatamIdoing (talk) 05:53, 24 December 2019 (UTC)

There is no line, it is a grey area that would be undecided in many cases until local consensus was developed for that case. This is a tedious business and generally goes to the most persistent group who will not accept no for an answer. My opinion at present, as a non-expert with no dogs in the fight, is that prices are too volatile and variable to be useful to the reader in most cases, and that the onus is on the person wishing to add a price, to show that the price is encyclopedically relevant in that case. Where a price is considered relevant, I do not think it should go in the lead unless obviously noteworthy in context. · · · Peter Southwood ^(talk): 17:21, 24 December 2019 (UTC)

Peter Southwood, thanks for your comments. I do fear it is hard to create a rule that lays down the law that works in all cases, and perhaps should not attempt to ask for one. The problem with a "local consensus" cop-out is as you say the most persistent group will not accept no for an answer. I wonder if you will look at the drug articles I mention in my collapsed box (Colin's Background) below. Those are examples of the sticking point this RFC is trying to address. They are typical of hundreds of prices in leads of drug articles. I would like a question that answers whether or the article text we have is acceptable. Of course we agree on the question about "Whether drug prices should be widely presented on drug pages" or some variant of that. The second of Tryptofish's question assumes that we have some text about prices and just need to find an appropriate source. But I maintain that the problem with all the drug prices on articles currently is that they attempt to boil a drug down to one formulation, one tablet size, one unspecified indication, one dose and present one price. So before we ask what sources support it, we need to consider if that is even a sensible approach to presenting prices (never mind for now where: lead, body or infobox). Then for sourcing my concern about asking an open question is we don't get decisive answers, and perhaps it is easier to give examples from existing articles and ask folk if that use of the sources is acceptable. I'm sure that would also provoke some people to give alternatives. If we close the RFC with "there is no consensus approach to drug prices, editors should seek consensus per article" then we are back to the RFC of 2016 and the effect of that is that prices were added to all drug articles anyway, and discussions are ongoing to add prices and links to GoodRX on the drug infobox of all articles. Can you think of a question that will resolve whether the text+sources in current drug articles is acceptable to the community? -- Colin°^Talk 19:09, 24 December 2019 (UTC)

Colin so the idea of a baseline (prices are/aren't encyclopedic) doesn't seem helpful to you in this current discussion? Best, Barkeep49 (talk) 19:33, 24 December 2019 (UTC)

"The second of Tryptofish's question (sic)"? I was not asking two questions. I was stating two kinds of issues that need to be addressed. And I wasn't assuming anything. There could certainly be a consensus that drug prices should not be routinely included in articles, and that drug pricing should be discussed only when there are independent reliable secondary sources. To the best of my knowledge, those kinds of sources do exist. --Tryptofish (talk) 21:25, 24 December 2019 (UTC)

Colin, I am still thinking about when and why the price/cost/pricing/whatever of a drug would be encyclopedic information, and when it would not.
When it is not encyclopedic, leave it out.
When it is encyclopedic, a reliable source is necessary, that supports the information included. It is not "medical" information, it is commercial information, so MEDRS does not apply, just regular RS.
It is necessary to specify the context of the information in the article in such a way that the reader understands its scope and is not misled by how it is expressed - the information must be presented in a neutral way, and must be accurate, both in space and time.
The information should be widely valid unless there is an encyclopedic reason to provide information that is limited in space and time. When it is limited in space and time this limitation must be specified/explained.
In many cases the sources may not be suitable for inclusion without some numerical processing. There we run into the risk of original research, and the scope of the numerical processing should be specified to allow the reader to check the math. I think the sources should usually be explicitly stated in the text, as well as being reliably referenced.
We are trying to establish a principle here, rather than to judge specific cases, so I am not convinced that looking at specific cases at this point is either necessary or useful, as it may prejudice fair consideration of the principle. First we establish the principle, then we compare specific cases to the rule.
The local consensus to include is a vexatious issue, as we have editors who do not appear to understand the meaning or process of consensus building, and persist in claiming that they are right in the face of evidence to the contrary, often with a barrage of marginally relevant shortcut policy links rather than logical reasoning and supporting evidence, and frequently accompanied by a group of like-minded involved editors. Closure by an uninvolved neutral third party is a reasonable remedy.· · · Peter Southwood ^(talk): 07:01, 25 December 2019 (UTC)

Kashmiri, I am concerned that unless the original currency and date are quoted, the validity of the claim will be variable over time and partly obscured. There is already a problem with a single quoted price that it generally applies to a specific place at a specific time. In some cases this is relevant, but it is generically not very useful. If converted to a common currency, which one would be used, and why, and how would this be kept current?· · · Peter Southwood ^(talk): 07:50, 25 December 2019 (UTC)

@Pbsouthwood: Yep, but we don't want prices in South African rands, RTGS dollars or Chinese yuans, do we? Also, how do we approach currencies which quickly lose value due to inflation, like the Venezuelan bolívar? Anyhow, I proposed to have the currency component as a part of the question this so that responders have a pause over what voting "Yes" would involve. Maybe it can be reworded, though, although I do think that the RfC questions should touch upon the currency issue. Also, I am toying with the idea that we should somehow indicate that this would prices of pharmaceutical products (i.e., goods) and not prices of the chemical compounds being article subjects, with all the challenges related to different brands, formulations, dosages, package sizes, combination drugs, etc. — kashmīrī ^TALK 18:27, 25 December 2019 (UTC)

Kashmiri, you say we don't want prices in South African rands, RTGS dollars or Chinese yuans, do we? - Why not? Wikipedia is an internationally targeted encyclopedia, what currencies do we want prices in?
Also why specifically prices of pharmaceutical products, not the compounds, and could you elaborate a bit on the challenges related to different brands, formulations, dosages, package sizes, combination drugs, etc? · · · Peter Southwood ^(talk): 19:23, 25 December 2019 (UTC)

@Pbsouthwood: Why not compounds? Because chemical compounds are usually traded by weight or volume, unlike drugs. For instance, Salbutamol is a crystalline powder[75] and is indeed traded wholesale by kilograms[76]. But it is not an approved drug in the powdered form - approved drugs are various products (formulations) that contain salbutamol as their active ingredient: a sugary syrup, NO₂-propelled inhalers (in varying volumes and concentrations), a variety of capsules and tablets (2mg, 5mg, 10mg, normal release, controlled release, etc.); some of them may or may not be approved depending on jurisdiction. I think it would help if the RfC question informs editors about this aspect. — kashmīrī ^TALK 20:45, 25 December 2019 (UTC)

In general, prices change. Whatever prices are quoted should be explicitly linked to the dates and places involved. I do not know how the proponents of including generic/average/median/whatever prices plan to keep them updated. · · · Peter Southwood ^(talk): 07:50, 25 December 2019 (UTC)

Inconsequential formatting change

Kashmiri, in retrospect, I should have re-labeled my suggestions to C and D, to distinguish them from the earlier A and B. If you agree, would you change yours to E (and simultaneously change mine to C and D)? Thanks, I think this will be clearer and allow for other options to come forward without confusion. SandyGeorgia (Talk) 13:53, 24 December 2019 (UTC) @SandyGeorgia: Sure, go ahead, it's only a technicality. — kashmīrī TALK 15:07, 24 December 2019 (UTC) OK, it was me to do it.  Done then! — kashmīrī TALK 15:09, 24 December 2019 (UTC)

Table of Despair

I think that the main question to be settled is how much WEIGHT do we (by default) put on sources about prices (as understood in the dollars-and-cents model, not the general how-many-people-can-afford-that sense). Does it fall into (or near) the category of basic information that User:Bluerasberry calls Wikipedia:Defining data, in which case we need to include something about it whenever reliable sources permit us to say anything at all? Or, alternatively, is this content something that we should normally not include, and only mention when we have especially good sources (e.g., multiple high-quality sources that discuss the price at length). Here are some examples that we might consider:

Caption

Subject	Source type	If we put a lot of weight on prices	If we put less weight on prices	Notes
Insulin	Many long articles in news media and academic journals, including claims of price gouging and people dying because they couldn't afford the drug	In the US, between 1987 and 2017, the wholesale price of long-acting insulin increased from US$170 to $1,400 per vial.	In the US, between 1987 and 2017, the wholesale price of long-acting insulin increased from US$170 to $1,400 per vial.
Valproate	A 2017 peer-reviewed journal article, and some data points in various databases (i.e., independent primary sources)	According to estimates published in The BMJ in 2017 for drugs on the WHO Model List of Essential Medicines, the cost of manufacturing the active ingredient in this drug in India, is approximately a couple of US cents per pill.	(Nothing – this is a single primary source)	The cost of manufacturing the active ingredient is reasonably consistent worldwide. India is the biggest producer of these generic small-molecule drugs. But no retailer or consumer buys just the active ingredient.
Denosumab	An article in a pharmacy industry magazine (independent and possibly secondary)	Shortly after its original approval in 2010, Medicaid's US average wholesale price for a 60-mg prefilled syringe of denosumab was reported at US$990 per dose, with two doses expected each year to treat osteoporosis.	(Nothing here, but maybe something in the manufacturer's article)	Just one dose (of two for this drug), in just one country, at just one point in time, using just one metric (of many).
Golodirsen	An article in a biotech business magazine reporting on an Earnings call (independent and primary)	In the days after Sarepta Therapeutics received permission from the US FDA to market the drug, the net annual cost was estimated to run around US$300,000 per treated patient, assuming the patient was a child weighing 25 kg (55 pounds).	(Nothing here, but maybe something in the manufacturer's article)	No actual sales and little non-business coverage at that point, but high-cost drugs tend to attract attention, so maybe more sources would appear later, at which point it might be treated more like the Insulin example.
Abacavir	A routine entry in a government database	According to the Drug Pricing and Payment database maintained by the US Centers for Medicare and Medicaid Services, the National Average Drug Acquisition Cost for 300 mg pills of abacavir was US$0.77 each in December 2019 in the US.	(Nothing, because it's a primary source)

All of these examples have been mentioned in the discussions leading up to this point. If anyone feels like any of the examples are misrepresented, please let me know. WhatamIdoing (talk) 05:41, 24 December 2019 (UTC)

Despair sets in; we won't get people to digest this much information. SandyGeorgia (Talk) 05:59, 24 December 2019 (UTC)

I don't have a problem with any of WhatamIdoing's examples above appearing in an article on the drug (assuming that they are correct at face value). I would not expect to see any of them, except possibly the insulin information in the article lead, in which case I would expect a section in the article discussing the price rise and its impact in more detail. · · · Peter Southwood ^(talk): 08:08, 25 December 2019 (UTC)

Colin's Background

I think we have rather forgotten why we are having this RFC. I'm here because MEDMOS got edited to state a medication price agenda, and despite discussion on talk, received further non-consensual edits supporting drug prices in articles. I'm here because I saw the edit war at ivermectin. The article currently says "The wholesale cost in the developing world for the tablets is about US$0.12 for a course of treatment." with a source link that doesn't work. When we look at a fixed link we find the source is the Costa Rica Social Security Buyer price, which is never regarded as an International Reference Price (the median of multiple seller prices is required). We also find that $0.12 is not mentioned by the source nor is the quantity for a "course of treatment" mentioned by the source. This has already been pointed out and no WP:MED editor has joined with me in criticising it nor fixing it since January 2016. Rather than fixing current problems in article text, WP:MED suggest we make it even worse by put prices in the infoboxes. Indeed at Template talk:Infobox drug#GoodRx we have a proposal to include prices and links to GoodRx on all our drug articles. One editor noted "I just realized that since some brands contain multiple dosage forms..." but still wants to pursue it. It is claimed prices are "defining data" suitable for wikidata, despite every evidence to the contrary. The lack of any logic supporting that idea has not stopped it being repeated multiple times. I'm here because valproate currently says "The wholesale cost in the developing world is about US$0.40 per day as of 2015" despite the source not including a price per day and despite the source being a cherry-picked (from 6 record choices) awkwardly large 500mg tablet with one supplier rather than the therapeutically more convenient 200mg table that has six suppliers. This has already been pointed out and no WP:MED editor has joined with me in criticising it nor fixing it since it was added in December 2015. I'm here because we keep pushing wholesale (or retail) official list prices yet also keep mentioning a desire to meet the needs of those who pay out of their own pocket. It has been claimed "The wholesale and retail price vary little from each other" yet WHO remind us that even standard medicines like an Salbutamol inhaler is "virtually unavailable in the public sector of many countries" and "purchased from the private sector, can cost the lowest-paid, unskilled government worker several days’ wages". I'm here because Ethosuximide currently says "The wholesale cost in the developing world is about US$27.77 per month as of 2014" despite linking to a single supplier that only delivers by their own fleet of trucks to the north of the Democratic Republic of Congo (not "the developing world") and despite the source not giving a price "per month" at all. This has already been pointed out and no WP:MED editor has joined with me in criticising it nor fixing it since it was added in December 2016. I hope a pattern is emerging. We aren't having an RFC about some new or alternative idea for prices in Wikipedia articles. We are having an RFC about the actual current prices in actual hundreds of articles. We're having an RFC because of an impasse between two editors. And we're having an RFC because WP:MED has completely avoided making direct explicit criticisms of that text or of fixing any problems in the past three years. WP:MED is clearly not going to fix this and we need input from the wider community and neutral editors to contribute. We need to offer a question that directly resolves this matter, rather than creating new ones. -- Colin°Talk 10:01, 24 December 2019 (UTC) Colin, I am going to put this comment inside your collapsed box, so that honoring the spirit of Barkeep's intent in this new thread, we can stay focused on formulating the questions. But I want to point out: we once had an RFC/U process-- Requests for comment on User conduct. We don't have those anymore. I agree there are behavioral aspects to everything occurring at WPMED, and I regret that not being heard when you discuss policy and provide incisive analysis is causing frustration. It is also frustrating that most WPMED editors are not even engaging this page to help formulate an RFC, while you continue to work towards explaining and applying policy. You are heard. But we don't have Requests for comment anymore for looking at user conduct specifically, and we are where we are. We need to work with what we have. Could you take what I did do with the two questions, and merge them back to your one suggestion while preserving anything of value in mine? SandyGeorgia (Talk) 13:50, 24 December 2019 (UTC)

I think question A (by Colin) is a necessary and sufficient condition for the current text in hundreds of drug articles to be kept. Splitting in two like Sandy proposes weakens this, especially the open question B which may not likely lead to anything other than a random mix of opinions. The problem with a "What kinds of sources" question is that it always depends what you use if for and "for pharmaceutical drug prices in drug articles" is not specific enough. Many people like to view sources as adjectives. So "MSH" or "data.medicaid.gov" or "WHO" or "BNF" will be viewed as "reliable sources" and as "secondary sources" (they aren't the primary source of their data). So I suspect we'd just get comments like "Must use high quality secondary sources like the BNF" without stopping to think that the BNF may list 30 prices from 10 manufacturers for a drug, or even working out what the two prices the BNF list for each record actually mean. I'll try to find some representative article texts later. Essentially we want an "Are you happy with this?" question around existing practice. -- Colin°^Talk 10:21, 24 December 2019 (UTC)

• I'm behind on reading everyone's comments, and busy as well. More than a reminder to myself than anything, with no attempt to create simple questions for an RfC: When are specific prices due mention in an article at all, and when in the lede? How much pricing information needs to be included with any specific prices to make those prices meaningful to the reader? --Ronz (talk) 19:24, 24 December 2019 (UTC)
  • NOTPRICE needs to be supported. The WP:Prices essay deleted/moved. --Ronz (talk) 17:52, 25 December 2019 (UTC)
• So once again, Colin states his support for question A by Colin. Someone please alert the news media. So far, I'm not seeing any single question that adequately covers the two issues that I identified, without being confusing to editors who are coming new to the RfC, or being non-neutral. What comes closest are Sandy's combined C and D, with some further wordsmithing. But alas, that is not a single question. --Tryptofish (talk) 21:33, 24 December 2019 (UTC)

Putting this here at the bottom, to encompass much and good feedback above. Please, people, come on ... take Barkeep's suggestion and put up concrete suggestions so we can start wordsmithing and discussing specifics. Once the proposals are up, we can see the issues and refine. I am at the limit of my wordsmithing ability, and despair has set in; length has again taken over this discussion, and we have nothing concrete. Trypto and Peter and Ronz, give it a go even if you aren't yet fully satisfied with what you might intially propose. SandyGeorgia (Talk) 19:02, 25 December 2019 (UTC)

Thanks, Sandy, but given the responses I've gotten so far when I've previously given it a go, I'm going to sit back and see what others can do. I appreciate the fact that you, alone, covered the things I think need to be covered, in your two C and D proposals, which together seem to me to be the best so far. But, given Barkeep's request for a single question, and given what seems to me to be the predominant opinion of editors here, I think we need to make it into a single question. Without creating new problems in doing so. I'm not seeing any way to accomplish that. --Tryptofish (talk) 23:00, 25 December 2019 (UTC)

Who are "others" that we are waiting for to see what they can do? All of the "Me, too, per editor-so-and-so" !voters, who don't engage to develop consensus? Are you expecting someone new to show up here? It looks to me like we're looking at us, and if *we* don't come up with something, we end up at Arbcom. It should be well evidenced by now that I suck at formulating RFCs, so what are we thinking is going to happen next here? SandyGeorgia (Talk) 23:25, 25 December 2019 (UTC)

SandyGeorgia — It's Christmas day, even if you feel dispair that no one is engaging right now, we need to keep our cool and wait until at least the beginning of January. With the moratorium on change wrt prices, the fact that prices have been included for years, and that this debate has raged for over a month — there is hardly to be any cataclysmic effect of waiting another week or so. Carl Fredrik ^talk 23:41, 25 December 2019 (UTC)

<sigh> ... you are right, CFCF. It happens that, with a big Christmas party behind me, the food done with, and gifts purchased wrapped and delivered, today was the first day I could really finally focus. You are right that today does not have to be the day. My despair is more related to seeing Barkeep try so hard, and getting no results yet ... SandyGeorgia (Talk) 23:47, 25 December 2019 (UTC)

SandyGeorgia, as I noted below to Typto I do think we're getting somewhere. As Carl notes we're in a slow period. But I plan to recollect the various wordings, some of which seem to have more support than others, to see if we can get to a final version (if one doesn't emerge organically). Best, Barkeep49 (talk) 23:51, 25 December 2019 (UTC)

Thanks again for still going above and beyond the call of duty. Lavendar herbal bath is calling my name. SandyGeorgia (Talk) 00:00, 26 December 2019 (UTC)

Well, I've made suggestions, and they have apparently been shot down. And, as I said, I'm not seeing a way forward under the terms we are working with here. --Tryptofish (talk) 23:34, 25 December 2019 (UTC)

I think it's possible we end up with more than one part (and said as much in my extended comments). However, when we started with multiple parts it quickly spiraled out of control and to a place we'd never achieve consensus. I'm hoping by staying smaller it will be easier for us to achieve consensus here on the questions and then to achieve consensus when it's formally putout to the broader Wikipedia community. But yes all this does require on people actively collaborating on wording or else putting forth concrete support for the wording of someone else. This conversation has generated some fresh perspectives so I'm still hopeful we're somewhat moving in a positive direction. Best, Barkeep49 (talk) 23:38, 25 December 2019 (UTC)

Honestly, Trypto, I haven't intended to shoot down or ignore any of your suggestions: they just aren't yet formulated in a way that I can get my arms around them. Because of the history of really badly formed RFCs leading to no conclusions, I think that Barkeep was right to suggest a format in which we could more easily see exactly what the questions would look like, and discuss from there. At this stage of frustration, I could better opine on your ideas if you would put them in the format as A, B, and so on above. WAIDs huge table is where my despair started; while *we* get it, I just don't feel most participants will. We have to stay simple. SandyGeorgia (Talk) 23:57, 25 December 2019 (UTC)

Tryptofish's two part question can be split into two questions. The second question is not relevant unless the answer to the first is "yes", so we can take the first part and ask it, and if the answer is no, the second question falls away. I see the second question as relatively straightforward, as it refers to economic content as opposed to medical content, but it may be that others have philosophical differences about the scope of medical content, and I am happy to leave that debate for another day. I present a modified version of Tryptofish's question. · · · Peter Southwood ^(talk): 06:57, 26 December 2019 (UTC)

• Trying to respond to several points above. User:Barkeep49 "a baseline (prices are/aren't encyclopedic)" question will need to be carefully worded. People have very different ideas about what counts as evidence of encyclopedic. Perhaps worth quoting WP:NOTPRICES here for reference:

An article should not include product pricing or availability information unless there is an independent source and a justified reason for the mention. Encyclopedic significance may be indicated if mainstream media sources (not just product reviews) provide commentary on these details instead of just passing mention. Prices and product availability can vary widely from place to place and over time. Wikipedia is not a price comparison service to compare the prices of competing products, or the prices and availability of a single product from different vendors or retailers

We have different opinions expressed about what that means and whether we should challenge it. WAID appears to want us to reconsider WP:NOTPRICES for any topic. Others believe WP:NOTPRICES already disallows the "two or three prices in the lead of every single drug article" approach that has occurred. When the above was quoted on 6 December, James wrote "And I agree with that. All the sources being used to discuss prices are independent. And we have sources that justify the importance of pricing information". We have some editors who think prices concerns in national newspapers is what counts as independent discussion and WP:WEIGHT to include. Others claim that multiple drug databases listing prices satisfies WP:WEIGHT and finding some obscure internal memo on the price of X is sufficient discussion, or that general concern about drug pricing is enough to justify the inclusion in every single drug article. So we need a question that cuts through all that to make a clear consensus. If we just end up with a reworded WP:NOTPRICES that is immediately ignored/evaded we haven't achieved anything.

Can we try not to personalise the ownership of questions either as a reason to negate someone's support for them or to get all defensive about them, though unfortunately to discuss options we sometimes have to label them with a proposer's name. I appreciate Tryptofish's two "ideas" weren't literal "questions". I think we should be free to state concerns about one question or another without having those concerns described as "shot down" or "rejected". One reason I so dislike it when wiki/commons jumps immediately to a vote is that all discussion just gets polarised into adversarial language supporting one's position and rubbishing the opposition -- we see that elsewhere in the MEDLEAD discussion. Sandy's C and D could well work. My nervousness with D comes also to the area User:Pbsouthwood is commenting on: which angle should we approach the problem? I could well be wrong and by asking an open question we get some really good clear advise (personally, I'd be hoping that we'd agree that database sources we use currently are useless to us in anything other than unusual cases, because of all the original research, cherry picking, synthesis, etc). I fear however we may get unclear unhelpful replies where words like "independent", "secondary source" and "reliable source" are used which I hope everyone here agrees are in the chocolate teapot department of unhelpfulness to us wrt the current dispute.

There seems to be some consensus around the C and D questions. I do think it would be useful to include examples, especially of current usage both for the "routine drug pricing citing price databases" and for the "exceptional drug pricing citing some newspaper or commentary source". -- Colin°^Talk 12:37, 26 December 2019 (UTC)

arbitrary edit break

F

Should reliably sourced and unambiguous dosage prices be routinely presented in articles on drugs (medications) in a prominent position (lead or infobox)
— User:Pbsouthwood

Pbsouthwood, can you give an example (current or invented) of such? I don't think I have seen any "unambiguous" (within article text) statement of dose (it has to be inferred from the database record cited, and in the case of some US prices, can't be inferred at all because the source links to a set of 25,000 prices for a given week). Take diazepam for example. It states "The wholesale cost in the developing world is about US$0.01 per dose as of 2014.". Again there are a whole host of problems with this because the source lists only the Buyer price in the Dominican Republic and in Peru, and lists no suppliers at all. This should have run alarm bells for such a huge drug. In fact the 10mg tablet would appear to be uncommonly used (hence no suppliers at all, and in 2015 only Peru was a Buyer) and the 5mg tablet here with eight suppliers is far more reasonable. But what is a "dose". Would someone writing "costs ___ for a 5mg dose" and citing this be "reliably sourced and unambiguous" in your book? The source does not indicate which tablet size to pick or what is a "dose". Nor, for other drugs, does it suggest whether to use enteric coated tablets, or suspensions. The MSH is a "reliable source" for some things (nobody is doubting it is generally likely to be correct about the prices it lists) but not a "good source" for others (the Buyer prices in Peru and DR are not considered representative of the "developing world" according to ANY recommended usage of MSH). Other articles give a cost per day or per month or per treatment. Are they "dosage prices"? By "unambiguous" does one need to include the exact indication too. For example, the BNF gives all sorts of dose options for various indications, many in a range. How would we pick which indication? And if we picked one ("Muscle spasm of varied aetiology") the adult dose is "2–15 mg daily in divided doses, then increased if necessary to 60 mg daily, adjusted according to response, dose only increased in spastic conditions." How does that translate to a "dosage prices" in an article? The BNF prices are here.

Btw, James has hinted he can't access the BNF (other than a paper copy which is very abbreviated wrt price). If others are having that problem, it would be good to know prior to the RFC. A VPN allows you to access the web from another country, and I've used that to access the US GoodRX website for example. -- Colin°^Talk 12:37, 26 December 2019 (UTC)

Colin, I have no example of an unambiguous dosage price, and am not sure I could find one if I looked for it. I consider the issue of medication dosages and pricing as outside of my skill set, and am not a big enough fool to try to bluff my way through. I do think I can recognise several varieties of ambiguity that might come up in an attempt to present such an unambiguous price, and you and others have mentioned a few of them already. My point is that it may be possible in some cases, and that it could be considered by the Wikipedia community whether, if and when it is possible, it is to be accepted as a standard component of an article on the drug. I am taking as a given that ambiguous dosage prices are not encyclopedic and are not acceptable, as being potentially or actually misleading to the reader, which is a MEDMOS issue. It is the responsibility of the editor adding the information to not only provide a reliable source, but also to ensure that the information provided is not misleading. Competence is required, both in researching the content, and in presenting it in an article. I may have the competence to identify a dosage price statementas badly expressed or ambiguous, but it is unlikely that I will ever try to add one. Your example for Diazepam for muscle spasm suggests that in many cases it is not reasonably practicable to provide an unambiguous and useful dosage price, even if restricted to a single market with stable prices. · · · Peter Southwood ^(talk): 16:27, 26 December 2019 (UTC)

PS: BNF is only available in the UK according to the website, so not accessible to me either. · · · Peter Southwood ^(talk): 16:27, 26 December 2019 (UTC)

Would someone writing "costs ___ for a 5mg dose" and citing this be "reliably sourced and unambiguous" in your book?
That price is for 2015, which is not mentioned in text.
The actual price is not mentioned in text. Several are available in the source, which one is actually used is important, and it should be mentioned what kind of price it is. Buyer or supplier, median, mean, or range.
The specific source is one of possibly may, and is not mentioned in text.
5mg dose is unambiguous, but unless the article goes into some detail elsewhere of what dose is appropriate in various circumstances, it is not very useful.
The dosage form (tablet) is not mentioned in text, but I don't know if this is relevant for this drug.
I would call it ambiguous. I have no particular problem with reliability of the source as such. · · · Peter Southwood ^(talk): 17:22, 26 December 2019 (UTC)

F is getting at part of something that must be addressed. This whole dilemma (according to some of us, anyway) is the result of the breach or misapplication of multiple policies and guidelines (NOR, WEIGHT, NOT and LEAD). Some are suggesting that we should revisit guideline and policy (overall, for the case of MEDLEAD for the purpose of translation, and NOTPRICE for the case of medical product pricing per WAID).

I am concerned that our questions must be asked in a way that addresses the NOR aspect of these drug databases, the WEIGHT aspect of whether we should include prices at all, and the LEAD aspect of whether they should be in the lead. If we really have so many core policies being challenged, what the heck. NOR, NOT and WEIGHT are policy; why are WikiProject guidelines and practices challenging policy and why is that not being done with RFCs on the core policies? I am going to end up dissatisfied if we don't have questions that will get us to the core problems. If we still had an RFC/U process for user conduct, we would be asking these questions there. SandyGeorgia (Talk) 14:56, 26 December 2019 (UTC)

Agree that three core policies are being breached, though we have rather lacked WP:MED regulars confirming this, leading us to only speculate why. This is partly why I am reluctant to ask a question that is essentially "Shall we allow several core policies to be broken for drug prices" rather than "Can we do _____ while still following core policy". The insertion into the lead only is an important problem, but perhaps not our first priority. Some have argued that some basic defining data could be in a lead (or infobox) and not repeated in the body. Of course nobody has satisfactorily explained why we should even consider price a defining data, other than to say so in order to justify inclusion in infoboxes or wikidata. The clear example is that drug prices can jump extortionately just because a generic manufacturer is bought out by a rival or decides to drop out leaving only one manufacturer, etc, emphasises that prices can be purely an artefact of the games soulless businesses play. The price of a drug in country YY can change purely because they adopt external reference pricing as a policy rather than cost-plus -- nothing to do with the drug itself. I can't really get my mind round the idea that we could take the dozens of prices on a BNF price page, multiplied by a handful of different indications and patient groups suggesting dose ranges, and condense that down to one entry in an infobox or one sentence in our lead. But that hasn't stopped multiple editors consistently and persistently saying they are fine with it and want to go further. So maybe we need to ask what seems to some of us as obvious, but we shouldn't be asking to be allowed to break core policy. -- Colin°^Talk 15:22, 26 December 2019 (UTC)

So, let's look at F from the perspective of the admin who will have to close the RFC. MEDLEAD is only guideline, so it seemed to be a good small piece to get out of the way early on. We see translation advocates agreeing that guidelines should be ignored for the "greater purpose" of "children in sub-Sahara Africa" (although the overall benefit to anyone, much less children in Africa, is highly dubious, but I digress). We see almost no one engaging the policy/guideline fundamental questions (partly because of the RFC framing, but that is precisely the problem we are facing here, hence the example). And we see a matter too complex and entrenched for average editors to engage.

So, let's consider the answers that F will generate, and how the closer will interpret those? F presumes a baseline understanding of and achknowledgement of reliably sourced and unambiguous dosage prices. We don't have that in this price dilemma. If we did, we wouldn't be here. So, what will the closing admin do with the "ILikeIt", "Me, too, per editor-so-and-so" responses that will not engage the core questions and policies? The RFC MEDLEAD shows we will get "because I like it" responses. We need to be highly specific in our questions about the core policies: NOR, WEIGHT, NOT. With the MEDLEAD RFC, a closing admin can argue that guidelines are flexible and can be ignored. In that case, with respondents not engaging the core questions, we end up with protracted local discussions to determine consensus on individual articles, where one group will argue LEAD and another group will argue MEDLEAD. We end up with articles that cannot be taken to FAC, because you can't please two masters. This is really not a big deal, because essentially no one at WPMED is attempting to write complete articles anymore anyway, and no one is maintaining most of the project's Featured articles.

Unlike the guideline LEAD, on core policy questions, the problem cannot be so easily overlooked. In this case, if we end up with an inconclusive RFC where respondents do not engage policy because we haven't asked the questions with great specificity, what's next? Protracted local disputes end up at arbcom. SandyGeorgia (Talk) 16:22, 26 December 2019 (UTC)

Peter Southwood, I think many article texts are ambiguous. They just say "per dose" or "per day" or "per treatment" and that is "ambiguous" in the sense we don't in-text name the dose or indication or any other factor that influences how much you give, how often and for how long. But suppose we did. Suppose we named a 10mg Diazepam tablet or we stated a full course of X antibiotic at 25mg per day for 8 days. Then the "a" word I have a problem with is "arbitrary". Someone picked 10mg dose and chose a tablet rather than a rectal suppository, and chose an indication for treatment. So while we can certainly make minor changes to improve the article text to be unambiguous and many will argue the sources we cite are already "reliable", neither seem to me to be sufficient to allow what we currently have, even improved with minor tweaks. So I wouldn't want the end result of the RFC to just be "we need to name the dose and formulation in-text".

Btw, I just installed windscribe.com extension onto my Chrome browser. It is free and I get 10gb a month through it. I was then able to select a server in Dallas US and access goodrx.com. I do suggest folk in this discussion do similar for a virtual visit to the UK and have a look at BNF for drug details including prescribing dosages and indications and medicinal forms including prices. I really wish the folk who think we can put this in an info box would look at that -- the infobox would be bigger than most articles. -- Colin°^Talk 17:21, 26 December 2019 (UTC)

Arbitrary is generally not useful to the reader, so I would consider it unencyclopedic. · · · Peter Southwood ^(talk): 17:30, 26 December 2019 (UTC)

Here are the things that I can identify that need to be addressed for F. When we ask editors whether that kind of information "should" be handled that way, we still need to somehow address "as opposed to what?" It's easy for an RfC to close as "yes, it can be done this way" and then have someone come along and assert "but it didn't say we cannot do it this other way" – or for it to be closed as "no, don't do it this way", and someone claims that "my way isn't that, so my way is OK". Also, however we present a question, it's important that we present specific examples of what it would look like in an article, for at least two contrasting alternatives, and that we present editors with concise policy-based arguments for or against whatever it is. --Tryptofish (talk) 18:56, 26 December 2019 (UTC)

Tryptofish, agree that there may be a problem with evading any decision by clever words/interpretation. That is a current concern wrt WP:NOTPRICES for many folk who disagree that drug prices meet that policy requirement for all drugs. If one person says "must use reliable secondary sources" to mean a commentary in a newspaper or journal explicitly talking about the high/low price and perhaps quoting a figure, another person might say "But MSH's prices are secondary sources (they aren't the supplier or manufacturer) and they are reliable". Wrt options, we need to always remember that one option is to not do it at all. For example, there are so many indications, formulations and manufacturers of diazepam that perhaps we conclude that wiki should not emulate the drug databases by trying to list them or nor to falsely condense them down to one price.

At the other end of availability, look at Terbinafine. Peter Southwood, you mention above whether "dosage form is relevant for this drug [diazepam]". It certainly always affects the price, so picking one form or table size can hugely affect the price quoted. Different forms are required for different patient groups/ages and indications. The most amusing I just found was Terbinafine. We cite a source commenting (as an aside) that the price for a 12-week treatment fell from $547 to $10 after the patent had expired. Is that price drop notable or typical? Anyway, the point is the source is talking about a lengthy oral (tablet) treatment for nail fungus. But we also, in the previous sentence say "The wholesale cost in the developing world is about 2.20 USD for a 20 g tube" which isn't for nail fungus (the cream isn't effective) but for athletes foot or other itchy skin, which typically clears up in days, not months. So we are being very ambiguous and the juxtaposition is very much against policy. But also importantly, the source from 2014 doesn't list any Suppliers at all and only one Buyer (an organisation representing 9 Eastern Caribbean states). And MSH doesn't have any entry for the tablet. The only reasonable conclusion is that Terbinafine cream or tablet are not generally available from suppliers to governments in the developing world, yet we are so desperate to give a price that we quote any old garbage database entry we find and claim falsely that this represents "the wholesale cost in the developing world". Mostly, we should pick the "say nothing" option, and our questions should allow for that. -- Colin°^Talk 09:11, 27 December 2019 (UTC)

Yes, we agree about that. I think there's a world of difference between taking terbinafine orally, under an MD's supervision and perhaps monitoring liver function, versus using an over-the-counter cream for an itch. --Tryptofish (talk) 23:26, 27 December 2019 (UTC)

NOR vs DUE

I really think this is important: NOR is not DUE.

Deciding which entry to cite in a database is not a NOR violation. It's (possibly) a DUE violation, but an accurate description of the content published in a reliable source is never NOR. This means that if you look up wonderpam in The Database, and it says "100 mg pill – generic – Specific Price Type – US$0.10 – December 2019 – UK", then writing "According to The Database, the Specific Price Type of a 100 mg pill of generic wonderpam was US$0.10 in December 2019 in the UK" is not original research. That's actually what the published reliable source said; therefore, that's not NOR.

Now, while that statement is not a NOR violation, NOTPRICE suggests that it's also probably not something that we want. It's possibly unencyclopedic, and it's very likely UNDUE emphasis (why that one size, that one date, that one country?), even though it's not actually original research. I don't think that we'll get a sensible RFC response if we go to editors and say "He copied this information straight out of a single reliable source – that's a NOR violation, right?" NOR means "material—such as facts, allegations, and ideas—for which no reliable, published sources exist". If you're copying it straight out of a single reliable source, then it's not NOR. IMO we need to stop talking about NOR (which is either not a problem at all, or is a problem that can be fixed by copyediting) and focus this discussion on DUE.

"Focusing on DUE" IMO means that we ask editors how much emphasis we should put on this subject. "How much emphasis" is partly subjective. Yes, you have to have the sources, but if something is "always" DUE, then you can/should write a weak claim from whatever source you can get.

We need to know whether editors want minimal emphasis (in which case, we remove a number of existing statements, or at least move them out of the lead), or whether they want significant emphasis (in which case, MEDMOS can provide information about how to write non-NOR statements), or something in between.

It would be good to educate the respondents about how complex the subject area is, but if they want significant emphasis on this, then the feasibility of implementing their goals in any given case is not the primary factor in writing our advice. After all, we put a huge amount of weight on the dates and locations of people's births, even though we know that can't source birthdates and locations for every single biography. Putting a huge amount of weight on it just means that if you've got any source at all, even if it's just one unimpressive primary source, then you include whatever you've got, with whatever WP:INTEXT attribution and careful description that allows readers to understand the limits of the source.

When this started, I gave a pair of examples in MEDMOS:

• Y Do this: In the US, between 1987 and 2017, the wholesale price of long-acting insulin increased from US$170 to $1,400 per vial.
• N Don't do this: Insulin costs US$1,400.

If editors want prices at all costs, we could add one that shows how to respect NOR while still including database-derived prices:

• Y Do this: "Prices vary according to dose, location, and other factors. As an example, according to Named Database, the government of Costa Rica paid US$0.10 per 100 mg pill for generic wonderpam in 2015."
• Y Do this: "Under the affordability model published by the World Health Organization and Health Action International in 2008, a medication is affordable if a month's treatment costs less than seven days' wages for the lowest-paid unskilled government employee. In the WHO/HAI database, generic wonderpam is classified as an affordable medication."

We could also add an example from news media:

• Y Do this: "Shortly after its original marketing approval by the US FDA in 2010, Business News estimated the manufacturer's list price of Wonderpam at US$1,000 per day in the US."

I know those statements can be sourced for some medications without transgressing NOR. IMO what we need is to know whether editors actually want us to do that.

With that in mind, I think that the question to ask is:

G

How much WP:WEIGHT should be put on the prices of pharmaceutical drugs?

This question can (and IMO should) be followed by examples (the despair-inducing table) and explanations (the impossibility of finding the One True™ Price for a drug that sells in 190 countries in six different doses and four common formulations under at least 90 brand names) and alternatives (we could skip dollars-and-cents and instead try to source a comment about affordable/expensive), and comparsions to similar subjects (e.g., how NOTPRICE is applied to other products), but I think that this is the most basic question to ask.

I do not think that we should be asking a yes/no question. I think editors should be encouraged to respond with both an overall view related to inclusion (e.g., always include, usually include, usually omit, only include under extraordinary circumstances, only for generic drugs, only for WHO Essential Medicines, only for drugs under patent protection, only for drugs with unusually high prices – whatever editors actually want) and with an idea of how to include (e.g., brief mention in the lead vs a whole paragraph or section, current prices vs original prices, etc.). WhatamIdoing (talk) 19:14, 27 December 2019 (UTC)

I do think we should give some serious attention to presenting the RfC in such as manner as to not conflate policies. But I also think that there is a serious need to keep things simple, rather than tl;dr. If we ask "how much weight?", there will be the issue of how to answer the question. How does one define "a lot of weight" versus "not much weight"? --Tryptofish (talk) 23:23, 27 December 2019 (UTC)

I'm still not seeing a simple RfC coming from all this.

Having information in the lead, an infobox, or in the article body are issues of weight. Presenting information in inappropriate context can be NOR, NOT, or POV issues.

I think this examples of what to do and not do are very valuable in moving us to some consensus, and giving proper guidance to editors who are trying to address these problems in articles. --Ronz (talk) 23:56, 27 December 2019 (UTC)

I, too, am still not seeing a simple RfC here. I also think that concrete examples are very useful – not only to what we are working on here, but also potentially to be included in some way in the RfC. I'm willing to be a bit flexible with regard to covering every issue, even every significant issue (infobox versus body text, for example), in the interests of a manageable RfC, but I also think increasingly that we will just be spinning our wheels if we keep trying over and over to craft a single question. --Tryptofish (talk) 23:53, 28 December 2019 (UTC)

WhatamIdoing, sorry but I think you are totally wrong, because you keep having a strawman argument about fictitious possible price/cost/affordability statements some fictitious article might contain. We aren't having an RFC because of fictitious possible article text. Of the 500+ drug articles that currently display prices, all of them required original research to make the statement they do. When we chose one unnamed tablet to represent "the cost", that was indeed original research and when we multiplied by a "dose" that was also original research. Have you looked at the medicaid source links? They cite a "prices for week xx in 2018" database of tens of thousands of records. You need to then, by hand, filter the results to the drug the article is about and then you need to look at different formulations and tablet sizes and try to reverse engineer which one was picked to get $50.45 a month or whatever we claimed. And of course the medicaid site doesn't mention a dose at all, so no "intelligent reader" could possibly work it out from that source -- the very definition of original research.

And we also made claims that are not supported by the source at all, like "the wholesale cost in the developing world" citing one buyer price in Costa Rica, or claiming the price ranged from $x to $y when that is also not supported by the source. None of actual articles formulate the price statement like you did. If they did, then you could indeed make the argument that it was undue weight to mention one tablet size from one drug manufacturer in one country. But they don't and the difference is not solvable, as you put it, by "copyediting".

We nearly always give a price per dose, per day or month (which also require picking a dose) or per treatment (which requires an indication and dose). But we always don't mention what that dose is nor do we mention what the indication is. Terbinafine was one example above, Aciclovir another I spotted today, where the article does not state whether the costs are for a cream for cold sores or a tablet for shingles or post-transplant infection of cytomegalovirus. Yes there would be a WP:WEIGHT problem with explicitly giving the price for just one indication, and there would be a WEIGHT problem if we were explicit about the formulation/supplier/etc we used for our maths. But we aren't even specifying the indication, never mind the other things, so our problem isn't WEIGHT but just being dishonest with our readers in claiming there is One Price. The "we could be as specific as The Database is" argument is a false one to make and knock down, because nobody is proposing it. As you say yourself, just giving an example of all the permutations should be despair-inducing. We need to also remember MEDMOS prevents us from stating drug dose information in articles (for good reason) so we can't even explain to our readers why we picked the 250mg tablet.

WhatamIdoing, I have complained about the horrendous problems with the lead text in 500+ drug articles, and I think an RFC that appreciates the problems with that text will result in all those lead prices sourced to databases texts being removed from all 500+ drug articles. Please, the RFC must resolve the conflict over existing article text, not generate imaginary conflicts with imaginary texts. I don't think your WEIGHT question resolves this conflict at all. It isn't even in my mind a sensible question to ask. WP:WEIGHT is determined by reading the body of literature on the article topic, not by Wikipedians expressing a personal opinion. -- Colin°^Talk 12:54, 28 December 2019 (UTC)

IMO, all of this is original research (not simple math). Let's not archive those sections yet. SandyGeorgia (Talk) 14:26, 28 December 2019 (UTC)

Help me out with this, User:SandyGeorgia and User:Colin. You've been consistent in thinking that it's NOR. If you're looking at a database record in a reputable, WP:Published database that says this:

"wonderpam – 100 mg pill – generic – Specific Price Type – US$0.10 – December 2019 – UK"

and someone uses that to write:

"According to The Database, the Specific Price Type of a 100 mg pill of generic wonderpam was US$0.10 in December 2019 in the UK"

where exactly in that sentence do you see "material—such as facts, allegations, and ideas—for which no reliable, published sources exist"? Or would this be fine (in NOR terms – I still have my doubts about it in DUE terms), and the problem is "merely" that none of the articles using this sort of source are following this model? WhatamIdoing (talk) 17:23, 28 December 2019 (UTC)

• WhatamIdoing, I'm working on a page that lists pretty much all the price statements I can find on our drug articles. Today or tomorrow I'll have something I hope. I think that should give us all a clearer idea of what is on Wikipedia and perhaps spark some suggestions about good things and maybe we can clear up some bad things. I strongly suggest you follow my suggestion above about getting a VPN for your browser and travelling over to the UK to have a peek at the BNF. It is easy to turn the VPN on and off and to travel to nice places. The BNF link for Colecalciferol aka vitamin D3 has 216 formulations/manufacturers listed. And that's only for manufacturers that supply the NHS, Amazon list a bazillion other options too. From what I've seen so far on wiki, a minority of articles mention a formulation/strength for a price, rather than the OR step of quoting a price per dose/day/month/treatment seen in the majority, but even those aren't making claims fully supported by the source. Rather than discuss imaginary databases, can you use one of our actual drug sources and see if you can come up with text you think avoids NOR and only has DUE issues. -- Colin°^Talk 17:47, 28 December 2019 (UTC)
  • Colin, thanks for this question at the end. This MOS guidance is about what we should be doing. This doesn't mean it always will be done, but does give editors who change content in that direction a basis for doing so. In my own area of content expertise, books, the MOS discourages sections about characters. Yet I frequently find character sections and when I do I normally just remove them referencing the MOS. The same will likely be true here. After the RfC there will be guidance about how price/cost should be used but it will still be up to editors to make that happen (both with existing and new content). Best, Barkeep49 (talk) 18:00, 28 December 2019 (UTC)
  • (edit conflict) Colin, I've done that. Remember the despair-inducing table? Those all have real-world sources behind them. I usually aim for obviously "model" sentences in guidelines, but if you want to see one that has real-world drug in it, with a real-world database cited, then that sentence could just as easily say "According to the Drug Pricing and Payment database maintained by the US Centers for Medicare and Medicaid Services, the National Average Drug Acquisition Cost for 300 mg pills of abacavir was US$0.77 each in December 2019 in the US.[77]" What I want from you is to know whether this model (which does not appear to have been used in any articles yet; I know) is, in your excellent judgment, a violation of NOR. I grant that it may have other flaws (e.g., DUE and NOT), but right now, I just want to know whether you believe that sentence, from that source, is a NOR violation. WhatamIdoing (talk) 18:10, 28 December 2019 (UTC)

• While a model statement might be a good idea in a guideline, we have historically had a problem throughout the price debate of claims being made about sources providing information that they don't. For example, it has been repeatedly and falsely boasted that the MSH database is a source of external reference pricing when in fact our up-to-date sources on that topic do not claim its use (instead, a basket of prices from specific countries are used). While the MSH records are no doubt generally reliable for what they are, if you ask me, I would say that it is not a reliable source for prices for "the developing world" because its records are extremely patchy. There is a reason why when WHO use it for price comparison studies, they restrict themselves to 18 common drugs at very specific formulations and strength -- those are ones they can be sure have a healthy set of supplier data.

So, back to your example. I clicked on the link and didn't see any mention of abacavir on the page. This may itself fail NOR requirements, because our dear intelligent reader has to figure out how to use the database, and it really isn't obvious. One might be tempted by the big [Search] box on the top, but that would be a mistake. So I click on the [View Data] button. It returns the first 100 rows out of 7,673,560! By a complete and lucky chance, abacavir is on this first page. But you and I both know that wouldn't generally be the case. So the reader has to figure out to insert "abacavir" into the [Find in this Dataset] box. Initially I got side-tracked by the [Filter] button but that led down a rabbit hole. (It isn't always the wiki article title you need: if you type "valproate" into the box, you'll get nothing). It isn't clear to me why I get 13 rows for the 300mg tablet for 12/18/2019 but at least they all have the same price of $0.77418 each. I don't know if that amount of manual-intervention and learning-how-to-drive-a-database is permitted for our sources? But assuming it is, then I would accept your article text does not breach NOR. Our current text, though, is in a galaxy far far away from that sort of sentence, and I sincerely hope nobody would want to write that in a lead. -- Colin°^Talk 18:39, 28 December 2019 (UTC)

Learning how to drive a database is permitted for sources. We may still have hundreds of probable NOR problems in articles right now, but we've got an agreement that something could be done with databases. I'll go add this to the Table of Despair. That database is probably not a good indication that including it is obviously DUE, right? So it'd be included if we put lots of weight on it, and not if we take a more stringent NOTPRICE approach. WhatamIdoing (talk) 18:52, 28 December 2019 (UTC)

It is worth pointing out that Abacavir currently says "The wholesale cost in the developing world as of 2014 is between US$0.36 and US$0.83 per day. As of 2016 the wholesale cost for a typical month of medication in the United States is US$70.50".[78][79] The DDD on the MSH site says 600mg daily dose, as does my BNF and Drugs.com. So that's two 300mg tablets a day as the developing world prices agree. But the US price of $70.50 is approximately the $70.35 I get by multiplying a 300mg tablet price of $2.34487 by 30. The actual 2016 US monthly price should be $140.69. (Why we have one price per day and another per month is beyond my understanding). Leaving aside that neither MSH nor Medicaid state what the typical therapeutic dose is, for us to do original research on, this is just mathematical incompetence. And it is very typical. I am repeatedly seeing prices citing the BNF that assume a pack of 28 tablets is a "month's cost", totally ignoring that a patient might taken more than one tablet a day. So, the evidence does rather suggest that disallowing original research is a jolly good idea, because we are crap at it. And then we see that the price hasn't been updated since 2016. As your citation shows, the equivalent price in 2019 for 60 tablets would be $46.45, which is about $100 a month less. No small change that, but nobody it seems, is interested in either the price in 2016 being right, or giving the right price for 2019. -- Colin°^Talk 21:40, 28 December 2019 (UTC)

@WAID, I'm not ignoring this, just really needed to back off here for a bit and let you all take the lead. I've got my own Table of Despair, that is awaiting your feedback. @Colin, add to all of this the prescribing practices I see in the free clinic for migrant workers without insurance. If a pill can be split, and is less expensive in a higher dose, the physician prescribes the higher dose if the patient appears competent to be trusted to split it. If the drug is on the 30-day $4 list, or the 90-day $10 list, they prescribe whatever is cheapest, even if splitting is needed, and even if they are only saving $2 on 3 x 30 vs. 90 days. SandyGeorgia (Talk) 22:25, 28 December 2019 (UTC)

Colin, I wouldn't be surprised if there were any number of accidental errors, but the goal here is to write advice on how to do it right. The fact that it's currently not right in many articles may be a disaster, but it's a nearly irrelevant disaster. Once we get some decent advice together, we can sit down and apply the advice. If we try to fix all of that before getting an agreement about what the right approach is, we might end up making the articles a different kind of wrong, rather than really fixing them. WhatamIdoing (talk) 06:02, 29 December 2019 (UTC)

WhatamIdoing, Abacavir is an interesting example for another reason. We can eliminate OR by stating a price for one thing -- with a barcode -- but then you argue the problem is DUE because there are multiple (dozens, hundreds even) of possible things with barcodes we could pick. But from my reading of most HIV pills, they tend to have one dose that everyone takes, and a limited range of suppliers. So the Abacavir 300mg pill in the Medicaid database is the only size in that database. It could then be argued that we could give the price of a 300mg tablet in the US. But we know the dose is 600mg, taken once or twice a day, and we can't tell the reader that because we aren't allowed to give dosing advice on Wikipedia. So the price of a 300mg tablet is fairly meaningless to our reader -- they don't know what to do with that information. That's probably why nearly all our price statements in articles give a price per day/month for an unspecified indication and unspecified dose and unspecified tablet size. So even in the few occasions where a particular tablet size is not undue, using the "price of a XXmg tablet" approach is not likely to be encyclopaedic. (Btw, the BNF has three suppliers for the 300mg tablet pack of 60: £177.60, £177.61 and £208.95 for the brand-name, as well as a 20mg/ml oral solution at £55.72 for 240ml, so darn the BNF for offering options!) -- Colin°^Talk 12:40, 29 December 2019 (UTC)

In terms of presenting this question, I think it needs a bit of explanation. The straight-up question is "How much weight?", but after that, some explanation is necessary. One way (of many ways) might be to explain the context, and then offer some considerations. It could look something like this:

All Wikipedia articles should present information with WP:Due weight. Generally, this means that the more our reliable sources talk about an aspect, the more attention that aspect should get in the Wikipedia article. However, there is some information that is considered so important that it is included whenever possible. For example, in a biography, we include information about the subject's birthdate whenever possible, but we normally mention the subject's hair color only if reliable sources dwell on the person's appearance.

Drug pricing and affordability is a significant area of discussion in reliable sources, but this discussion is almost always held at a general level, and does not extend down to individual products. The prices of individual pharmaceutical products vary so widely by place, time, dose, and other factors that general claims, such as "the price worldwide" or "the price in developing countries", are almost always incorrect. It is, however, frequently possible to source a statement about what a particular metric yields for the price of a particular size of a particular drug from a particular manufacturer in a single country on a given date.

Editors who work on medicine-related articles have recognized that much of the information about drug prices currently in Wikipedia articles is not an example of our best work. Much of it is outdated or otherwise incorrect. We want to fix it, but we have not been able to agree on the best approach yet. On the one hand, the cost of a drug affects whether people can get it at all, so some editors believe we should always include whatever we can source. Other editors believe that pharmaceutical drugs should be treated like any other manufactured product, and that means no prices unless we have multiple reliable sources discussing the price of that particular product in depth (as we do for some, usually because of very high costs). Editors fall across the whole spectrum from maximizing inclusion and prominence, through the middle grounds, to the opposite side of including as little price information as possible. All of us want to know other editors think, so we're asking you: How much weight should we put on drug prices?

To explain some of the positions, a few editors have offered background information that may be useful to you. We hope that you will join us in a conversation about the best way to handle this subject area.

(Collapsed – It's incredibly important) (Collapsed – You wouldn't believe how complicated and useless this is) (Collapsed – The middle road is WHO/HAI affordability, not prices) (Collapsed – People should care about pricing, not prices) (Collapsed – Where and how we mention prices matter more than whether we do) (Collapsed – What we could actually source is unencyclopedic) (Collapsed – Whatever other ideas/positions/recommendations I've forgotten)

My suggested "collapsed" sections could be written by different people, in the hope that editors would read more than just the headlines. Yes, it's long. That's not necessarily a showstopper. The important question is, if we asked this, do you think that we would get responses that would help us figure out how to clean up these articles? WhatamIdoing (talk) 06:53, 29 December 2019 (UTC)

I think we already know how to clean up these articles. And I don't think we can craft an RFC that will prevent this from happening again after the RFC closes. I am unsure if any of the very well-crafted text above should be positioned as referencing editors (plural) representing ranges of differing opinions, because that gives the equivalent of UNDUE weight to a very minority position that has led all of us to all of this effort to attempt to formulate an RFC, where we used to have an RFC/U process for these kinds of issues. Speaking relatively, I don't think it matters in terms of the respondents how we phrase the questions: we will generally get, "me, too, per editor so-and-so responses". That is why we are where we are. WAID, you are seeing this through your own lenses, as an editor who carefully reads and contemplates issues. Most RFC respondents won't do that as you do. It is a timesink to try to figure out how to phrase a response that will generate contemplative responses; we need to phrase an RFC that will generate something useful to the closing admin knowing that we will get irresponsible responses from people who don't read or digest or contemplate the problem. I am not sure those two sets of potential questions intersect. SandyGeorgia (Talk) 11:45, 29 December 2019 (UTC)

I shall think about it. I fully agree with your Wikipedia:Polling is not a substitute for discussion link and am not convinced voting is going to bring sufficient light here for all the reasons wiki has documented about why it is evil. The "It is, however, frequently possible..." sentence really isn't true. I think part of the problem has been upside-down thinking about our sources. We have an article topic, and what appears to be a database of thousands of product prices, and we type the topic into the search box. We get the price of a random sized pill in north east Democratic Republic of Congo in 2015. Or we get the price of a 20ml cream paid by the government of Sudan in 2014. It really isn't "frequently possible" to source drug prices at that level of precision for any given country or any given supplier/manufacturer. The MSH database is really way too sparse to be generally useful, yet I have seen it used desperately for everything including sanitising hand rub and chlorine bleach, which are both not even pharmaceutical supplies. The Medicaid database has average prices, not prices for a given manufacturer, and the drug descriptions can be so terse it isn't always easy to know what formulation is included. There is a suspiciously low range of pill sizes in that database compared to e.g. BNF or Drugs.com, making me suspect it only includes common ones or, worse, only the ones it got price data for. The BNF is almost the opposite in richness of price information, but we do need to bear in mind it is only concerned with drugs a doctor can prescribe on the NHS and are available via pharmacies. The wholesale price of drugs/products that are (also) available over-the-counter at a pharmacy or supermarket cannot be determined by citing the BNF. The BNF also fails to point out when drug prices are increased due to availability issues, or massively discounted due to (often confidential) agreements with the NHS on expensive new drugs. Also "frequently possible to source a statement " tends to suggest that doing so is actually a valid thing that makes sense and doesn't break policy. The 200+ prices for Vitamin D3 would suggest it really isn't "frequently possible" at all. So while we can get some arbitrary prices for mostly random countries or suppliers, it think it is dangerous to suggest that could reliably form a "statement" rather than merely form a piece of data that an editor may then wonder what to do with.

The "Much of it is outdated or otherwise incorrect." is really "Nearly all of it is incorrect and misleading, and most of it is many years out of date, often citing a source that stopped being updated in 2015". It would greatly simplify any RFC if we could simply get prior agreement that the MSH database is not fit for our purpose. It was fine for the 18-50 products that WHO/HAI studied in their global price analysis projects a decade ago, but pretty useless otherwise.

I think the "WHO/HAI affordability" has, with all good intentions, been rather over-egged as a solution. It is a historical project from 10-15 years ago. Look at the database of prices and expand all the countries. You get random dates of when a survey was done, mostly from 2001 to 2008 but a handful as recent as 2015. I wonder if it is not a coincidence that the MSH database stopped being updated in 2015. If you expand the Affordability tab to see all the drugs they have surveyed, you will find just 50 medicines (some in a few formulations/strengths). And if you consult the table of results, you see what we have noted previously, that many popular drugs are simply not available through the government health system and can only be purchased privately at extortionate retail prices. Those prices bear no relationship other than a tenuous much-greater-than the wholesale price. It is hard to understate how misleading and wrong our "developing world" prices are.

I agree with Sandy, that we must be careful not to claim one editor with strong views and who added nearly every single drug price to over 500 articles, is "some editors". We are here because that editor persistently resisted challenges to the text added, and because WP:MED failed to intervene in any meaningful way. WhatamIdoing, I have every confidence you can think wisely about the sources and could, if so inclined, add some excellent price information where and when it is justified to do so. When Sandy says "I think we already know how to clean up these articles" that means those currently participating here and it is very obvious who are not participating. Is this really a content problem we don't already know the answer to and need to ask the community, or actually a user problem that should be dealt with another way? -- Colin°^Talk 12:28, 29 December 2019 (UTC)

We are here because too many editors are not like WAID, and because in WPMED discussions that led to these impasses, the idea that Wikipedia:Polling is not a substitute for discussion was not in evidence. I appreciate WAID's care and consideration in crafting text and responses, but I fear we may be missing the obvious underlying tensions when we expect editors will engage extended commentary or discussion aimed at developing sound consensus. They won't (hence my reference to the table of despair). SandyGeorgia (Talk) 13:57, 29 December 2019 (UTC)

I think we are here because editors at ANI told us to start an RFC about "the question of drug pricing". My preference is to write the RFC in a way that results in improving this guideline, but SandyGeorgia's alternative (below) is also responsive to that direction. WhatamIdoing (talk) 21:41, 29 December 2019 (UTC)

I think it is useful to have different editors write different explanations, although there also will need to be some general discussion of it here before the RfC goes live, in order to have a consensus that views are being presented fairly, as opposed to being lopsided. But I think we have to be extremely careful of tl;dr. I'm not sold on the idea that the solution to making "how much weight" clear is to write a lengthy introduction to explain what we are trying to ask. Instead, I think it needs to be as clear as possible what responding editors are supporting or opposing. If we do have different editors each presenting different views, I could see an RfC where we do that, but without identifying views by editor names, and ask the community to support or oppose those views. It really looks to me like we have essentially two conflicting views that were laid out at ANI, one favoring widespread presentation of pricing, and one insisting on caution instead. In that way of looking at it, we really are not asking how much weight to put on prices, so much as how widely prices should be included. --Tryptofish (talk) 23:15, 29 December 2019 (UTC)

Start over again

SG's attempt at a new start over, abandon hope all ye who enter here

IMO, the reason we are unable to formulate an RFC question or questions is that the task we are undertaking is the wrong one: we are attempting to formulate a general RFC to address what is in fact a very specific dilemma relating to very specific databases, when the answers to our sourcing and content dilemma are already addressed by policy. We have no other example anywhere, after weeks of discussion, of any other instances of drug pricing in articles presenting a problem. There has been no problem except the database-style sourcing. We are attempting to generate questions that will get respondents to read and respond to what is (should be) a policy question, but we will get "because I like it" responses that will give us nothing useful as result. That is why we are here. Everyone who has participated in this discussion knows how to add price data according to WP:V, WP:NOR, WP:WEIGHT, WP:NOT and WP:LEAD. Why don't we forget all the general questions we are trying to ask to solve a non-existent general problem, and instead just get straight to the specific problem? Put up one example of database-sourced text (I have repeatedly asked the database advocates to give us the strongest example, and none has been produced) and simply ask if this text is supported according to policy, V, NOR, NOT, WEIGHT, LEAD. Then each respondent will lay out arguments of why it does or does not breach each policy. We are spinning our wheels trying to solve a non-existent problem, as if this has been a generalized problem across all drug articles. We have one problem only; over 500 articles using a database to source text. SandyGeorgia (Talk) 12:57, 29 December 2019 (UTC)

• H

  Do these examples of pharmaceutical drug prices comply with Wikipedia policies on verifiability, due weight, no original research, and what Wikipedia is not? If so, should this text be inserted into leads of articles?

• Ethosuximide: The wholesale cost in the developing world is about US$27.77 per month as of 2014.[80]
• Carbamazepine: The wholesale cost in the developing world is about US$0.07 to US$0.24 per day as of 2015.[81]
• Mebendazole: The wholesale cost in the developing world is between USD 0.004 and 0.04 per dose.[82]

But the examples need to be varied to included the other drug databases, and the other kinds of problems presented; I pulled these samples from one section above, but the three of them were to demonstrate one issue. The other kinds of examples should be give in place of two of these. SandyGeorgia (Talk) 14:23, 29 December 2019 (UTC)

Multiple examples of the same problem are useful, in that they demonstrate that the problem is not isolated. If we assume that Wikipedia:Nobody reads the directions, then it might be more effective to provide an explanation of the source in the RFC question, like this:

Ethosuximide:

What the lead says: The wholesale cost in the developing world is about US$27.77 per month as of 2014.[83]

What the source says: One organization said that they sold 250 mg tablets for US$0.1845 each (100 tablets per package). This organization only sells drugs only to government-recognized healthcare organizations in the Democratic Republic of Congo. The defined daily dose (a complex statistical concept; not necessarily the dose any person takes) is 1.25 grams.

I don't think that "if so, should it be in the lead?" is necessary. WhatamIdoing (talk) 21:55, 29 December 2019 (UTC)

I like the general approach here. I think it's much better than trying to ask a single question. What we could, in effect, do is to ask the community: do you support doing it this way, or that way? If we work on making something along those lines as clear and concise as possible, I think that would be the path to getting an RfC that results in an outcome that actually means something. --Tryptofish (talk) 23:18, 29 December 2019 (UTC)

I think this is a decent formulation, though we are presuming all respondents will both check the sources, and know what they are and how to interpret data from them. I'm not certain that will be the case, especially if the RfC is widely publicized. Seraphimblade ^{Talk to me} 03:48, 30 December 2019 (UTC)

Could a neutrally written background accomplish some of this? Best, Barkeep49 (talk) 05:12, 30 December 2019 (UTC)

Yes, if everyone could come to agreement on how to describe what they are, how they gather data, etc., that could be very useful knowledge for those who comment in the RfC. Seraphimblade ^{Talk to me} 07:06, 30 December 2019 (UTC)

This concept of how to address ask the RfC seems to have some traction. Pinging participants from the past week: Colin, Kashmiri, Pbsouthwood, CFCF, and Ronz. Best, Barkeep49 (talk) 05:12, 30 December 2019 (UTC)

I had wondered a similar thing to Sandy's idea: that we take each of the four(?) price-database sources and look at how they are being used and then together reach an agreement about the problems with them, what could be said using them, and what shouldn't be said using them. I actually remain optimistic that much of that analysis/discussion/conclusion could simply be done if we get a good-faith article-experienced bunch of editors to simply work together. All the previous discussion on specific problems felt like it was only me and James and that didn't work for various reasons. I'm not rejected the idea of an RFC, but it would be great to clear away a lot of the crap first, and there really is an awful lot of pretty straightforward crap we could eliminate IMO quite quickly, and focus an RFC on asking the community about price statements that are actually source->text honest and policy-legal. I would be much more confident that such an RFC would be successful in its goals (and to be honest, presenting all the awful prices to the entire community right now would IMO seriously dent WP:MED's credibility).

A mix of family priorities, the latest Star Wars film, and going back to work mean I've not yet finished some of stuff I was working on that demonstrate current text and current problems. I do hope I get a chance to put them up very soon. I think then it will be good if we can all see an honest full selection of drug prices in articles (rather than anyone accuse of picking a hard/easy example) and also quite a number of drugs where the price is genuinely notable and editors could make a really good effort to polish some excellent guideline text on how to present that to readers.

Barkeep49, could we have some kind of moderated workshop to focus on e.g. one source at a time. Advertised to the community. We want participation from wise owls or diligent investigators, rather than just ask for a big mob to vote all at once. I don't think we require medical experts at all, so please nobody rule themselves out on that regard. This is really straightforward source->text analysis. The only kind of ability I can see being useful is an appreciation of statistics to the degree that one can't make general statements from few data-points and to identify the weaknesses in one's data. But that is elementary stuff. We could do this workshop on this page even, and do one source at a time. I would certainly like all the current participants to help, but there is one obvious name who is not currently engaging in discussions, and who's participation is essential. --Colin°^Talk 10:06, 30 December 2019 (UTC)

I think this idea has potential, and I am willing to invest some time in it if and where I think my contributions will be helpful. If we are looking for a moderator, I propose Barkeep49. · · · Peter Southwood ^(talk): 16:15, 30 December 2019 (UTC)

Despair recedes ! SandyGeorgia (Talk) 16:20, 30 December 2019 (UTC)

But, as we present the different examples from the different databases, we should keep in mind that there is another outstanding issue: in spite of all of this discussion, we still have editors contemplating adding goodrx.com prices to infoboxes. SandyGeorgia (Talk) 16:27, 30 December 2019 (UTC)

One can only deal with so much madness at any one time :-). --Colin°^Talk 17:24, 30 December 2019 (UTC)

Adding prices to the infobox falls short of the community endorsed prohibition on adding or removing this information and you can feel free to point editors to that discussion in removing the prices/costs. Best, Barkeep49 (talk) 20:19, 30 December 2019 (UTC)

I, too, am feeling increasingly optimistic about this direction that the discussion is taking. One aspect where my own understanding is rather poor, and it would be helpful to me to better understand, is one of how many different competing opinions there are. In other words, are there really one group of editors who like Source A, but not Sources B, C, and D, a second group who like Sources A and B, but not C and D, a third group who like Sources A, B, and C, but not D, and so on? It kind of looks to me (but please correct me if I'm wrong) that it's not like that. It looks to me more like there are basically two "camps" that were in conflict at the ANI discussion: one that likes several of those sources and likes widespread presentation of drug prices, and another that opposes the sources that the first "camp" likes on the basis of things like NOR and opposes widespread presentation. So if – if – I understand that correctly, we might not need to evaluate Sources A, B, C, and D individually, but instead treat them as a group for RfC purposes. --Tryptofish (talk) 23:18, 30 December 2019 (UTC)

It's not useful to try to determine consensus based on "Me, too" or "BecauseILikeIt" declarations at ANI or anywhere else, and it's that sort of !voting rather than discussing that has plagued WPMED discussions. The answer to your question is elusive, because no one who may support these database prices has fully engaged this discussion. How are we to know if that means there is no defense of the sourcing used, or something else? The RFC needs to show sufficient examples and be phrased in a way that the closing admin can judge responses that engage policy (or not). SandyGeorgia (Talk) 23:29, 30 December 2019 (UTC)

Thanks, that makes sense. Would it be correct to say, conversely, that there are editors who would consider an entire group of such sources inadequate? --Tryptofish (talk) 23:33, 30 December 2019 (UTC)

I can only speak for myself. I see violations of V, NOT, NOR, WEIGHT or LEAD in every example given. I have asked for a best example of how to use these databases; none surfaced. SandyGeorgia (Talk) 23:36, 30 December 2019 (UTC)

Actually, that's not entirely true. WAID gave hypothetical, "if we believe this", "then we'll do this" examples. SandyGeorgia (Talk) 23:38, 30 December 2019 (UTC)

Tryptofish, pretty much all the drug-database-sources->wiki-text have fundamental original research and/or weight and/or competence and/or source-honesty issues. I don't think good-faith impartial editors should find any of that contentious. It is possible that a workshop can identify if there are any limited ways those sources could be used properly. The conflict stemmed from a lack of follow-through from "you can't do this because it has X/Y/Z wrong" -> "ok, I won't do that". Is there actually any camp that still claims a database result-set listing no suppliers whatsoever for a product should be used to make claims about "the developing world"? Is there actually any camp that still claims that a source that gives prices for many formulations and specific pill strengths but does not give any therapeutic dose or recommended tablet strength or treatment duration cannot be used to make claims about "monthly cost" or "cost of treatment"? Is there any camp that claims Wikipedia should continue to use a source that stopped being updated in 2015? Etc, etc. Camp B seems to have gone rather quiet. -- Colin°^Talk 23:46, 30 December 2019 (UTC)

(edit conflict) As you can probably infer, where I'm going with this is that I'm looking for ways to simplify the RfC structure (without ruining it, of course). Would it be appropriate to render this as: one perspective is that the sources all flunk policy, and the other perspective is that the sources can be OK subject to certain caveats? Adding after ec: so from what Colin says, that may perhaps actually be the case. I've become increasingly aware of that lack of responsiveness during my brief involvement here. In one fish's opinion, if editors don't engage, they are entitled to one last alert before the RfC goes live, but if they still don't respond, then they lose out on determining the parameters of the RfC. --Tryptofish (talk) 23:55, 30 December 2019 (UTC)

All of those suggestions seem to be leading. The problem is how to strike a balance between my "Is it in sync" (which got responses that didn't engage at all, rather "Me, too'd", and only later read each point to change their position), which attempted to not lead at all, and providing more text without leading?

As to editors who aren't engaging an RFC formulation they know is underway, and have forbidden pings and talk page posts, I'm happy leaving notification decisions to Barkeep. SandyGeorgia (Talk) 00:02, 31 December 2019 (UTC)

Hmmm... I'm just thinking out loud here, but: It sounds to me like it would be pretty feasible for the editors who are participating actively here to articulate arguments against all of the sources in question. The dilemma is how to formulate the RfC with respect to perspectives that support some or all of the sources. I'd like to avoid making the RfC creation process more difficult than it needs to be. --Tryptofish (talk) 00:24, 31 December 2019 (UTC)

I've been regularly pinging people who appear to be engaged with the process while respecting those who, for whatever reason, have stopped participating. I think to Trypto's point we're close to being ready for a ping to WT:MED that could invite people who haven't been participating to join in before the RfC. Ultimately if editors who have a certain position choose not to help formulate the RfC they lose out on the chance to structure the debate but obviously not the chance to weigh-in during the RfC. I do think we the end RfC will be better if editors who end up with contrasting thoughts at the RfC help formulate it, but as a volunteer project we move forward with the volunteers who are willing to spend the time. Best, Barkeep49 (talk) 00:38, 31 December 2019 (UTC)

I agree, and I think a message to all interested editors at WT:MED (perhaps what you meant by a "ping") is entirely appropriate. --Tryptofish (talk) 23:59, 31 December 2019 (UTC)

I think we can simplify SandyGeorgia's proposed question:

I

These examples of pharmaceutical drug prices have been taken from the leads of some articles. Do they comply with Wikipedia policies on verifiability, due weight, no original research, and what Wikipedia is not?

It could have a line about "If you think this needs to be improved, then what changes would you advise?" Editors could then provide their specific opinions (like explaining the contents better, moving it out of the lead, using it as an external link, not including prices at all, or whatever else they think).

On a related point, I'm not sure that addressing "all of the sources in question" in a single RFC is a good idea. The inherent problems with the NADAC database are fewer than the inherent problems with the MSH database. I think we are better off asking about one set of problems, and then having another RFC on another source later (if necessary). WhatamIdoing (talk) 01:02, 31 December 2019 (UTC)

I like it. <oops ... yes, I said that>. SandyGeorgia (Talk) 01:50, 31 December 2019 (UTC)

I like the idea of presenting the community with specific examples of article content, and asking whether they approve or disapprove. I think specific examples are much better than asking about generalizations. I'm weighing, however, how best to pose such a question. A bit higher up in this talk section, Sandy quite rightly pointed out the issue of not making any aspect of the RfC a "leading question". With respect to I, I could imagine editors coming to the RfC would surmise that we would never open an RfC asking about policy compliance of selected examples from leads unless there were a significant concern about policy violation. Another point that occurs to me is that I think we want to get community feedback about how widely drug prices should be included on pages, and it seems to me that any consensus derived from this question would only be applicable to lead sections. The "what changes would you advise?" approach could also lead to a very large number of proposed changes without leading to a consensus about which changes are the best. --Tryptofish (talk) 00:17, 1 January 2020 (UTC)

I would like a bit more time to prepare some data pages like I've linked below. There's also a couple of wiki articles on price-related topics that still need a bit of work. I'm keen that we present real data rather than appear to cherry pick. As Sandy notes below, for many of the sources, other editors will not easily grasp how on earth the price was derived from the source. It has taken a while for us to figure this out (while picking jaws up from the floor at the amount of original research and arbitrary choices made). WhatamIdoing you ask "Do they comply...."? Is there anything about this that the group here currently discussing prices disagree on or don't know the answer to? I still think a workshop could be a better approach to tackle and resolve the basic stuff that isn't opinions about what is or isn't encyclopaedic or what does or does not belong in a lead. Those are questions we could ask an RFC, but there are basic mistakes with all the texts & sources that really we don't need to ask the community. Do we? Having an RFC to resolve issues with what one editor alone has written and defends is a bit weird if you ask me. If you ask the above question, it is admitting that WP:MED has no competence to discuss, reach consensus and write honest source-based facts about drug prices. That it hasn't a clue whether the prices in the articles are good or bad and needs some help from Pokemon editors and Historical Fiction writers. And while I think that has been true, that WP:MED has been incompetent here and has failed to address or even examine the problem over many years, it doesn't need to be. -- Colin°^Talk 14:24, 31 December 2019 (UTC)

Colin, how I wish that this all could have been resolved through discussion. However, past attempts at discussion left you frustrated because a number of editors showed up to disagree with you without engaging you on the merits of what you were saying. And then the discussion to the extent it did happen became heated - in part because of comments like "Having an RFC to resolve issues with what one editor alone has written and defends is a bit weird if you ask me." which takes a jab at an editor (who while unnamed is clear to us who've been following this) and which could have been omitted without diminishing the larger point. The broader community decided that the way forward was through the RfC process, a way of gaining binding consensus. Any consensus reached without an RfC will need to include the consent of editors not currently participating in the discussion but who are invested in the outcome. Unless those editors choose, voluntarily because this is Wikipedia and we all have options afforded to us by being volunteers, to agree to that consensus then it will need to go to RfC. I wouldn't say that the chances are 0 of finding consensus without an RfC but they are slim. And if we can't get to an RfC then the only option is to focus on the behavior issues first (through ArbCom) in hopes that this then creates an atmosphere where consensus can be found on the content question. And even then we still might end up with an RfC. Best, Barkeep49 (talk) 17:43, 31 December 2019 (UTC)

Barkeep49, because we have to so carefully choose our language here, I want to be certain I am understanding. Are you saying that, if the editor(s) who made all of these edits were to voluntarily withdraw them (although that case appears unlikely), then an RFC could be avoided? Or are you saying an RFC has to proceed regardless if any pings to editors not participating result in a chance in stance? The time we are spending on this (and related) is what continues the despair. Regardless of the outcome of either RFC, I am unsure we will have changed the overall picture with all of this effort; I would rather be improving content (eg Epipen). SandyGeorgia (Talk) 17:50, 31 December 2019 (UTC)

SandyGeorgia, if the editors who are members of the WikiMed Project and who have, in a variety of venues, supported inclusion of prices/costs where you (and others) have not, agree to standards about where and how to include prices/costs then no RfC would be needed. This could happen even if they don't voluntarily withdraw them - though if they were to voluntarily withdraw them that would be an even stronger indicator of support for the consensus (but is as you point out not likely). I think it unlikely that that consensus can be reached without an RfC. Not impossible but unlikely. And hopefully, at the end of this, we have standards that let us focus on improving that content because I agree that improving content is multiple more rewarding than this :). Best, Barkeep49 (talk) 18:12, 31 December 2019 (UTC)

Thanks for the clarification, which clears up my uncertainty. SandyGeorgia (Talk) 18:16, 31 December 2019 (UTC)

My view on the "Do we hafta?" question is even more restrictive than Barkeep's. ANI said that nobody gets to touch that content until there's been an RFC. Fine, let's have an RFC (or several). The fastest way through that restriction is probably having an RFC that says "Is this stuff okay, or not?" Fine details, like what ought to go in MEDMOS, can be handled later. WhatamIdoing (talk) 23:39, 31 December 2019 (UTC)

I've posted a reminder of this discussion at WT:MED#Plans for RfC about drug pricing. --Tryptofish (talk) 22:12, 3 January 2020 (UTC)

Existing Prices

I have created User:Colin/ExistingPrices that is an automated extract of drug prices from drug articles. I got the list of drug articles by looking for external links to the MSH price guide, the Drugs.com price pages, the Medicaid NADAC pages, or referred to the BNF. It isn't all the drugs, but it 530 is good sized sample. I then extracted lines containing the word "price" or "cost" and did a bit of hand-editing on the result.-- Colin°^Talk 17:24, 30 December 2019 (UTC)

User:Colin, would you please change that to put a plain <references /> tag at the end of each ===Example===? I'm pretty sure that a simple regex find-and-replace across the page would do it, but I wasn't able to figure out the right combination. The main ref tag was updated to auto-limit itself to only the stuff in between the current one and the previous one, so that will get us the refs in each section (so people are more likely to look at them) without exceeding the template limits. WhatamIdoing (talk) 20:01, 30 December 2019 (UTC)

WhatamIdoing, with Colin's indulgence (and at the risk of taxing my poor Mac with the copy and pasting I did), I have  Done this. Best, Barkeep49 (talk) 20:35, 30 December 2019 (UTC)

Thanks Barkeep49/WhatamIdoing. I didn't know you could have multiple reference groupings. -- Colin°^Talk 23:14, 30 December 2019 (UTC)

The mw:Editing team worked on Cite.php a couple of years ago. This was one of the happy results. Another is automatic display in 30em columns (at wikis where this is enabled, including this one), although that somehow broke last week (and will remain so until next WP:THURSDAY). WhatamIdoing (talk) 00:50, 31 December 2019 (UTC)

I have also created User:Colin/MSHData which lists every MSH Price Guide reference along with the data year cited and the number of suppliers and buyers. In the WHO/HAI price survey methodology, how representative reference prices are depends on the number of suppliers quoting for each product. Because of this, they focus on a small set (14 or 18 products) that have good supplier data. We can see that 30% of our drug citations have no suppliers at all, yet we claim a price in "the developing world". A further 28% only have one or two suppliers, which makes the claim to be representative of "the developing world" a tenuous one. The majority (58%) of our MSH citations for "the developing world" refer to fewer than three suppliers. While some suppliers are international in scope, many target a single country or even just one part of a country. Nearly all (92%) of our MSH prices are from 2014, five years ago. The remainder are from 2015. The guide used to be updated annually but has not been updated since 2015. -- Colin°^Talk 23:14, 30 December 2019 (UTC)

Colin, let's assume everyone coming to the RFC is going to access this data, and do what I did: check meds they know. I checked the most commonly prescribed at the clinic where I volunteer. First, readers trying to understand this data are going to get incredibly frustrated, because it is so hard to find what one is looking for in those sources, and then when one figures it out, it's unbelievable.

• Levothyroxine claims "In the United States, a typical month of treatment costs less than US$25" based on an old hard-print source I can't access. That information is not very useful, so sure does not belong in a lead, but this is crazy. Levothyroxine is on every $4/30-day and $10/90-day list. Walmart sample. Everyone gets 90 days (you take it for life, why get 30?). It's $3.33 per month ... <sheesh>. Inserted into the lead three years ago.

• Chlorthalidone claims "In the United States the wholesale cost is about US$13.50 a month", based on NADAC. So, the average reader goes to that source, finds a bunch of gibberish, and finally figures out that some real person, seriously, took the wholesale cost for a 25-mg tablet and multiplied the 0.45 cents per tablet x 30 for a monthly supply.[84] But the standard dose for chlorthalidone for high blood pressure is 12.5 mg. Since it only comes in 25 mg tablets, it has to be split. How much of this kind of error do we have? Do we have no drug editors reviewing this data? The wholesale cost of how many days to treat what? It doesn't even say. And, by the way, since I assume we are not supposed to mess with price data before the RFC, we are supposed to leave this error in the lead for how long?

• Lisinopril gives me "In the United States the wholesale cost per month was less than 0.70 USD as of 2018", based on NADAC. I can't figure out why that information is useful or belongs in the lead. Basically, lisinopril is a generic that nobody pays anything for anywhere. Yes, the wholesale cost is extremely low, and pharmacies give it away. So our information here isn't blatantly wrong, just not at all useful.

So, that's my sample. Number one and number two top prescribed meds in the US. This is wacky. With the extent of the problems you have raised on this page, how are we going to choose which drugs to use as meaningful and representative examples to put forward in the RFC? Seeing this amount of gibberish in leads of articles, I am becoming convinced again that we have to deal with the lead problem. SandyGeorgia (Talk) 02:43, 31 December 2019 (UTC)

And I'm afraid I may be going backwards on my suggestion that we only need to deal with the databases. One of the two errors above (levothyroxine) is not due to a database source: it is from a hard-print source. I may be coming back to we have a problem bigger than the databases. We have sources BEYOND these databases being used to insert formulaic undue information into leads. SandyGeorgia (Talk) 03:20, 31 December 2019 (UTC)

Sandy the prices from the Tarascon Pocket Pharmacopoeia are based on a $, $$, $$$, $$$, $$$$$ pricing symbol much like your holiday guidebook indicates if a restaurant is a cheap-eat or an expensive night out. I complained about it earlier at WT:MED. Here's what they the book says about its symbols: (the underline italics is theirs)

RELATIVE COST

• Code / Cost
• $ = < $25
• $$ = 25 to $49
• $$$ = $50 to $99
• $$$$ = $100 to $199
• $$$$$ = >= $200

Cost codes used are "per month" of maintenance therapy (e.g. antihypertensives) or "per course" of short-term therapy (e.g., antibiotics). Codes are calculated using average wholesale prices (at press time in US dollars) for the most common indication and route of each drug at a typical adult dosage. For maintenance therapy, costs are calculated based on a 30-day supply or the quantity that might typically be used in a given month. For short-term therapy (e.g., 10 days or less), costs are calculated on a single treatment course. When multiple forms are available (e.g., generics) these codes reflect the least expensive generally available product. When drugs don't neatly fit in to the classification scheme above, we have assigned codes based upon the relative cost of other similar drugs. These codes should be used as a rough guide only, as (1) they reflect cost, not charges, (2) pricing often varies substantially from location to location and time to time, and (3) MHOs, Medicaid, and buying groups often negotiate quite different pricing. Check with your local pharmacy if you have any questions.

So whenever you see "is inexpensive" or "under $25" that came from a "$". If you see "between $25 and $50" that came from "$$". And so on. If you see "more than $200" that came from $$$$$. So even if the drug costs an eye-popping $9000 a dose, we'll just say "more than $200". And as you point out, the majority of drugs are "under $25" even if actually they are just a few dollars. None of the values 25, 50, 100, 150 and 200 appear in the source-data for the wholesale price of those drugs. Those values are all artefacts of Tarascon's price grouping into $ symbols. Our readers couldn't give a damn about Tarascon's price grouping. Reverse-engineering a $ into "less than $25" is a heinous crime. Saying a thousand-dollar drug is "more than $200" is a heinous crime. -- Colin°^Talk 13:58, 31 December 2019 (UTC)

Lovely. (I unfollowed WT:MED because of the bullying; sorry I missed that.) So, we're using a source to insert UNDUE information into leads that is good for nothing. Walmart is a plague that has infiltrated every part of the US, and at Walmart, you get levothyroxine for $3.33 per month. If you're going to a different pharmacy than Walmart, most likely, your insurance is covering the drugs (I just picked up four prescriptions for my household, and paid a Big Fat $0 because we chose the right insurance).

In thinking about how this relates to or affects the general question we've posed, I decided to look at the other med we deal with most commonly in the clinic where I am an interpreter. I was not as familiar with metformin pricing, because we are often giving away free samples. (Actually, many patients have their family send metformin from Mexico.) So, I decided to check that one (diabetes). Wikipedia has:

• In the United States, it costs US$5 to US$25 per month. That is sourced to drugs.com.

I can't find those numbers in there; perhaps it is similar to the restaurant guide above. At any rate, fourth most commonly prescribed drug in the US, and it is also on the Walmart $10 for 90-day list. In fact, I believe Walmart may have $24 for 180 days of metformin. We have useless, UNDUE, inaccurate information in leads.

What remains astounding about this formulaic editing is that, in one demonstrable case where we SHOULD have information in the lead about price (epipen), there is NONE. We could give that as an example, but we can't edit prices right now. (If we decide we need that as a good counter-example, we can do a mock-up.)

In re-thinking how all of this impacts the formulation of the RFC questions, I am coming back to the lead problem, which must be dealt with.

J

These examples of pharmaceutical drug prices have been taken from the leads of some articles. Do they comply with Wikipedia policies on verifiability, due weight, no original research, and what Wikipedia is not, and do they reflect the guideline on leads?

We have to deal with the lead problem; that is where WPMED editing is focused these days, and that is where we are consistently finding problems. Alternately, I could be convinced that we could leave LEAD out of the RFC question, so as not to dilute the policy questions, but the fact that we are finding bogus drug price info in leads will be noted by those who bother to read and they can comment that it doesn't belong there. Open to ideas between I and J. SandyGeorgia (Talk) 14:31, 31 December 2019 (UTC)

Firstly the Drugs.com links mostly don't contain prices on the pages being linked-to. I assume Drugs.com moved their site content about a bit since the first link. Either that, or the editor was citing Drugs.com for other info and couldn't be bothered creating a separate link for the price info. If you look down the page, you'll see a "Pricing & Coupons" link. Alternatively, just Google "NameOfDrug Drugs.com prices" and that should find it. Metformin prices are here. The article also gives a developing world price of "between US$0.21 and $5.55 per month as of 2014" and links to the 500mg tablet. The MSH site says the Defined daily dose is 2g (DDD is what is being used for "typical daily therapeutic dose" even though WHO says it is not a typical therapeutic dose and should not be claimed to be such). Drugs.com agrees that a typical adult dose is 2g taken in divided doses. The $0.21 is clearly coming from the 0.0070 unit-price from cheapest supplier IMRES (which is an international supplier) multiplied by 30 -- so this is actually only one 500m tablet a day not four. I can't make $5.55 no matter how I try. The dearest price is 0.0372 unit-price from supplier MEDEOR/TZ (which you might guess, supplies just Tanzania) and would be $1.12 a month for 30 or $4.46 if we take four-a-day like the doctor ordered. I have absolutely no idea how to get "$5 to $25 per month" from Drugs.com, it seems to depend how much you order and even then there can be a huge price range. Metformin, according to Wikipedia, is the fourth-most prescribed medication in the United States, so you might hope WP:MED would care if the price is correct or even makes any kind of sense.

You can see from my Existing Prices page that most of the specific price information is only in the lead, and occasionally copied in the body. I think adding "should it be in the lead" to the RFC at this stage is premature, since "Is this all a steaming pile of ____?" is more immediately relevant. -- Colin°^Talk

I found Epipen info at Epinephrine (medication), which claims that "In the United States, the cost of the most commonly used autoinjector for anaphylaxis was about US$600 for two in 2016, while a generic version was about US$140 for two." We paid $280 for 2 last year, so I don't even know where to start. But the controversy about Epipen pricing is not even addressed in that lead, and here we have one instance of where it should be. We have formulaic insertion of dubious price data, with no regard to how a lead should be contructed. SandyGeorgia (Talk) 15:42, 31 December 2019 (UTC)

I need to dig up all the articles I previously have mentioned...

Pyrimethamine (edit | talk | history | protect | delete | links | watch | logs | views) is another example where there could be a great deal in the lede about pricing, but instead there is just the basic price information as found elsewhere. --Ronz (talk) 17:27, 31 December 2019 (UTC)

Draft:List of drug prices has an interesting lede, and appears to have entries not mentioned prior. --Ronz (talk) 17:34, 31 December 2019 (UTC)

I wish I could unsee that. When despair sets in, the healthiest thing I can do is go volunteer at the local food shelter. Every time I think we are getting a handle on this ... SandyGeorgia (Talk) 17:38, 31 December 2019 (UTC)

Draft:List of drug prices has entries that appear to have been written by other editors. --Ronz (talk) 17:58, 31 December 2019 (UTC)

I was just asked about that on my user talk (which is I suspect where Ronz saw it too). If an editor wants to invest time in something that might ultimately lead nowhere well that's their choice. It's even possible that the work helps move this conversation forward by providing "real life" examples that can be used in the RfC, or in this discussion leading to the RfC. This page needs to be where the conversation continues collaboratively but like with work Trypto and Colin have done doesn't mean work can't continue on the side also. Courtesy ping to QuackGuru. Best, Barkeep49 (talk) 18:00, 31 December 2019 (UTC)

Barkeep49, I could have this wrong, but please check the timestamps closely re "which is I suspect where Ronz saw it too". The timestamps give the appearance that your courtesy ping may have been redundant. SandyGeorgia (Talk) 18:03, 31 December 2019 (UTC)

Nope you're correct. My apologies. I've struck that comment. Best, Barkeep49 (talk) 18:06, 31 December 2019 (UTC)

As I thought. QuackGuro, since you seem to be following this discussion so closely that you posted to Barkeep's talk within moments of Ronz raising this issue, I would point out that your participation here could be useful. This discussion has revealed extensive instances of dubious price information in drug articles; your opinions on the topics might help us advance towards consensus. "Student editing" has not been shown to be a way to produce good outcomes, btw. SandyGeorgia (Talk) 18:18, 31 December 2019 (UTC)

I've been searching for other related articles, and came upon the draft earlier. It's an example of someone else's work on the topic.

For another example, here's an entry to the body of an article written by a SPA: [85] --Ronz (talk) 18:20, 31 December 2019 (UTC)

I'll probably create a workspace as I search articles, but I've found an example of (apparently) good price info in the lede, added by Nbauman (talk · contribs) [86] --Ronz (talk) 21:11, 31 December 2019 (UTC)

Remember that Abatacept the price statement is 4 years old. The lack of any community maintenance/upgrade/fact-check of these prices is a major issue. Also note that in the UK the price is confidential (in 2013) no matter what number appears officially in the BNF. In the UK, the powder for infusion is a tiny fraction of the price of the pre-filled syringes for injection. Getting the right price for this could be a challenge or impossible. -- Colin°^Talk 21:34, 31 December 2019 (UTC)

Yes, and it's a far better example, a far better source, and clear why the editor thought it met NOT. --Ronz (talk) 02:07, 1 January 2020 (UTC)

Status check

• We've more or less settled into a question that is approximately like H/I/J.
• We've got a few examples that we can explain in detail. (These could be posted upfront, or they could be part of individual editors' personal responses.)
• We've got pages (which we can link) that list hundreds of examples of current content.

What else needs to be done before the RFC is officially launched (other than waiting a couple more days)? WhatamIdoing (talk) 00:03, 1 January 2020 (UTC)

I guess it depends on how one interprets "approximately like", but I'm not sold on the idea that we have really settled onto anything. I'd say that the more recent suggestions are regarded as moving in the right direction, relative to the earlier ones, but I'm not satisfied that we are close enough to our destination. --Tryptofish (talk) 00:20, 1 January 2020 (UTC)

I'd like to see us write up the specific examples soon, so we can see exactly what we've got. And to be sure we have one example from each "database", the "restaurant guide" hard-print source, and one example not using this kind of sourcing. I'd also like to see a mock-up of the whole thing. Obviously, I'm gun-shy :) :) I certainly thought a broad, simple, one-question (is it in sync) was the right way to go because it wasn't leading, but it was only the right way to go to get "ilikeit" responses! I really want to see how this thing "looks" before we launch. That's why I keep saying we need to define the name of the page we will put it on, and start working there. SandyGeorgia (Talk) 00:36, 1 January 2020 (UTC)

Yes, I think the next step really should be to settle on that page name. --Tryptofish (talk) 00:49, 1 January 2020 (UTC)

Here are the ones I put up earlier ... we have evolved :)

1. RFC on NOPRICE and pharmaceutical drugs
2. RFC on pharmaceutical drug content
3. RFC on pharmaceutical drug pricing and sources
4. RFC on pharmaceutical drug prices and sources

SandyGeorgia (Talk) 00:53, 1 January 2020 (UTC)

As we have evolved, I like prefer #4. Have we decided if we still need a preamble discussing Ronz's price/pricing terminology distinction. I think we could do it in one or two sentences, and hope Ronz will do that. SandyGeorgia (Talk) 01:30, 1 January 2020 (UTC)

Maybe just RFC on pharmaceutical drug prices? WhatamIdoing (talk) 06:28, 1 January 2020 (UTC)

Fine with me. Happy 2020 to all! SandyGeorgia (Talk) 06:32, 1 January 2020 (UTC)

Likewise for me, thanks! --Tryptofish (talk) 21:35, 1 January 2020 (UTC)

Barkeep's Update

I had written a long update to the community spurred in equal measures by how close we are to the tipping point here (either towards success or towards a failure to formulate an RfC) and QuackGuru expressing a desire to appeal an aspect of the prohibition. Quack has now said they're not going to appeal so with half the justification for the update gone I undid it.

Barkeep's Summary at AN

A little over three weeks ago I closed a long and unsurprisingly acrimonious ANI thread relating to the behavior of several editors. The ANI discussion also had heavy elements of a content dispute around what should be or not be included in the Medicine Manual of Style page. The major finding was that an RfC was needed. In the time since I (as penance for closing that discussion) have been helping interested editors move towards an RfC. I am updating the community now both because I think we're going to be at a tipping point soon as to whether an RfC will ever get crafted (I'm hopeful but it's hardly a sure thing) and because an editor has expressed desire to appeal to the community for an exemption to one area of that close which I expect will be forthcoming soon. All are of course welcome to read the whole long discussion but here's my summary of major points since then: The prohibition on adding and removing pricing/costs information has largely held and worked without any editor being sanctioned. The initial efforts to formulate an RfC on the topic of price/costs, led by Guy quickly sprawled. Just over a week ago I closed other discussions and attempted to restart the discussion. Bad news: this new discussion is beginning to sprawl - and could reach a point where it becomes clear that no RfC formulation can be found with the current set of participants. Good news: there has also been general agreement around an RfC concept (labeled H/I/J on the page). Bad news about the good news: specific examples to include in the RfC still need to be found and agreed to. This could prove tricky in and of itself. A discussion about guidance for the lead of articles gained no participation and is now at its own RfC which is itself long and complex and I will not attempt to summarize here. There has been a collegial and collaborative atmosphere among those participating (with the only clear behavior line-crossing occurring at the LEAD RfC and quickly walked back by both editors when requested). Part of the reason for this is because few editors who are broadly in favor of pricing/cost information remain involved in the conversation. This is obviously their choice, but does present risks to whether the RfC will be successful even if it can be formulated. While all of the editors participating are incredibly skilled and knowledgeable about the topic – far more so than I – sometimes that understanding of complexity sprawls the discussion in interesting, relevant, and important to the topic ways but not necessarily ways that are helping lead to the RfC. I am hopeful that this update is helpful when considering the appeal that is about to come and in the interests of having some more uninvolved editors who can help move the RfC to launch. Best, Barkeep49 (talk) 01:08, 1 January 2020 (UTC)

If this is going to get to RfC the more focus we can have on the specific examples to be used the better. Naming is not unimportant but if we can't finish finding the examples for H/I/J the name won't matter. I suspect that this next week is going to be all the difference as to whether we will get the RfC launched and have it find a consensus (still my hope and well with-in our grasp) or not. Best, 01:32, 1 January 2020 (UTC)

Wrt examples, are we going to focus on one database/source or pick examples from several. If we pick one, then that should perhaps be part of the RFC page name, for we will certainly need other RFCs to tackle the other sources. If we pick one, then the MSH International Medical Products Price Guide seems an obvious first target. -- Colin°^Talk 11:22, 1 January 2020 (UTC)

Barkeep49 notes the lack of recent participation by those wanting prices. Indeed the last comment I can find was on the 16th December by Doc James, who suggested the RFC questions: "Sure we could start with "do you think the approximate price of a medication can be estimated for various regions of the world?" And "Should Wikipedia articles contain information about the cost of medications?". Furthermore I think it is relevant that when WP:NOTPRICES was quoted earlier, James claimed to agree with it, despite clearly disagreeing with others as to what it means. I note at Talk:Ivermectin#Price that Seraphimblade wrote "Pricing, per policy, is not permitted in articles, with rare exceptions when the price is a significant part of something's notability or is very extensively commented on (not just mentioned).". Ronz linked here saying "All discussion so far supports removal" and James yesterday replied "Ah lots of discussion supports keeping it.". James's two questions aren't a million miles away from my Question A: can we establish "the price" "for various regions of the world" and "should articles contain them" (though most of us want a "while obeying policy" included in that question). Rather than addressing that bigger question, we seem now to be focusing on chipping away at the prices, either by establishing each source shouldn't be used like it has been (and possibly discover if there is any way it can be used), or get community agreement that the current examples fail policy. Maybe that is the correct approach, but we need to be clear that the chipping-away RFCs are not asking the-big-question. Even if we conclude that our uses of sources A, B, C and D are all awful and should never have been permitted and must be swiftly removed, we still haven't resolved WP:NOTPRICES because some editors read it to mean one thing and others read it to mean another. -- Colin°^Talk 11:22, 1 January 2020 (UTC)

I think examples of content created by other editors are important to have, especially any that don't have the NOR, POV, and NOT problems. --Ronz (talk) 19:36, 1 January 2020 (UTC)

Dealing with a single source (MSH is used in more than half of the identified articles) means a simpler task for respondents. They tend to have the same set of issues, such as the question of how to using DDD to estimate a monthly cost and reporting a few data points as being representative of the entire developing world.

If you want a "good example", I could re-write one of the MSH and NADAC sources to (IMO) avoid NOR problems. We'd have to pick different drugs to deal with the DUE and NOT problems. WhatamIdoing (talk) 19:56, 1 January 2020 (UTC)

I think we have to decide next who is going to be the editor to start chunking text in where. My brain is linear. I am to a point where I am stalled until I see something on a page. And we know we don't want me doing the writing. SandyGeorgia (Talk) 20:21, 1 January 2020 (UTC)

If we focus solely on MSH, we'll have further RfCs, and need to make that clear.

I think WhatamIdoing's suggestion of having some rewrites would be very helpful. --Ronz (talk) 21:39, 1 January 2020 (UTC)

I don't know about that, Ronz; if we focus on MSH, and get a clear result, that should take care of the similar. Remember, in responses to the RFC, people can extend beyond the question asked. @WAID, I don't see how you can get a good example out of MSH, that doesn't muddy the picture wrt DUE WEIGHT. SandyGeorgia (Talk) 21:44, 1 January 2020 (UTC)

Barkeep used the formulation of good news/bad news, and I'm going to do my own version of it. First of all, I think it's entirely due to say out loud that we all owe Barkeep a debt of thanks for his very helpful guidance here. Thanks! I also think it's good that we seem to be having a growing consensus that it's a good thing to present editors in the RfC with specific examples to evaluate, as opposed to asking about generalities. And I do think that we are making genuine progress. For me, that's the good news.

Now for the... you know what. As much as we really are moving forward quite well, and despite the fact that we are approaching our self-imposed deadlines, I think that we have yet to resolve some really important issues for the RfC, and we cannot just wish them away. I'd rather get it right, than get it fast. We've been saying that G/H/I/J are getting close to what we want. But I want to be honest about that: I'm not really seeing a consensus that we are there yet.

I said something earlier, and I'm going to repeat it because some editors agreed with it then and I don't think anyone has really objected to it. I've said that the RfC needs to address two issues, both of them in a way that will lead to a clear consensus, one way or the other:

1. Whether drug prices should be widely presented on drug pages, and
2. What kind of sourcing is needed to support such content.

Those are not proposed questions. They are issues that the community needs to answer, in a way that no one will be able to argue against once consensus has been achieved. Even if we get a consensus that, no, we should not be citing drug prices to those sources, we still need a further consensus that, yes, we should present the information this other way. That's important: we need to get consensus for something and not only against something else. And I've also said that I don't think that we can really accomplish that in a single question. I also don't want to leave the RfC format so open that we fail to get focus in the responses.

So: it seems to me that we need to think outside the box, compared with G/H/I/J. I've been thinking about this hard, and it seems to me that we need to present the community with two or more specific choices. For each of those two or more, there should be specific examples of what it would look like on the page, how it would be sourced, and the policy-based rationales for it (or against something else).

If editors here are receptive to that, I can propose what it might look like, but I don't want to do that unless there actually is interest. --Tryptofish (talk) 22:16, 1 January 2020 (UTC)

Tryptofish, I think the consensus now seems to be a step-by-step approach rather than trying to formulate an RFC that solves all drug price issues in one big bang. So let's focus on that. I think an aspect of your first issue might lead many editors to optimistically / naively say "Yes sure, that sounds really useful to our readers" but when we go into the specifics of "is there really One Price" and "do we have any sources for that" then the response is more like "dang, this is hard and complicated and messy". For the second issue, I don't think it has ever been contentious that high quality commentary sources on drug prices can be appropriate for drug articles, provided we are conscious that "current affairs" sources may be temporal in nature and need reconsideration after a while. So there's a core of drug prices that we all have always agreed on, which leaves just the "routinely added to drug articles" contentious aspect. I think we are likely to make progress on the "As much as you'd like it, it really can't be done" approach to this. -- Colin°^Talk 22:49, 1 January 2020 (UTC)

Trypto, we must simply move forward now, even if we only take baby steps, even if we don't resolve everything. It has been a month, and issues are beginning to fester. Yes we owe Barkeep an enormous debt of gratitude, but I suspect he is beginning to weary, too. We need to get on with it. We have about a dozen editors here expending crazy amounts of time on something we will never get perfect. Let's set a goal to launch within a week (Jan 8); people will rise to the occasion. I don't want to be doing this the rest of my life, and if this is what Wikipedia is going to be about, I've got better stuff to do in real life. SandyGeorgia (Talk) 23:03, 1 January 2020 (UTC)

PS, Trypto, which doesn't mean I'm saying you shouldn't do a mockup of your suggestion, but let's not go back to square one here. SandyGeorgia (Talk) 23:04, 1 January 2020 (UTC)

Well, if the response is going to be negative, I too have better things to do. But I'm not going to support an RfC that I think is going to fail. If the consensus here is to go forward with a format to which I object, then so be it, but if I'm objecting to it, I think it's pretty likely that you'll get a negative reaction from the community, and I don't see what good that will do.

I do not think that there is a consensus that we should have a preliminary RfC now, and then have a follow-up RfC some time later. The virtue of having the community respond to specific examples is that we don't have to solve "all the issues" via some kind of complicated question; instead, we can have a consensus that a specific way of doing it is preferable to a different specific way – and that will cover everything that we need to cover for now. Of course we don't want naive answers, but that should not happen if, as I said, we have well-crafted explanations of how policy applies. But are we going to present those explanations in what would be, in effect, a POV way? Are we really going to present it as "As much as you'd like it, it really can't be done"? Talk about a rigged question! But if editors are presented with a choice: one is "it really can't be done, and here's why", while the other is "yes it can be done", then the community will see that as a neutral RfC and they can weigh which argument wins the day.

If I were to ask everyone here to go back to square one, well, I would not do that. I'm not doing that. That would not be fair. I'm saying that I can do the lift of showing how it could be done, and everyone else just has to take a look at it. That's not so hard. But I'm not going to do it if I'm going to be wasting my time. And I'm not going to do it if it's just a grudging "go ahead and make a mockup but we're just going to say no." But here's the alternative as I see it: you all go ahead and propose a revised version of G/H/I/J that actually works. Personally, I think that's near to impossible. But if you think that's going to be less work for you, go right ahead. --Tryptofish (talk) 23:38, 1 January 2020 (UTC)

PS: Let me put it this way. Please go ahead and do try to create a version of G/H/I/J that actually works. If it flies, I'll gladly support it. But if it turns out as I expect to be harder than it sounds, then I hope you'll have an open mind. OK? --Tryptofish (talk) 23:55, 1 January 2020 (UTC)

User:Tryptofish, I've just started a draft at Wikipedia:Manual of Style/Medicine-related articles/RFC on pharmaceutical drug prices. I don't think it will answer all of our questions, but I do think it is capable of producing some useful information. I'm interested in whether you think it has a chance of producing some useful information.

User:Ronz, on the subject of a re-write, I think the NOR problems could be avoided by re-writing the first to say "In 2014, a non-profit organization sold 250 mg tablets of ethosuximide for US$0.1845 each to recognized healthcare organizations in the Democratic Republic of the Congo. I'm not sure how to put that in the RFC "question", but I think it would be easy to post that as a response to the RFC (e.g., "It violates NOR, which can be got around by copyediting to say ____, but then the DUE problem becomes more obvious and insurmountable, especially if it's placed in the lead"). WhatamIdoing (talk) 01:37, 2 January 2020 (UTC)

The Congo example is NOTPRICE and UNDUE, though. We need policy-compliant examples. @Trypto, our choice is to end up at Arbcom, or put up an RFC to get us moving towards something, even if partial. Arbcom has already mandated that Where there is a global consensus to edit in a certain way, it should be respected and cannot be overruled by a local consensus, which gives us a good idea of how they might lean on multiple WPMED matters. I really think we should not be looking Barkeep49's gift horse in the mouth, and it's time to get an RFC going. We have to stop the paralysis by analysis, and accept a partial solution. SandyGeorgia (Talk) 03:05, 2 January 2020 (UTC)

Do we actually need policy-compliant examples in *this* RFC? WhatamIdoing (talk) 03:30, 2 January 2020 (UTC)

The example I've given above has only minor problems, at least from what's been pointed out so far. --Ronz (talk) 04:31, 2 January 2020 (UTC)

WhatamIdoing, Not necessarily, and WAID, your implication that the existing text is not "policy compliant" just confirms my frustration with having to ask this RFC. But, here we are. -- Colin°^Talk 09:20, 2 January 2020 (UTC)

Well I have been suggested the need for a prioritized solution from the get go. However, it is important not only that we get the RfC out (and that is important) but we do so in a way that will get the kinds of responses necessary for consensus for those priorities. Put another way, it is important that the right priorities are put forth and it is important that the broader community doesn't reject the RfC. I have been pushing a single question not because I think that's what is ultimately going to work (I don't - H/I/J would really be three questions after all) but as a way to try and focus on what's important and a way to focus the question so the community will provide feedback that leads to a consensus. I am not sure if the worse outcome of all this is no RfC or an RfC that comes up with no consensus. If there's no consensus because the community as a whole is split the way editors here are split well that's one thing. But if it's no consensus because of an ill-formed RfC well that's just an unforced error that will have spent a lot of editor time for naught. Best, Barkeep49 (talk) 03:32, 2 January 2020 (UTC)

User:Tryptofish, I never suggested "As much as you'd like it, it really can't be done" would be a question or that the RFC should take a non-neutral stance. Of course those crafting the RFC want to see an answer that is useful rather than one that does not resolve anything. I think there may well be a popular (though hopefully minority) vote to support prices per the agenda (Big Bad Pharma want to hide prices from consumers; Wikipedia is Not Censored), but an RFC must ensure that respondents carefully analyse whether this is possible. I object to merely asking the question as if the wish could be granted. I also don't know why necessarily we might conclude "a specific way of doing it is preferable to a different specific way", when generally not doing it is likely to be preferable, especially considering that "generally not doing it" has been WP:NOPRICES official policy for years.

User:WhatamIdoing, can you please avoid using the word "copyediting" like above. A copyeditor fixes minor issues with prose without changing the meaning or adding and removing facts or points made. Changing

• "The wholesale cost in the developing world is about US$27.77 per month as of 2014"

to

• "In 2014, a non-profit organization sold 250 mg tablets of ethosuximide for US$0.1845 each to recognized healthcare organizations in the Democratic Republic of the Congo"

Is very very much not copyediting. If a statement is untrue and not per-source, no amount of copyediting will fix that. -- Colin°^Talk 09:20, 2 January 2020 (UTC)

Dear friends, I am unwatching this page because, IMO, our considerable efforts here are being undermined by back-channel conversations, and the topic of drug pricing has not been confined to this remit as required at ANI. A few of us are doing all the work to solve problems that aren't being addressed as the ANI receommended. Sorry, bye. Ping me if there is any urgent need for my useless and verbose opinions. I will keep the RFC formulation page started by WAID watchlisted. Regards, SandyGeorgia (Talk) 14:18, 2 January 2020 (UTC)

Sandy I'm really sorry to read this and this is rather a blow to making further progress. We are already hampered with lack of recent participation from anyone on the strong-pro side of the debate. Btw, Barkeep49, I'm not sure if I broke the rules with comments at Wikipedia talk:Manual of Style/Medicine-related articles/RFC on pharmaceutical drug prices. I thought that comments specifically on WAID's RFC text were appropriately made on the RFC talk page. But perhaps that's prohibited and we need to continue to discuss that here. If so, you are welcome to move the text over here in a sub-section, or ask me to do it. -- Colin°^Talk 14:25, 2 January 2020 (UTC)

Colin, going forward it would probably be helpful to reply here (though maybe collapse it?) to honor the discussion in one place element of the ANI close. Also I will join you in sadness of Sandy actively withdrawing from here. The lack of participation will be a problem in general, and the lack of participation of those most in favor of pricing does carry risks to a successful RfC. Best, Barkeep49 (talk) 22:54, 2 January 2020 (UTC)

I've moved it all here. I've also asked User:JzG and Iri's talk page stalkers to look over the draft and see whether it makes sense to people who are smart and highly experienced, but who haven't been actively involved in developing it. WhatamIdoing (talk) 23:16, 2 January 2020 (UTC)

If there are any "back door" discussions, I'd like to know where they are. (If it's just the talk page at WAID's draft, my opinion is that it's no big deal, other than just the practical issue of keeping the discussion in one place. But for that reason, I'm going to comment about it here, not there.)

Just so everyone knows, I'm watching this page closely, so you don't really have to ping me. (But I'm not troubled if you do.)

WAID: thank you very much for setting up that mock-up page. I've looked at it, and I'll comment here rather than there. I raised a concern about that format before, and I'll repeat it now. If we present the RfC in that way, we will potentially get a result that says that the mshpriceguide.org website (the only one cited in the three examples) should not be cited in that way. Then, a few months later, an editor will come along with a different source and make the same kind of widespread edits using that different source. And they will say: "but I didn't use that source that consensus said not to use, so I was honoring the RfC consensus." So my point is that, to get a useful RfC, we need to get consensus for something, in addition to getting consensus against something else. I see that as being a fatal flaw in the single-question approach to formatting the RfC. I also continue to believe that the way that draft presents the question comes across as a loaded question. It implies that there must be some sort of problem with policy noncompliance. And I can predict that the result of going that route is that editors (some of whom may have been quiet so far) will show up at the RfC quite loudly, and say that the RfC is bogus, and it will go off-track. And these concerns are the same ones I've been raising all along in these discussions, without them really being addressed.

On the other hand, I've gone ahead and created User:Tryptofish/Drug prices RfC draft 2, which shows a format for the RfC that I think will work. Please don't anyone get too hung up on any of the wording that I used there, because that will be easy to revise. --Tryptofish (talk) 23:36, 2 January 2020 (UTC)

Tryptofish see my user talk for more on back door discussions. Best, Barkeep49 (talk) 00:04, 3 January 2020 (UTC)

Thanks. I was just going to post that I've seen that, and also at Doc James' talk page. --Tryptofish (talk) 00:08, 3 January 2020 (UTC)

Tryptofish, I have two basic thoughts about your POV (which I mostly agree with), and if you promise to uphold my reputation for being verbose, I'll try to keep it short:

• Yes, we need to get consensus "for", not just "against". But does it have to happen in *this* RFC?
• Do we really think editors are so far gone into the voting mentality that we won't get responses that help us figure out what people are "for" as well? I keep hoping that we'll get responses that say rather more than just "N Bad dog. No biscuit!" or "Y Who's a good boy?!" WhatamIdoing (talk) 03:49, 3 January 2020 (UTC)

I'm sure in either the HIJ format or Trypto's format we'll get many responses, some of them at substantial length, beyond yes and no (support, oppose whatever). The two questions I'd throw out is which format poses the most important question (there will be voter fatigue, especially with one MEDMOS RFC already out there) to get a sense of consensus for/against and which format makes it more likely to get a consensus. Best, Barkeep49 (talk) 03:55, 3 January 2020 (UTC)

Potential accusations of bias

WAID, my answer to your first question would be "yes". My answer to your second question is that it will be a mess if we plan on figuring out after the RfC what the community is for. Who will make that decision? What happens if the closing admins see it one way, but some editors who are very active disagree? It's far better to ask, and get an unambiguous answer. Also, as I'm about to explain some more below, I've just revised my draft page in response to the feedback from Doc James and Colin. You, in turn, might want to revise your draft page in response to the feedback that I have given. --Tryptofish (talk) 21:30, 3 January 2020 (UTC)

@WhatamIdoing: I'm pinging you because I'm not sure whether you saw my comment immediately above. And I want to ask you this question: Let's say we are a few days into the RfC and editors who heretofore have not been active in these discussions suddenly find their voice and complain loudly that the RfC is hopelessly biased and should be discarded, because it basically presents only one "side", and then uninvolved members of the community start agreeing with them. What is you plan of action in that event? --Tryptofish (talk) 20:42, 4 January 2020 (UTC)

I appreciate you asked WAID, but IMO I really hope you don't try to sabotage the RFC by "complain[ing] loudly that the RfC is hopelessly biased and should be discarded" just because it isn't yours. That would not go down well for you at arbcom. Will others? I don't think the main question Do you think that this content complies with Wikipedia's standards for verifiability, due weight, no original research, what Wikipedia is not, and how to write a lead section is biased. It is the kind of question we ask of content every day. It is a very ordinary question for Wikipedians. It is really very neutral. Are the examples neutral? Well look at User:Colin/MSHData and see if you think they are representative. I think (with the substitution of diazepam for mebendazole) they are. If you think the background and info is biased then please argue specific problems. It is also asking for comments, not a vote, which is something I very very much support. All wiki wisdom suggests a plain vote on polarised options (which is your RFC) is a recipe for polarised and adversarial comments and disunity and in the end "consensus" by counting votes. I have no doubt that some in the pro-prices faction will totally ignore the factual, source and policy issues, and present their case based on Big Pharma suppressing prices and Wikipedia being Not Censored. We will see sweeping statements that of course WHO/MSH and Medicaid and BNF are totally reliable secondary sources. And drug pricing is such an obvious world concern you'd have to be a drug company shill to want to censor prices. And some people will be totally cool with wiki-docs doing original research. This will happen. Let it happen. If Wikipedia decides those things are more important then so be it. -- Colin°^Talk 22:16, 4 January 2020 (UTC)

As you know perfectly well, I did not say, loudly or softly, that "the RfC is hopelessly biased and should be discarded". I cautioned that other editors are likely to do so, if we do not get it right. I'm trying to get the RfC done right, and I'm quite comfortable with how that will go down if this ends up at ArbCom. --Tryptofish (talk) 22:22, 4 January 2020 (UTC)

Thanks for the ping, Tryptofish, because I did miss that.

I've been one of the regulars at WP:RFC for about a decade now. I consider accusations of a "biased" question to be par for the course on contentious subjects, and if you want to search the old archives, you will find that my opinion is consistently that claims that "The question is biased!" mean "My side is losing!" I'm not afraid of seeing those accusations, and my plan is to ignore them, because they say more about the fears (and usually inexperience) of the accuser than about the question. (Now, if you personally thought that the question was biased, I'd be very concerned, but that doesn't exactly seem to be your concern.) If you'd like, we could ask the other RFC regulars whether they think the question is biased.

The other theme in your comments is that the results might be hard to interpret. I agree with you: That's a risk. However, I see this phase as information-gathering, and then (as stated in the RFC "question"), I expect the next phase to involve a proposal that people can be for or against.

The way I expect to handle this is to care less during this RFC about summative "votes", and more about the rationales. Let's say, hypothetically, that editors read the three RFC examples and say, "Ugh, that's all wrong". What actually matters to me is the next sentence: Do they think it's bad because per-pill costs were converted to monthly costs? Or is that okay, and the problem is that the sentences are in the lead but not in the body? Or because MEDRS suggests a five-year timeline for sources (all the WT:MED regulars know what I think about that) and 2014 prices are now technically six years old? Or because they think NOTPRICE for a drug requires a source saying that the price of specifically ethosuximide itself is terribly important to the world, and no amount of sources saying that the price of WHO Essential Medicines for epilepsy can possibly suffice to prove that the price of ethosuximide is worth mentioning? Once we've seen which points of policy and common sense people care about, I think we can build on your draft to make an actual proposal for adding a section about prices to MEDMOS. Or, to put it another way, the purpose of this RFC is to make yours produce a clear consensus for fixing this guideline. This is just the necessary baby step to get us on that path. WhatamIdoing (talk) 01:52, 6 January 2020 (UTC)

Convenience break

We are currently using secondary and government sources for prices such as Medicaid, MSH, and Tarascon. Unless people consider those primary sources? So not sure why "Secondary sources should not be required. Appropriate primary sources include: example, example." Doc James (talk · contribs · email) 05:42, 3 January 2020 (UTC)

Doc James, most databases are considered primary sources under Wikipedia's system. (The MSH and NADAC databases are second-hand or secondary data, which is not the same as a secondary source.) Primary sources can be perfectly reliable. That sentence could be completed with "Appropriate primary sources include the MSH and NADAC databases". WhatamIdoing (talk) 00:37, 4 January 2020 (UTC)

Responding to Barkeep post of 03:55 3 Jan, made while I was in Ceylon, asking which format will get more responses. Neither RFC is ideal, but IMO we were charged with formulating an RFC to address the wrong (indeed, a non-existent) problem, so it has been difficult to nail down. We typed for a month. We did our best, and we came to WAID's formulation. Should we switch gears now? I believe that Trypto's formulation will suffer the same fate as my ill-formed RFC on the medical disclaimer of several years ago. When you present options, people pile on other options. So, I go with WAID's, even if we are only addressing one issue at a time. SandyGeorgia (Talk) 14:53, 4 January 2020 (UTC)

Sandy, the solution to that would be to specify in the RfC introduction that no new versions should be added after the RfC begins, although editors are free to comment about alternatives within the discussion. And there is also that danger in the other format: editors could make all kinds of comments about how that example with [name of drug] could be rewritten in a particular way, without any consensus about which "particular way" is the best one. --Tryptofish (talk) 20:37, 4 January 2020 (UTC)

Tryptofish I appreciate that asking only about MSH-sourced text will not reliably put out the fires at NADAC or BNF or Tarascon or Drugs.com sourced texts. But if anything can be seen from the volumes I've written about the flaws in our texts, is that it is amazing that one can make so many mistakes and commit so many policy crimes in just a few words. I think if we try to explain why the "drug prices everywhere" approach is not just not-encyclopaedic but also totally impractical, for all examples, we will get totally bogged down. Fatigue will then likely to see over-simplistic replies like "Support: we using secondary and government sources for prices such as Medicaid, MSH, and Tarascon". We need to allow participants a chance to focus on a smaller level of practical issues than "all drug price sources".

I think Wiki largely discourages "prices everywhere" on fundamentally practical grounds, more than on "encyclopaedic" grounds. We aren't a standard paper encyclopaedia and already contain much trivia or dubious lists of facts. Practically, there are plenty sites (GoodRX, Drugs.com, BNF in the UK) where patients and readers can reliably find out about drug prices in their country, and get figures that are accurate TODAY for the indication or prescription they are concerned about, rather than citing a book from 2015 for a totally unknown indication and dose. All the evidence suggest Wikipedia is crap at this. Same goes for the prices of other things from mobile phones to insurance to properties in your area to train tickets.

I think it is a huge mistake to frame drug prices round concepts like primary and secondary sources, which Wikipedia has historically had mixed interpretations of meaning. James says above that the BNF is a secondary source, and it is true that it gets its information from the NHS, who in turn give a mix of regulated price and/or indicative price for the drugs, the latter of which comes from drug companies and pharmacy contracts. A better distinction is that all those sources (with the exception of Tarascon's $$$ symbolic prices, which have their own huge problems) are simply databases of prices of products with barcodes, and all those sources give their own unique kind of price (all different variants of wholesale or retail, some actual, some list, etc). Those sources are raw data, at a level of multiplicity and complexity that none of us are proposing is reproduced on Wikipedia.

Wrt notability of prices, saying "secondary sources have written about issues that are specifically about the pricing of those medications". Repeatedly it has been claimed that the fact that e.g. BNF include prices for all their drugs (similar for Drugs.com and Tarscon's book) means they have been written about. Clearly the authors of those sites/books, when writing about each drug, considered it relevant to give price data. And Google will find someone somewhere mentioning that X is a low cost drug compared to the new drug Y which is expensive. But doh!, all new on-patent drugs are expensive and most existing generic drugs are cheap, so that isn't exactly news to anyone that someone might make that remark in print or online. The advocacy argument for including prices is so strongly held by some, that we need a much higher barrier-to-entry than simply being written about in secondary sources.

I'm very nervous about get-out-of-jail clauses like "or used only with care". We aren't here to redefine fundamental policy. Anyone can argue they are careful. We see in the lead RFC that e.g citation excess is justified on the grounds that there is no policy against citations: any guideline-caution or recommendation to seek per-article consensus about citations is simply cast aside. An "or used only with care" clause simply says one can ignore the preceding text.

So I don't think, sorry, your RFC is appropriate today. Let's start with the RFC on MSH-sourced price statements today and see where that takes us. We can learn lessons from it. It may be that Wiki so clearly rejects raw-database-sourced drug prices that existing policy on WP:NOPRICES becomes the clear consensus, and we all do already know how to write about prices when newspapers, etc have made comments about them and give us a price-to-treat or a price-per-year without us having to get our calculators out. -- Colin°^Talk 08:54, 3 January 2020 (UTC)

Doc James and Colin, thank you both for the feedback. I've just made some revisions in response to that feedback (and I think it might be a good idea to, likewise, revise the H/I/J draft to address the feedback that I've given about that). Here's a quick summary of what I changed. I removed all mention of primary and secondary. Thanks for correcting me about that; I hadn't understood that properly before and I appreciate the opportunity to clear up my understanding. I changed the description of what had been primary sources to instead be sources that are databases of prices (based on Colin's mention of "raw-database-sourced drug prices"): please check me on that, as I can certainly change it again. I also removed that "get-out-of-jail" clause. As for the broader point that, presumably, everyone should already know that Wikipedia largely discourages "prices everywhere", I'm pretty sure that if everyone really agreed about that, we would not be having this discussion or this RfC. And as for the difficulty of tl;dr if we try to cover many database-style sources, I'm fine with having just a few, not every one. We can do as many or few as we wish. But we should do at least two, because then there will be an established general principle as opposed to something that could be gamed as being about just one source. --Tryptofish (talk) 21:44, 3 January 2020 (UTC)

More looking for "Medication prices can be included when appropriate sources are avaliable." Doc James (talk · contribs · email) 23:36, 3 January 2020 (UTC)

I appreciate the revision attempt. I think, though, the issues are too complex and require some effort on the part of voters to look at actual article text and actual usage to appreciate the problems. Simply asking people to support polar-opposite A or B approaches is wrong. Particularly as B is effectively asking people to support original research and undue weight and out-of-date sources. I don't think we should be asking people to support a fundamentally broken option. Nor do I remotely support the kind of question James poses above, which is leaving "when appropriate sources are available" to the judgement of whoever edit wars the best. -- Colin°^Talk 13:29, 4 January 2020 (UTC)

Tryptofish, I think your draft is too much of a skeleton at this point in the game. We've been discussing this since October and fatigue is already causing some to unwatch and drop out. I think we should put our energy into WAID's draft. It doesn't meet exactly what any one of us would have wanted, but nor does it ask to community for permission to do unacceptable things, which is what your question B asks and what in practice James's question would continue to permit to do. -- Colin°^Talk 15:02, 4 January 2020 (UTC)

Thank you both for your replies. James, my concern with "when appropriate sources are available" is that it presupposes that sources are "appropriate", which is something very much at the heart of the disagreements here. Colin, I would say the same thing about your objection to asking about part B: you are assuming that the consensus must be that the sources are inappropriate. But these are the things that we should be asking the community to determine, as opposed to, in effect, rigging the RfC to get a desired result. As for the skeleton, my intention is that interested editors here would be able to fill those things in themselves to complete the RfC, particularly when there are those of you who have already written and analyzed so much about it. I'm sure that you, for example, could easily provide examples and pro/con rationales. You probably could even copy-paste them. Of course, not doing that and then complaining that no one has done it ends up being a self-fulfilling prophecy. --Tryptofish (talk) 20:32, 4 January 2020 (UTC)

Notifications given

• [87] If there are any "back door" discussions, I'd like to know where they are. ... --Tryptofish (talk) 23:36, 2 January 2020 (UTC) Putting this here per your request, SandyGeorgia (Talk) 23:38, 3 January 2020 (UTC)

Sandy, I'm going to WP:AGF and assume that you did not see: [88]. --Tryptofish (talk) 20:06, 4 January 2020 (UTC)

Sorry, Trypto ... I was in Ceylon for a day :) Anyway, this section gives a place for others. My apologies, SandyGeorgia (Talk) 20:46, 4 January 2020 (UTC)

Newbies to the debate

User:Tryptofish thank you for the notification at WP:MED. I was not following the debate but would like to participate as a newbie. As I understand it, there area two parts of the debate. The first is regarding point #5 of the WP:NOTDIRECTORY which states,

"Sales catalogues. An article should not include product pricing or availability information unless there is an independent source and a justified reason for the mention. Encyclopedic significance may be indicated if mainstream media sources (not just product reviews) provide commentary on these details instead of just passing mention. Prices and product availability can vary widely from place to place and over time. Wikipedia is not a price comparison service to compare the prices of competing products, or the prices and availability of a single product from different vendors or retailers."

From what I can read, part of the RfC will debate whether or not medication prices falls into the exceptions listed above (e.g. is there a justified reason). The remaining question will debate which medications and which sources are acceptable? Can you confirm if I have this correct and if so, is the draft on your talk page now? Also which opinions you're seeking at this time. Thank you for shepherding this topic. Ian Furst (talk) 15:06, 4 January 2020 (UTC)

Ian, there has been much discussion about how the RFC should pose questions, give examples, be structured. There are currently two drafts. The first is Wikipedia:Manual of Style/Medicine-related articles/RFC on pharmaceutical drug prices and is being discussed at the #Debugging the draft section below. The second is at User:Tryptofish/Drug prices RfC draft 2 and some comments about that have been posted above. James, above, has also proposed a one-line question, but nothing more than that. IMO the former, by User:WhatamIdoing is further advanced (the other being just a skeleton) and by focussing on one source and a few concrete examples, is best placed to uncover the issues surrounding the routine inclusion of drug prices. -- Colin°^Talk 15:50, 4 January 2020 (UTC)

The formulation of an RFC was remitted to this forum almost a month ago by ANI. WAID's draft is further developed, and while addressing only one part of the overall picture, I believe it has the best chance of giving us a clear answer on at least that one part. Depending on what kind of response it generates, the topic might broaden, but I fear that Trypto's format will not lend to a clear answer on any one part of the broader issue, as it will encourage others to suggest even more options. We have been working for a month; we should finalize and launch WAID's RFC in the coming week.

Also, welcome Ian! Since we have put so much work in to this, there is voluminous information to read on the page. I hope you will find the time to read through the entire large page, as it is late in the game to re-hash territory already covered. Regards, SandyGeorgia (Talk) 16:26, 4 January 2020 (UTC)

Ready, steady, go

User:Barkeep49, I think Wikipedia:Manual of Style/Medicine-related articles/RFC on pharmaceutical drug prices is ready. As I understand it, ANI thought it would be best if an uninvolved admin endorsed its neutrality. Can you post a request at some suitable forum to find volunteer for that step?

Here are my thoughts about how to manage the RFC once it's underway, and I'd like to hear all of yours, too.

First, I'd like to encourage all the "principals" in the original dispute to stand down for the first day or two. Let's imagine that an admin certified it as neutral very soon. In that case, I could probably take the draft tag off and list it as an RFC as early as Tuesday. If that's the schedule we end up on, and if you think that other people might think you've got a dog in this fight, then please stand down until at least Wednesday. There's nothing like long or angry posts, especially from our recognizable community leaders, to scare off some contributors. I may not post my own views at all, and I certainly don't plan to do so during the early days. We can let this run for weeks, or even months if we're still getting good responses. There's no magic timer for RFCs. I promise that you will have a chance to have your say, even if your post is #10 instead of #1.

Second, I want us to be encouraging the uninvolved editors to engage in this RFC enough to tell us what they think. Some people will just want to dump a drive-by vote on the page, but if they're willing to explain their thought process, then I want to find out more. I am discouraging straw-poll or "survey" approaches, and I hope that you can all support that in practical ways, like breaking long discussions into sections with useful names like === Thoughts on X ===. Getting detailed explanations from our volunteer editors is a gift that we should treasure. I expect to be asking some editors questions to encourage discussion. If you think that you can ask a question that will draw out more details from an editor or that will encourage that editor to connect with another editor, then please consider doing that. Something like "Do you feel like your idea relates to what User:Example was saying last week?" or "Do you think that might work better with <this slight change>?" could be good. The goal is to get the other guy talking. A good, responsive question, phrased with respect, can be an excellent tool for producing further explanation. An amazing success looks like a couple of editors putting their views together to come up with something that's better than what any of them started with.

Third, if you see an opportunity to meatball:DefendEachOther, especially if it's someone from the other "side", please do so as quickly and as gently as you can. Nobody involved in constructing this RFC wants Wikipedia to get worse. We all have the same ultimate goal. If you'd like, I can ask the WT:RFC regulars to help out with this.

Fourth, I personally don't feel like I'll need an official "closing statement" to know what I've learned from this RFC. However, if you do, then please be bold and speak up now, especially if you'd like to have a "team of three" approach. Recruiting three people after an RFC has ended can be difficult and result in needless delays.

If others have advice they'd like to add, or would like to suggest a different approach, please post here. I'll make time to check this page between meetings (probably in ~12 hours or so). WhatamIdoing (talk) 06:41, 6 January 2020 (UTC)

User:WhatamIdoing I think it is important to mention that one seller like the IDA Foundation sells at the stated price in more than 130 LMIC. LMIC make up about 80% of the global population. And most people in these countries pay for their medications out of pocket.

Per "many are out of date" is fairly non neutral. Many of the prices are from 2015. That is not out of data but simple from 2015. Additionally many of them are not otherwise incorrect. They are simple approximates. Yes sometimes 30 days rather than 31 days is used for a month (using 30 days for a month does not make an approximate "incorrect".

That is a fairly one sided background. It is missing details such as that the lack of transparency around prices results in deaths per Doctors Without Borders.[89] Or that prices of electronics are often included in Wikipedia articles including featured articles. Doc James (talk · contribs · email) 07:11, 6 January 2020 (UTC)

I don't think it is important to mention one supplier (IDA) when there are 35 in the database and a third of all uses of the MSH database cite records with no suppliers at all. The majority of our citations rely on a buyer price, either because that is the only price or because it has been used for the upper-bound. The Guide says "Buyers: These prices should not be used as international reference prices". WHO/HAI says: How representative reference prices are generally depends on the number of suppliers quoting for each product.. We do not have a source saying "IDA is representative of wholesale prices in the developing world".

The "pay for medicine out of pocket" argument is also not directly relevant to quoting an international reference price. Again WHO says "in many low- and middle-income countries medicine prices are high, especially in the private sector (e.g. over 80 times an international reference price); availability can be low, particularly in the public sector (including no stocks of essential medicines); treatments are often unaffordable (e.g. requiring over 15 days’ wages to purchase 30 days’ treatment); government procurement can be inefficient (e.g. buying expensive originator brands as well as cheaper generics); mark-ups in the distribution chain can be excessive; and numerous taxes and duties are being applied to medicines" they give an example: "The price of originator brand atenolol 50 mg tablets is over 20 times the international reference price in all the countries except India (where it is still high at 5 times the reference price) and Kazakhstan. Even the lowest-priced generic is very expensive in all countries". So basically, the wholesale price of generic medicines to the state healthcare is irrelevant if the state healthcare has no stock of that medicine and the patient has to buy a premium-brand version on the private market. And that is the norm.

Wrt "Many of these are out of date or otherwise incorrect" I would agree we can state that more neutrally. The 92% of MSH prices in articles are from 2014 so lets say that "Nearly all cite records from 2014". As for being incorrect, we are not talking about number rounding or the exact number of days in a month, but basic maths errors with multiplying (or forgetting to multiply) or basic beginner-level statistical errors with how the data is offered. Perhaps a more neutral statement is that "the mathematical and statistical correctness of figures given has been questioned".

The background is merely the background on the source, and the question asked is simply about source->text policy. We are not at this stage asking that we should include prices in Wikipedia because otherwise PEOPLE WILL DIE. Let's leave advocacy out of Wikipedia please.

Similarly the WP:OTHERCRAPEXISTS argument also has no place in deciding if text meets policy.

In summary, James, you are welcome to make these points when you make your case in the RFC, but these points are not neutral facts, they are easily challenged, and most are not directly relevant to this RFC. -- Colin°^Talk 09:05, 6 January 2020 (UTC)

James, the notion that "prices are included in Featured articles" was covered above. If that faulty, WP:OTHERSTUFFEXISTS emerges during the RFC (as it is an oft-repeated meme), then I will have to unpack the analysis even further on the RFC. FAs are from perfect even when they get the star, and are even less so now that FAC and FAR have died (FAC stats-- I was 2007 thru 2011, scroll down), and there is basically NO ongoing review of older, out-of-compliant FAs. One of the FAs held up as an example was written by an editor whose socks supported the FAC, whose FAs I was forced to promote by "consensus" (of his socks), even though all of his FAs were awful. Someone will introduce that bogus argument to the RFC; I hope I don't have to unpack it there, and I hope that won't come from you. This particular RFC isn't looking at the broader pricing issue; it's looking at drug prices. And even at that, in some cases, the prices that are in Featured articles are in compliance with WP:NOTPRICE. I hope this RFC doesn't have to unpack all of this, which is beyond the scope of a first step. SandyGeorgia (Talk) 11:29, 6 January 2020 (UTC)

WAID, I fully endorse your launch-step commentary, and agree that most of us should initially stand down. I hope that admins intend to assure that the civility sanctions in place are, at this stage, aggressively defended (as by now, everyone is aware), and that in particular, as I mentioned explicitly in the ANI, CFCF is held to the civility restriction. Thanks for all your work!

Re: your mention of what we've learned: regardless of outcome, I learned a lot already. Only yesterday, when I saw your comments as you wrapped this up, did I understand that by staying focused on "what should MEDMOS say", the result is that we additionally cover the NOR/SYNTH problem. You were focused on "what should MEDMOS say", while I was focused on, "is there an example of these sources not using NOR/SYNTH". The search for the elusive example consumed too much bandwidth in this discussion, and that one's on me. SandyGeorgia (Talk) 11:40, 6 January 2020 (UTC)

SandyGeorgia — I do not believe I have on any occasion been uncivil, having with arguments pointed out what I percieve to be WP:GRIEFING. I understand that this position may be percieved as insulting as such, but it does not violate WP:CIVIL. If it did, we could never act in any way against the will of an individual editor (see: Wikipedia:Our social policies are not a suicide pact)
I have argued for my position and it is no matter how we look at it — a fully legitimate position to hold. If anything I find that the collegiality within WP:MED has been taken advantage of — and abusive and disruptive behavior has been let slide too long.
What is however noteworthy: chosing to single me out in the manner above, despite no comments from my part for a week, and me not having commented on the issue of PRICE for over 3 weeks. That does not seem in line with WP:CIVIL. Carl Fredrik ^talk 15:19, 6 January 2020 (UTC)

@CFCF: While I may disagree with some parts of your post, I agree with and acknowledge that singling you out here was not at all helpful or wise. I thought of this within minutes of posting, but unfortunately, pings are not undoable (another of the many reasons I hate the pingie-thingie-- you can't walk them back), so it didn't seem that striking it would be helpful. I do apologize for singling you out. At any rate, I hope you will agree with WAID regarding what she hopes to see in the conduct department from recognized leaders of WPMED. I will do my best, and hope you do as well. Once again, my sincere apologies. SandyGeorgia (Talk) 15:30, 6 January 2020 (UTC)

Debugging the draft

Draft RFC at: Wikipedia:Manual of Style/Medicine-related articles/RFC on pharmaceutical drug prices

Generics

Thanks for getting this moving, WAID. I am unclear if we are actually restricted to generics? Why do our questions not apply to all drugs? SandyGeorgia (Talk) 03:08, 2 January 2020 (UTC)

If we are focusing on the MSH International Medical Products Price Guide, then it will be mostly generics and drugs from WHO's list of essential medicines. I'm not sure if some HIV drugs are still under patent (and of course some countries have different rules about patents, so produce generics when e.g. the US is still paying the patent price). But the HIV drugs are also a bit weird in terms of supply to developing nations and I think actually there's a better database for prices for those (can't recall right now). -- Colin°^Talk 11:42, 2 January 2020 (UTC)

The considerations for small-molecule generics are different from brand-name drugs (including most biologics, even past patent expiration). You'll get some news sources for shiny new things, and the price (and therefore sales) has a considerable effect on the company's stock price, etc. I specified generic because all of these are, and none of those exceptions apply. WhatamIdoing (talk) 22:01, 2 January 2020 (UTC)

Which examples to use

Colin might want to switch out the sample drugs to include one where it's not even clear what the dosage is or what is being treated. SandyGeorgia (Talk) 03:10, 2 January 2020 (UTC)

I'll have a think. Look at User:Colin/MSHData also. If a third of our drug prices have no suppliers at all, then I think an example of that needs to be picked. Ethosuximide is very much a one-indication drug, though the dose will vary by patient. Carbamazepine's main use is epilepsy but there are also several important secondary indications (and sometimes these, often off-label, indications can actually outstrip the original licensed one, and I'm not familiar enough to know if that is the case here). Mebendazole is just used for parasitic diseases, but there are different treatment regimes for different kinds of infection. So perhaps there are other/different examples to cover the spread of issues with us giving one dose as though there is only one indication and one kind of patient. -- Colin°^Talk 11:42, 2 January 2020 (UTC)

I pulled these straight out of a previous page. I don't think that we want to present the most complicated cases. We should start with the simpler cases, and see what information we can get. WhatamIdoing (talk) 22:04, 2 January 2020 (UTC)

I'm not suggesting we pick complicated cases, but also not just pick easy ones: we need to be representative of the usage. Considering that a third of our drug price statements have no suppliers, I think we must include one as representative. An example we have previously discussed is diazepam which says "The wholesale cost in the developing world is about US$0.01 per dose as of 2014"[90]. This is also a drug people will be familiar with and a good example of the common situation that an old drug has many indications and possible doses.

I think we can abbreviate the per-drug commentary about "what the source says" by moving some of the background about "organisations" into a sub-section below. I think we need a short sub-section about MSH International Medical Products Price Guide. If you'd rather not include the "multiple formulations" fact in the per-drug section, then this would be the place to mention that Carbamazepine has 4 variants, Mebendazole has 5 variants and Diazpame has 3 variants in the database, each different strengths or formulations (tablets, syrup or ampoule). We could link each time to the by-name database search result. We can briefly explain buyers and suppliers. It is also a hard fact that it was updated every year from 1986 to 2015. This "is no longer being updated" not a minor fact when our readers consider policy for inclusion. I don't think we should keep information like that hidden from anyone. -- Colin°^Talk 10:15, 3 January 2020 (UTC)

Which example do you (you-Colin or you-anyone) think should be removed, to make space for diazepam? I'm happy to run any three MSH-based examples that people want. I prefer sticking with MSH (because one database is enough for people to wrap their heads around) and with only two or three examples, but I do not care what the examples are. WhatamIdoing (talk) 00:26, 4 January 2020 (UTC)

I think if we are going to replace, then dropping Mebendazole is my choice: it has similarities with Carbamazepine and I do want to be representative of real usage and issues. Wrt the database update, readers will not know it was typically updated annually. How about: "This database was updated annually since 1986, though has not been updated since 2015". -- Colin°^Talk 11:36, 4 January 2020 (UTC)

What the source says

WhatamIdoing, I think there is an initial problem with the "what the source says". We need to back up a bit to see what the source says about the drug, and not just one variant tablet size or formulation: Ethosuximide, Carbamazepine and Mebendazole. We can see from that:

• Ethosuximide is only listed in one 250mg tablet formulation. A wee problem is that if we update the search to 2015 we get no results at all! Indeed if you look at the chart at the bottom of the 2014 page, you see that there was supplier data between 2000 and 2009 but after that, only 2014.
• Carbamazepine initially looks like it has 6 variants, but two are duplicates. So we have a 100mg/5ml suspension, a 200mg plain tablet, a 200mg sustained-release tablet and a 400mg sustained-release tablet.
• Mebendazole has 5 variants: A 100mg/5ml suspension, a 100mg plain tablet, a 500mg plain tablet, a 100mg chewable tablet and a 500mg chewable tablet.

So we need to be up-front that the source offered options, and the editor chose one of them. We need to be careful the examples don't offer illusory easy answers to picking one of them (such as, there's one variant that has lots of suppliers and the others have none, or that the 100mg and 200mg tablets work out the same price by dose anyway). The existing method for choosing a variant isn't foolproof: the diazepam article picked the wrong tablet with no suppliers whereas a different tablet size has lots of suppliers. -- Colin°^Talk 11:42, 2 January 2020 (UTC)

[snip]

We are giving the prices and DDD but not actually helping readers with the maths. This matters because sometimes the DDD is being used and sometimes not. With Ethosuximide, if we multiply the 0.1845 250mg by 5 (1250mg DDD) and by 30 we get £27.67 (not sure where 27.77 comes from). With Carbamazepine, we are multiplying these 200mg tablet prices by 5 to get 1000mg and using that for the daily dose. With Mebendazole, we are actually just giving the price of the 100mg tablet, not the DDD of 200mg. So what do we even mean by "dose"? Since the DDD isn't being used here, it may be worth me explaining to you guys using the BNF that you can't read in the USA, which likely focuses on the kinds of parasites we get in the UK

• Threadworm infections.
  • 100 mg for 1 dose, if reinfection occurs, second dose may be needed after 2 weeks.
• Whipworm infections, Hookworm infections
  • 100 mg twice daily for 3 days
• Roundworm infections
  • For Child 1 year: 100 mg twice daily for 3 days.
  • For Child 2–17 years: 100 mg twice daily for 3 days, alternatively 500 mg for 1 dose.
  • For Adult: 100 mg twice daily for 3 days, alternatively 500 mg for 1 dose.

So a "dose" could be the 100mg one-off dose, or the 200mg daily dose for three days, or the 500mg dose you take once. How on earth do we cover this? For many youngsters on Wiki, they may only be familiar with ibuprofen tablets and the contraceptive pill, and unaware that medicine dosage and indications for drugs are complex. Maybe we need a little side-box for each drug, that explains things that Wikipedia generally is forbidden to cover like how the dose is recommended for each indication/patient-group. -- Colin°^Talk 11:58, 2 January 2020 (UTC)

Hi, User:Colin. I agree with you overall, but I think this initial RFC needs to focus on the simpler cases of whether the text matches the already-cited source. I would be perfectly happy if an editor says that the cited source discusses one pill size, but there are several other pill sizes, and why was that one size (or that one database) selected? However, the question in this RFC is smaller: we've got a sentence, it's got a source, and are people okay with that particular combination?

Similarly, an uninvolved RFC respondent might notice that there's a typo in the one price, and someone might opine on the subject of whether calculating price per DDD is a routine WP:CALC or a case of WP:SYNTH. Editors might have strong views on whether it's important, or on geographical bias, or any number of other subjects. That's fine, and I want to hear everything that's on their minds, but I don't want to push those questions. I'm just looking for a starting point with this RFC. WhatamIdoing (talk) 22:19, 2 January 2020 (UTC)

Hmm. I'm quite fundamentally opposed to us neglecting up-front to say that the price link used here is merely one from several options. I don't think it is acceptable to wait for some participant to notice, halfway through the RFC. The multiplicity reduced to One Price is a intrinsic problem with MSH, NADAC, Drugs.com and BNF sources. I appreciate we need to be compact, but we absolutely must IMO state that "the source" actually lists several variants each with their own price sets, and the editor has picked one.

The fact that the editor can link to just one pill size by URL is a artefact of the MSH website. With NADAC one can't even link to the drug at all, just that week's database of a bazillion prices. With BNF and Drugs.com one must link to a page that gives prices for dozens of formulations. So I don't think we should hide the fact that when the editor, seeking price information about this drug from MSH, typed the drug name into the search box, they got multiple results and picked just one. It is fundamental to what "the source says". -- Colin°^Talk 09:07, 3 January 2020 (UTC)

Colin, I've mentioned this, but perhaps you will think it would be better to add, for each example, something like (This is for the 100 mg pill; there are are also 50 mg pills, 250 mg pills, and a liquid in the database.) If you all think that would be better, maybe someone could make a list of the size/forms for each example? WhatamIdoing (talk) 07:14, 4 January 2020 (UTC)

WhatamIdoing, the improvements look good. Can we change "such as the size and dosage form " to "such as the strength and dosage form ". I would change "some drugs are available in different amounts" to "most drugs are available in different strengths", as "amounts" is ambiguous with bottle or package size, and it really is the norm for there to be multiple records. See User:Colin/MSHData where I started adding a count of variants but only got to "C" -- 65% of entries have 2 or more records. And the number of variants in the MSH database is much lower than the number of variants in Drugs.com or BNF. So multiple variants is very typical. This fact is a little buried in a (necessarily) dense intro section, so I do think it very much worth briefly repeating the explicit case for each drug exactly what you put in quotes above. When mentioning the guide, I suggest "International Medical Products Price Guide (website)". I plan to try to improve the article more: there's more room to expand in the article, but it is necessarily restrained by policy unlike a more casual intro we can do here. -- Colin°^Talk 11:30, 4 January 2020 (UTC)

Thanks for the strength/amount note. "Strength" is what the database uses, and it's what I should have used from the beginning. I think I've {{resolved}} every request in this section. WhatamIdoing (talk) 02:44, 6 January 2020 (UTC)

Buyers and sellers

Wrt wording "One organization said that they sold ... in 2014" should really be "One organization said that they sell... in 2014". We have price data but no evidence they actually sold any. It might be simpler to call these "organizations" "suppliers", especially as that's what the source calls them, and we will end up discussing buyers and suppliers. -- Colin°^Talk 11:42, 2 January 2020 (UTC)

In the Carbamazepine example, you have focused on the suppliers. In fact, the highest price in the article text ($0.24 per day) is taken from one Buyer (SICA: System of Central American Integration). So, to discuss what's gone on in that article text, we need to talk about Suppliers and Buyers, and should really enlighten readers that for example WHO encourage we take the median supplier price, and only consider this representative of an international price if there are many suppliers. We can source this and if necessary quote verbatim. -- Colin°^Talk 11:42, 2 January 2020 (UTC)

In the Mebendazole example, the highest price in the article text ($0.04 per dose), comes from one Buyer (South Africa Department of Health) and in a package of 6 pills, not 1000. -- Colin°^Talk 11:42, 2 January 2020 (UTC)

On these points: I wrote that they "sold" at that price, because it was five years ago. That's ambiguous, but saying they "sell" at that price also feels wrong (because prices change).

I listed supplier prices because those were the ones used to create the text. There is quite a lot in the source that I didn't reproduce. WhatamIdoing (talk) 22:27, 2 January 2020 (UTC)

Wrt "I listed supplier prices because those were the ones used to create the text", no it was not just the suppliers that were used to create the text. In two of the three drugs, the upper range comes from the buyers. I believe the upper and lower prices were drawn from both supplier and buyer datasets in all cases there is a range, and it is mere chance whether upper or lower bounds come from one or another. Indeed, a third of the time, only buyer prices have been used because there are no suppliers at all. We need to bring in buyers, and two explain those two price datasets are not the same thing. We can't have someone claim "these are the prices offered by major suppliers to governments in the developing world", as has been done, when in fact one third of the prices have no supplier data at all, another sixth only list one supplier and another sixth only two. The reliance on and usage of inferior buyer prices is a fundamental problem with a large portion of drug prices. -- Colin°^Talk

 ToDo Thanks for explaining it again. I thought those only used buyer prices, but I was wrong, and that has definitely got to be fixed. I'm willing to do it (if nobody else gets there first), but I won't get it done within the next couple of hours.

I like the idea of explaining the MSH database in its own section. That could reduce the repetition. WhatamIdoing (talk) 00:22, 4 January 2020 (UTC)

 Partly done I've got two paragraphs up about the database. I have not yet looked up the high/low/median/ratio numbers for the buyer prices for the second or third example. I'd rather do that after we decide whether we're replacing one of them with diazepam. WhatamIdoing (talk) 07:19, 4 January 2020 (UTC)

I think it better to use "supplier" than "seller" since that's the terminology the source uses and what we've used throughout discussion and the terminology on the wiki articles. I suspect the reason for them using that word is that people might think the sellers are the drugs manufacturers themselves, or retailers. They aren't typical wholesalers either. -- Colin°^Talk 11:39, 4 January 2020 (UTC)

 Done and  Done. I no longer consider the RFC to be blocked on these issues. Speak now or forever hold your peace, and all that. WhatamIdoing (talk) 02:51, 6 January 2020 (UTC)

Missing links

The bits at Additional information can be found at... and Previous discussions on this subject include... still need to be finished. Anyone's welcome to add whatever they want there (or post it here, and I'll add it). I'd really appreciate some help with finding all the relevant things. WhatamIdoing (talk) 22:27, 2 January 2020 (UTC)

I'm not clear where you are quoting from. A workspace on another page? --Ronz (talk) 00:55, 3 January 2020 (UTC)

Two paragraphs in the middle of the draft's ==Background==. I tried to fill them in with something, but I probably missed some appropriate links. WhatamIdoing (talk) 03:44, 3 January 2020 (UTC)

General

I don't want to be here, but here I am. Please feel free to ignore me, but. I still think there is no need to present three examples from the same database. We will get a result about one database. We could present three examples from the three different sources that have been incorrectly used, and get a broader result. It is the same principle; we have no good data on drug prices anywhere. SandyGeorgia (Talk) 12:30, 4 January 2020 (UTC)

I really don't want to get bogged down in explaining the differing problems with each database. It is hard enough with just one. The problem of just picking one medicine is that it doesn't show up the range of issues. If we only pick an example with supplier data, voters may think all records have supplier data. If we only pick an example with one pill size in the database and no syrups or ampoules, voters may ignore the problem with that multiplicity. If we only pick the worst possible example (there is in fact a usage where the "developing world" price is a totally invented work of fiction with no records at all) or most complex example, then folk may argue that this is unrepresentative and that that kind of problem is rare. I think three representative examples are not overloading folk too much. There are of course issues with MSH that are shared with other databases, and folk may also want to post a general comment that they think routine prices are wrong full stop, or a wishful-thinking comment that surely we could find better sources, or a non-policy comment that we should trust editors to do original research. -- Colin°^Talk 13:37, 4 January 2020 (UTC)

As I said, I have no problem if you decide to ignore me, since it is well demonstrated that I am the world's worst at formulating an RFC. But that's my opinion :) SandyGeorgia (Talk) 13:39, 4 January 2020 (UTC)

Well I don't want anyone to be ignored. None of us is getting exactly the RFC we'd have wanted, but I remain hopeful. -- Colin°^Talk 14:31, 4 January 2020 (UTC)

I won't feel ignored; I am perfectly fine with you all proceeding without my advice. But ... if I turn out to be right, I'll get 18 Girl Scout points and big, "I told you so" :) :) SandyGeorgia (Talk) 14:37, 4 January 2020 (UTC)

After a couple days away, I just looked at the RFC page, and it looks ready for launch to me. What work is remaining? Nice work, WAID. SandyGeorgia (Talk) 18:06, 4 January 2020 (UTC)

THanks, SandyGeorgia. I agree with the sentiment that it's hard enough with just one. After we get through this, we'll probably need to have (at least) one more RFC. The next one should be broader. I hope that we'll get enough information about the community's overall sentiment that we can proceed directly to writing a section for MEDMOS on what/when/how/why to handle drug prices. WhatamIdoing (talk) 02:55, 6 January 2020 (UTC)

General, version 2

Actually, I'm feeling ignored, too. I think it will be a fiasco if that version of the RfC gets launched. And I've been saying so quite clearly for a long time now. If the consensus really is to do it that way, I won't stand in the way but I also won't refrain from expressing my concerns. I feel like a very small number of editors are, in effect, trying to run out the clock, and aren't really engaging with my expressed concerns, even though I've been revising my suggested version of the format in response to their concerns.

So here is what I would like. I hope that editors who have been watching here, or who have just started watching here, will provide additional opinions about the two draft versions: Wikipedia:Manual of Style/Medicine-related articles/RFC on pharmaceutical drug prices and User:Tryptofish/Drug prices RfC draft 2.

I'm not asking for editors who have already expressed an opinion to repeat themselves here. I'm hoping that more editors will express their opinions, giving a wider range of input. Thanks. --Tryptofish (talk) 20:52, 4 January 2020 (UTC) A part of it struck. --Tryptofish (talk) 23:04, 4 January 2020 (UTC)

User:Tryptofish Would change "Drug (medication) prices should routinely be included in as many drug articles as possible" to "Drug (medication) prices may be included in medication articles when such prices are avaliable in suitable sources." Not sure how you are wanting people to weight in. Doc James (talk · contribs · email) 07:23, 6 January 2020 (UTC)

Feedback from Sandy

OK, if you don't want other than new opinions, where do you want those of us who have been at this for a month to put suggestions and questions? SandyGeorgia (Talk) 03:38, 5 January 2020 (UTC)

I studied RFC draft 2, and went to the talk page there to enter some feedback. Tryptofish, you told us about this draft (above) on 2 Jan,[91] I took a day off, and today when I go to give you some feedback, I see you could have notified us a week sooner that you were working on it, when you notified Barkeep on 26 Dec. Then I got involved responding to Barkeep and didn't get to give you any feedback, but ... could you have told us sooner? I still don't know how to fill in a single part of Proposal 2, because I'm still waiting for an example from these databases that works. I am not aware of any way to usefully use these databases: could you fill in the examples for us? I don't know what to do next to help advance your draft. Corrected typo in date from Jan --> Dec in original post with timestamp 06:26. SandyGeorgia (Talk) 13:31, 5 January 2020 (UTC)

Sandy, I see that you have also reached out to me at my user talk page, over some of the same points, so I'm going to focus here only on the best ways for editors to work together on drafting the RfC. I'm very happy that you intend to give some feedback on that draft, and I look forward to making good use of it. And it's OK if you haven't gotten to it yet. About my notifying editors here, there was a consensus that we would take things a bit easy through January 1, during the holidays, and then get serious about it around January 2. And January 2 is when I posted here about it: [92]. It's certainly true that I started working on it before then, and I don't think there's any problem with my having done that. But I continued to work on it, on and off, in late December through January 1. So I didn't really think it was ready to show the rest of you until when I posted here on the second. I certainly was not trying to hold anything back from anyone, and I cannot imagine why anyone would think so. In addition, I actually wanted to see the other draft before I posted anything here, because I felt that if that other version were superior to my own, I would just discard my own. Again, no bad intent about that. On January 2, WAID posted her draft, and I gave my feedback on it as soon as I could, and then I notified editors here about the draft I had made (same diff above). That was all on January 2. The reason that I notified Barkeep earlier was because of this comment he had made to me on my user talk page: [93], and I felt that he should be aware that I had started working on it, as opposed to having it ready for discussion.

About filling in the draft that I started, I had hoped that editors would do that collaboratively, and I was only trying to show a format. I may be the worst person here to fill in the details, because the rest of you know a lot more about it than I do. If you and anyone else who is interested in the Prop 1 part of it want to put in whatever occurs to you, that would be great. I'd appreciate it if Doc James (or anyone else who wants to) could fill in some of Prop 2. I'll also take a stab at filling things in now. But I think that it's essential that editors recognize that one can put something into the draft and then someone else can revise it to make it better. This isn't like there's that part that says X but it really should say Y, so the entire format is unusable. Better just to change X to Y. And in one fish's opinion, WP:There is no deadline about having it ready to go live. We all are eager to be done with it and move on, but personally I'd much rather get it right than get it fast. --Tryptofish (talk) 19:05, 5 January 2020 (UTC)

User:Tryptofish how do you want me to contribute? Edit directly? Simple make suggestions? Are we allowed to use the talk page of the draft? My computer is having problems with the size of this page. Doc James (talk · contribs · email) 07:28, 6 January 2020 (UTC)

Yes to there is no deadline. But. On re-reading and recalling some parts of the activity on this page, I can see that Barkeep's Update on 1 Jan read to me (perhaps incorrectly) as if we needed to kick this into gear and get something up soon. It does seem like a month is a long time, and there is no doubt fatigue has set in (to wit, Ceylon). Looking at it from the perspective of a week later, that may not have been possible even when Barkeep posted his 1 Jan update to get this to launch, especially since the ANI close was predicated on the premise that the terms of RFC launch were: "an uninvolved admin after confirming that the question(s) are neutrally worded". In hindsight, I saw that three weeks had gone by, we still didn't have a single example to work with, so we needed to scale back what we were trying to accomplish and wrap this up.

This page has become sprawling; we seriously need to archive the older parts to a separate, RFC archive. At some point on this page, Barkeep indicated that you had presented a sample. I don't know where that is. If you have a sample, could you please add it to your RFC draft 2 page? Or should we wait to get the chlorthalidone sample clarified? As I've said elsewhere, I don't know how else to help on that draft without an example. And it's in your user space; the way we are having to work here is really constraining. Do we ALSO have to debug your draft on this page, or do we go to your user space talk?

I raised the dates because if the rest of us had known your were working on a draft, we might have focused on it; by the time we did focus, I at least felt we were under the gun to get this done. Again, water under the bridge, so onward and upward.

On filling in the draft, we are constrained that the ANI indicated Colin shouldn't do the drafting, and I believe my ability to formulate an RFC is already on display, so neither should I. What would be most helpful at this point is to get an example of how those databases can be used correctly. I am trying to work on the chlorthalidone example below, but it is in the middle of a sea text that does not belong on this page. On Part 1, I keep saying, what about something like Epipen, and ... talk about feeling ignored :) I hope you realize that there is too much volume on this page for us to all hear everything, and ignoring anyone is not intentional nor a matter of bad faith. And no, we can't put stuff in to your draft; you set it up in your userspace, and Colin can't edit it no matter where it is.

Considering it now appears we are not close to launching anything, we need to deal with the page sprawl here before we continue. SandyGeorgia (Talk) 19:29, 5 January 2020 (UTC)

Please bear with me, because I'm juggling these discussions on multiple pages now. I have no idea about Barkeep saying that I had a specific sample of that sort. What I did, outside of comments on this talk page, is User:Tryptofish/Drug prices RfC draft and User:Tryptofish/Drug prices RfC draft 2. (And I hope it's clear that I constructed v.2 based on comments to me about v.1.) About scheduling, I'm going to defer to Barkeep. Let me say for the record, and to everyone, that everyone who wants to is explicitly invited to make edits to draft 2, regardless of it being in user space. That has my explicit blessing! Doc James already made some edits there (and I reverted him and then changed it to something that I think accomplished the same thing better!). Suggestions on the talk page are welcome, too. As are suggestions here on this talk page; probably in a new section and not here. I can certainly look at and copy-paste what look like good examples amid the discussion here. And then I hope other editors will pitch in and correct the things that I will doubtless botch up. Ceylon? Sail on! --Tryptofish (talk) 19:44, 5 January 2020 (UTC)

I found Barkeep's words ... "Though I will note that at times both WAID and Trypto have presented modified versions that they feel is policy compliant. Those examples have not been satisfactory to you which is fine." It reads as if you presented a policy-compliant example that I rejected :( :( :( If that happened, I am not aware of it. SandyGeorgia (Talk) 19:53, 5 January 2020 (UTC)

I see. I'll let him clarify, but I would understand that as the two draft versions of the possible RfC, as opposed to an example, shown within an RfC, of a policy-compliant edit about a drug price. --Tryptofish (talk) 19:56, 5 January 2020 (UTC)

yea ... that's not how it came out to me. I think I am supposed to be seeing that somewhere on this page, valid examples of how to use those sources have been presented, but I have ignored/rejected them. :( SandyGeorgia (Talk) 20:00, 5 January 2020 (UTC)

User:Barkeep49, others, I think things have gone a bit off the track here. This top-level section heading was created to discuss WAID's draft and is now discussing the genesis of and participation at Tryptofish's draft. The former is imo nearly ready to roll and the latter little more than a sketch of an idea (sorry, but true). After almost a month since the ANI was closed and we were asked to form an RFC, it seems we are being dragged back to square one, still scrabbling around to find even one example to put into this second rival draft RFC. It seems we have two styles of RFC:

• WhatamIdoing: present some examples using one source, with some background on that source, and ask for comments about compliance with policy and to offer an open space to discuss guideline advice on drug prices. Emphasise that this RFC is not the final say but a step towards something.
• Tryptofish: present two polar-opposite positions on drug prices in articles and appropriate sources, along with some samples and (presumably brief?) arguments pro/con. And then ask everyone to support and/or oppose both options separately. Aim to settle the matter in one go.

The point of having the RFC is that we do not claim to know community consensus on this matter, and since polling is not a substitute for discussion, it seems wrong at this point to go for a simple poll, and to offer a false dichotomy that voters must pick positions on. Polling is known to polarise debate and separate voters into adversarial factions, so this would not seem to be a wise option when we have already seen incendiary claims of collusion with Big Pharma to censor Wikipedia and conceal prices from patients. We don't need closed-minds formulating some sound-bite that will destroy the opposition's argument. We need open minds to look honestly and carefully at this complex issue. So complex, that I think we do need to concentrate for now on one source and one region (developing world). I keep saying, please lets choose the least-conflict option, and an RFC that divides Wikipedians into two camps to try to outvote each other is not that, imvho. -- Colin°^Talk 21:43, 5 January 2020 (UTC)

My thinking is that Trypto's Draft will need an example, which it appears from a month of discussion may not exist. WAID's draft looks ready to go, and while it answers a much lesser question, at least it answers one question in the overall picture. I can see a scenario where we launch the version we have ready to go, and depending on the outcome, Trypto's becomes the followup to address other questions. I think we can do both here, but agree we need to maintain focus on finishing the one that is ready. SandyGeorgia (Talk) 21:50, 5 January 2020 (UTC)

Sandy, when you and I started discussing things earlier today, I became very optimistic about the direction things were now heading. Reading the newest comments here, that optimism has significantly diminished. I don't think that the community will respond favorably to having a preliminary RfC to be followed by another one later. And let me be very clear: have a fail on whatever is the first RfC, and it's going to be "ArbCom here we come". As for an example that may not exist, I think I addressed that just above. And as for Colin, if the ANI consensus was that he should not be the person to formulate the RfC, maybe he is also the person who should not be trying at every opportunity to shut down the discussion. Yes, I said that, and if anyone does not like it, I really do not care. Barkeep, if you think this is the wrong header level, or something like that (oh, the horror!), please feel free to refactor the page sections. But I don't think this talk page was created to discuss only one possible version of the RfC. The idea that the format I recommend is presupposing the view of the community or closed minded or polarizing, whereas the other one is nice and neutral, is laughable. --Tryptofish (talk) 22:21, 5 January 2020 (UTC)

Just to make it clear, there was not an "ANI consensus [] that [Colin] should not be the person to formulate the RfC". I thought someone had indicated that at some point (at ANI or here) but I may be mistaken because I can't find it. I am not going to add or edit any RFC text directly. Wrt "closed-minds" "open-minded" I should clarify I'm talking about how any RFC makes voters/commenters react, not about anyone here drafting an RFC. Hence, closed vs open question in the RFC. I am not going to respond to the other points here. -- Colin°^Talk 10:08, 6 January 2020 (UTC)

Re, Just to make it clear, there was not an "ANI consensus [] that [Colin] should not be the person to formulate the RfC". This is correct. There was one comment in the ANI: There may very well be problems, but I'm unconvinced that Colin is the best person to be the primary one drafting an RfC to deal with them. Nil Einne; I believe that Colin has acted correctly in respecting the spirit of that one comment, and leaving the actual drafting to other parties. I have full confidence in WAID because of her RFC experience and long-evidenced neutrality in all matters WPMED. She has done a fine job in laying out one part of the problem that needs to be examined before we can move to the broader issue of what to include in MEDMOS, while not giving respondents too much to deal with in one shot. Trypto, I hope you will back off on the term "laughable" and that we can move forward step by step, which is something we've discussed many times on this page. SandyGeorgia (Talk) 11:15, 6 January 2020 (UTC)

Explain to me please what you mean by "have a fail" on a first RFC; I'm not following. Could you give me a timestamp for what you are referencing with "I addressed that just above", so I know the reference? (We now have James participating to help generate an example, is my understanding of where we stand on that.) And since we have multiple times on this page talked about needing more than one RFC, I am confused (and saddened) that this has caused your optimism to diminish. I thought that possibility has been on the table for weeks?? It sounds like you are now saying that your RFC is the only acceptable one in your view ?? SandyGeorgia (Talk) 22:30, 5 January 2020 (UTC)

If the community finds that the first RfC is flawed and reacts negatively to it, that will be a fail. Timestamp: 19:05, 5 January 2020. I have never said that this is about me wanting it to be my proposal, but I am definitely saying that I expect to be able to have my comments discussed reasonably instead of being treated the way that is happening here, and I am trying my hardest to avoid having such a fail occur, which is above all what I really do want. --Tryptofish (talk) 22:42, 5 January 2020 (UTC)

OK, I understand your "fail" concern; it strikes me that WAID has put a lot of thought into it and that doesn't look likely to happen. But we know my record. At 19:05, amazingly a ctrl-f shows three of them on this page (we need an archive), but I located your post, and it follows what I am saying: we think/hope James will provide an example. OK, the page is a mess, and we are all possibly talking past each other, but you clearly feel you aren't being heard and you aren't getting breathing room in here. My plan: I will continue the discussion with James to try to get an example, work on trying to get the page archived, answer any direct queries or anything Barkeep asks, and otherwise back out of other discussions for a few days. I don't want you to feel ... ignored, disappointed, etc. I hope that will help. It's risky for me, because the last time I unwatched, and completely missed Draft 2, then you felt ignored. I'm at a loss :( SandyGeorgia (Talk) 23:12, 5 January 2020 (UTC)

Much better, thank you for that. I mean it. As you were writing this, I have been filling in the requested stuff at draft 2, insofar as I understand it. It will still require a lot of vetting and fixing by other editors here, but at least this is getting the ball moving. --Tryptofish (talk) 23:28, 5 January 2020 (UTC)

OK ... just to make sure you understand ... except for two initial and very simple questions on WAID's draft, I didn't help there at all. Equally, I will back out and not help on the wording of your draft. Not "ignoring" you, but just giving equal treatment to each. Deal? When (if ???) we get to a point of (I guess?) Barkeep asking for final opinions pre-launch, then I'll re-engage on wording, if I see anything significant. SandyGeorgia (Talk) 23:35, 5 January 2020 (UTC)

As far as I'm concerned, we are all volunteers. At this point, I have filled in possible examples for everywhere editors have asked for examples. Very likely, I made significant mistakes. But those mistakes can be corrected. I'm tired now, and will work on the "rationales" tomorrow maybe. In the mean time, everyone should feel free to make changes or suggestions if they want to. --Tryptofish (talk) 23:39, 5 January 2020 (UTC)

I don't think this talk page was created to discuss only one possible version of the RfC

By my count, at least four different RFC models have been discussed on this page (questions that would let us formulate text for MEDMOS, Tryptofish's two userspace drafts about voting on dueling texts for MEDMOS, and the narrower RFC on whether the existing content is our best work), so that's not likely that this page is only about one possible version of the RFC. ;-) The narrower RFC that I have drafted at Wikipedia:Manual of Style/Medicine-related articles/RFC on pharmaceutical drug prices was not originally my idea, but I no longer remember who first mentioned it.

I think my reply at Wikipedia talk:Manual of Style/Medicine-related articles#Potential accusations of bias applies here, too. WhatamIdoing (talk) 03:15, 6 January 2020 (UTC)

Collapsing as suggested by Barkeep49 Nil Einne (talk) 13:15, 6 January 2020 (UTC)

I got pinged here. I'm not going to read the whole discussion but I've read some brief snippets. I will first say I don't see anything wrong with Colin being involved in a community drafting process. I do think it will be a mistake for them to be the one to draft an RfC on the matter by themselves. Frankly, I think it will be a mistake for anyone to draft the RfC by themselves but from what I saw at ANI, I stick by my view that it will especially be a mistake for them to be sole drafter since the way they approach things IMO doesn't help earn support. OTOH, if most others, including those like Doc James and others who have been in opposition to Colin's views, feel it will be best for Colin to draft the RfC by themselves, then I defer to their view since they surely know far better than me. Also, while I understand the need to avoid excessive complexity, I would think carefully before using 2 RfCs. From what I've seen in the past, by the time a second RfC rolls around, there tends to be a great deal of 'not this crap again' i.e. community frustration and disengagement from the process. If editors do feel it's best to hold 2 RfCs, then at the very least this requires very careful planning and engagement IMO. Finally, I guess there's something major I'm missing since I don't understand how there can be any attempt to 'run out the clock'. This is not an issue which is likely to go away anytime soon or one which is most urgent at a certain time. (E.g. I can understand how something on US presidential elections will be less important in a year's time.) Editor's should give ample time to any discussion to come up with the best RfC taking on board everyone's views as far as possible, preferably reaching a clear consensus. (I know it may be weird to talk about a consensus on wording an RfC and I'm not suggesting people do an RfC on the RfC wording. But this page seems to have enough participants with diverse viewpoints that I would hope something good is achievable.) In some cases maybe even an informal straw poll will be useful although I'd also urge caution since I'm sure we all know it shouldn't be a substitute for discussion. Edit: Read a bit more. If editors are starting to get frustrated with this process and disengage I agree that this is a problem. I think this is a case where there's a delicate balance act between not letting perfect be the enemy of good; and not producing something which will leave those involved feeling their concerns were ignored or produce an outcome with is disputed or unclear. Edit 2: I should perhaps clarify that as my concern is with Colin specifically being the sole drafter, they may alienate everyone from supporting their view by their approach. So frankly if those who, for lack of better word, "support" Colin want them to be the sole drafter, that is good enough to allay that concern as long as they understand if things go pear-shaped they shouldn't blame the community. But this won't deal with the fact that an RfC drafted only by Colin with no feedback by anyone else may be biased or otherwise problematic, the same as with an RfC drafted only by Doc James etc. Nil Einne (talk) 11:38, 6 January 2020 (UTC) Thanks for commenting, Nil Einne. I think we've had enough words said on a hypothetical situation that is going to remain hypothetical. Please lets return to discussing the RFC rather than specific editors or admins. -- Colin°Talk 12:42, 6 January 2020 (UTC) @Nil Einne: Apologies but I was pinged here and I don't know why. I don't think I've visited this page, let alone commented in a long time. I think the ping was because of something I said at ANI about you drafting the RfC but frankly the discussion is so long that I don't think I can be bothered figuring it out. If it is only hypothetical then I'm unsure why I was pinged, since I'm fairly sure I never said you could not be involved in drafting the RfC along with the community. If I did, I apologise without reservation. In any case, I wanted to make my views clear if they're being discussed and misunderstood. Or if they were correctly understood, I felt it best to re-iterate the point I intended to make. I stand by my views expressed here and I think at ANI. But do agree it's not that important since ultimately it will pan out how it pans out. If mistakes are made in the drafting which destroy any result, well all I can say is me and others tried to help in suggesting the best path forward, it's unfortunate if it wasn't followed. I will collapse this discussion as suggested by Barkeep49 somewhere above. Nil Einne (talk) 13:15, 6 January 2020 (UTC) @Nil Einne:, I pinged you because I did want to make sure we had stayed true to your concern, so your feedback was useful. Sorry if my ping created discomfort. I really despise this new-fangled pingie-thingie, as it creates so many side issues (recalling that you have tried to help me learn how to use it, but now I want to unlearn it :). In the "olden days", I would have gone to your talk page, and notified you that I had mentioned you, and explained why. I may go back to the old ways. Thanks for weighing in, SandyGeorgia (Talk) 13:26, 6 January 2020 (UTC)

Discussion about how this was worded

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

---

I'm having a hard time with the phrase "run out the clock" after we typed ourselves blue in the face for a month ... ??????? Many many many times I asked, please name the RFC, put up a mock up; you didn't, WAID did. I feel that a month is a pretty long clock. SandyGeorgia (Talk) 22:32, 4 January 2020 (UTC)

Well, I put up two such markups, but let's not let the facts get in the way. As I said, I'm not asking for editors who already expressed an opinion to repeat themselves. --Tryptofish (talk) 22:36, 4 January 2020 (UTC)

Barkeep49 I know you said that certain concerns should be done on a user talk page rather than a guideline talk page, but I'm prevented from doing that. I have to say I'm finding these allegations by Tryptofish to be rather against our assume good faith policy, and not conducive to collaborative working. Trypto, here you are frustrated that I haven't helped fill in your skeleton RFC and suggest I'm deliberately leaving it bare in order to comment that it is unfinished. The allegation that folk are "rigging the RFC to get a desired result", and query to WAID about her response to a potential complaint raised "that the RfC is hopelessly biased and should be discarded" do not make me feel comfortable. Your "let's not let the facts get in the way" is fightin' talk, Trypto, please tone it down. Some of us have been at this since October! Let me be clear: I am not going to write any RFC, and I don't think James should write one either. I can express my opinions about the content/focus of the draft RFCs and you guys can accept or reject it same as with any other editor here. I think we are all rather tired, and frustrated at the lack of participation by experienced editors in helping drafting any RFC, but lets please collaborate and compromise, not fight among ourselves. -- Colin°^Talk 22:45, 4 January 2020 (UTC)

Well, that's one way to frame it. In any case, I think this demonstrates that it would be useful for more editors to participate in the discussion if they want to. And if consensus is against me, I won't stand in the way. --Tryptofish (talk) 22:52, 4 January 2020 (UTC)

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Post-ANI Discussion and Clarification

There has been a fair amount of requests for clarification and interpretation occuring at my talk page (see [94]). Out of respect for the consensus that "The question of drug pricing is remitted to a single venue" I note the above diff and then copy over the current discussion that had been occurring there to here to be continued. Best, Barkeep49 (talk) 15:51, 3 January 2020 (UTC)

Content from Barkeep49's Talk Page
The following is a closed discussion. Please do not modify it.
Just a note to those around here, I was just asked on DocJames talk page specifically about updating. I had discussed this previously with two other uninvolved syosps on IRC and all three< of us feel that updating information already present in an article does not qualify as "adding or removing pricing" and can be done even while the RfC creation process plays out. courtesy pings to @Ronz, Colin, SandyGeorgia, and QuackGuru:. Best, Barkeep49 (talk) 06:25, 2 January 2020 (UTC) I think that tagging/amending/updating prices on articles should be included in the moratorium. While I have no intention of editing the prices at present, the contentious issue is not just whether prices are included or not but that existing prices make false claims, do not represent what the source says, cherry-pick one database record rather than another, use prices drawn from buyers and suppliers, etc, etc. All these issues have been a source of conflict over the years. I wouldn't support the idea that James alone is permitted to modify their own price statements, and all evidence suggests that anyone else amending them will result in conflict. It would also be highly disruptive if editors modified the article text of drugs being used as RFC examples such that they deviated from the examples presented at RFC. We don't want the RFC discussion on what should be said to be played out by edit warring on the articles. It would simply be better that any statement about prices be left alone for now. -- Colin°Talk 10:40, 2 January 2020 (UTC) Thanks for the ping, Barkeep; I am following the discussion at Doc James talk. 1. These personal exceptions for Doc James are part of why we are where we are. It is awkward that we can't consolidate conversations because of a) prohibitions on pinging/posting and b) lack of engagement in one place. One point of the ANI close was to prevent disputes from spreading. At ANI, "The question of drug pricing is remitted to a single venue". Full Stop. Your words. I agreed with JzG on these conditions instead of supporting a topic ban on Doc James as requested by another sysop (Nil Einne). Doc James has not come to the single venue to discuss, and one party is prohibited from posting to Doc James talk, and yet that party can be discussed there. This is a formula for dispute-spreading. There should be, in fact, no discussion of the price dispute at James talk, other than your notifications and further clarifications with James and Peter. 2. I have always been and still am opposed to back-channel decisions on IRC. (That is not "a single venue".) Conversations regarding a sysop who is continually given exceptions for behavior (eg edit warring) should be in full view of everyone else affected. Who are the two sysops who agreed that we can grant this change to the clearly established conditions? By conducting business off-Wiki, your excellent conduct in this matter so far puts you at risk for being drawn in to the precise pattern that needs to be addressed. Please provide the names of those sysops. 3. My recommendation is that you walk back this private discussion and decision, and leave the ANI close as was clearly agreed. I agreed with JzG's approach only because it was so carefully worded, and having been down this road before, knew what might happen. It has. Doc James breached the close. Please respect your own ANI close, and do not complicate a years-long dispute with non-public discussions where three sysops decide to exempt one sysop from a community agreement. OWNERSHIP is at the core of the pricing dispute and all previous disputes: exempting one editor, who has not engaged with the rest of us who are attempting to solve the problem, from a community-wide decision which accounted for this very possibility furthers the very problem we are attempting to solve. SandyGeorgia (Talk) 13:55, 2 January 2020 (UTC) @SandyGeorgia: I was one of the editors who was consulted on IRC. The off-wiki consultation was clearly disappointing for you, and I'm sorry about that and will aim to be more transparent in this matter. It probably won't make you feel better, but I was presented the question in a very abstract sense and was not aware of the identity of the participants, and certainly did not intend my comments to be about giving a particular sysop an advantage or special exception over others. My reasoning was that the spirit of verifiability doesn't allow us to keep old bad information if newer, equally- or better- sourced information is available -- it must be updated or removed, and the closure explicitly prohibits removing it, so updating it is the only option. Additionally, updating pricing doesn't seem to involve the same dispute as to whether prices should be included or not. Thanks for your comments -- I hope I have addressed some of your concerns, and if not, please don't hesitate to reach out further. Best, Kevin (aka L235 · t · c) 18:22, 2 January 2020 (UTC) it must be updated or removed Howso? What's the hurry? Is there some BLP-like requirement? --Ronz (talk) 20:27, 2 January 2020 (UTC) @SandyGeorgia: thanks for all your thoughts here. I want to acknowledge that I've seen them. I think some of your criticisms of me are more than fair but want to take a few hours to think on them before replying (especially because other pieces of the analysis I respectfully disagree with at first blush). Best, Barkeep49 (talk) 15:40, 2 January 2020 (UTC) For today at least, pinging me will only increase my need for a calming cup of tea. That "tipping point" has been passed by sysops allowing the very behaviors we seek to address, to fester and grow by a participant who has not engaged. This Is The Pattern. Regular editors, who want to add medical content, have no place on this project. We have competent, qualified medical editors who have written numerous guidelines, featured articles and featured lists, but who do not have a sysop flag attached to their account, who are doing everything they can to assure medical content on Wikipedia is accurate, and being undermined by those who have a flag attached to their name. This is why good editors quit. This is why medical content is suffering. This is why good editors no longer engage to remove vandalism and quackery. This is why good editors stop trying to bring medical content to featured status. This is why I stopped editing for years and unwatched hundreds of medical articles. This is Wikipedia. SandyGeorgia (Talk) 16:02, 2 January 2020 (UTC) Sandy, (who I'm intentionally not pinging based on what she wrote above as there's only so much calming tea in the world but who I will ping tomorrow) thanks for all your thoughts. As I've expressed to you in the past, when you were working on the edit request, I knew that being the sole sysop involved in these issues was going to be trouble sooner or later. And now we're here, in part, because I took a half measure in this regard. There had been an increasing number of decisions that I had been asked to make about this situation. While the all the decisions had been respected I had become uncomfortable being the sole uninvolved sysop attempting to properly carry out the community's will in this regard. What I should have done was to solicit wider feedback formulate a different version of the update I'd posted and then reverted at AN. What I did instead was to reach out to others on IRC. I understand why this opaque measure rubbed you the wrong way. While I appreciate L235 chiming in, what I should have done after making the first mistake of consulting on IRC rather than AN was to just own the decision as my own - as I felt that updating prices was ok before the discussion and after. I included the IRC mention as a way of being clear what I had done but rather than being transparent it diffused responsibility for the decision in a way that it shouldn't have. Why was updating ok with me when I'd said even adding a tag was not? In my thinking it's different because it's not adding something new and because the embargo is designed to ameliorate the conflict not preserve bad information. In the time that I've been working on this conflict I've worked hard to treat all editors with respect. Especially because the core group of editors, whether they carry the sysop user right or not, deserve immense respect for the longstanding ways they've made Wikipedia a better place. I can, and will if you or some other editor would like, explain how I've tried to treat all editors the same regardless of sysop status. But essentially my approach has been to make a polite request for specific action where I've observed a line being crossed. So far every time I've made a specific request to take/rescind an action the editor has agreed. I have also, attempted to follow the ANI close statement that there will be "no rehashing of grievances" by acting on not what's happened in in the past with editors but what's happening now. All of this has applied to all editors including DocJames. But I also remember acutely what it's like when you don't have the sysop flag. I wrote, in a line I had to remove from my ACE statement due to space constraints, "I remember what it’s like to feel put down not or otherwise dismissed because" I wasn't a sysop. We lose so many good editors for so many reasons and I'm sorry that the conditions here are such that we're going to (potentially) lose you. The places you've chosen to contribute in this dispute have been made better because of your contributions. I can only hope you decide Wikipedia remains worth it. As to the specific asks you've made, I have struck my statement here and on Doc's talk that refers to the discussion IRC. I am willing to discuss the decision to say that updating is OK further and have now laid out my thinking rather than just giving a "because me and two people you can't comment about said so" reasoning. Just to reiterate, updating feels different than adding or removing because the ANI close was about ameliorating the dispute and where the information is already present updating serves our readers without changing the scope of articles involved in the dispute. If you don't want to discuss it further we can all head back to AN/ANI or go to ArbCom as has been discussed below. If we head to AN/ANI (and obviously if it ends up at ArbCom) it will likely mean I step back from attempting to moderate the dispute for at least a bit to give the community space to weigh-in and indeed offer feedback on my own actions. As for the final bit of keeping discussion at WT:MOSMED you're absolutely right. It has, at minimum, spread to this page, Doc James' user talk, WT:Prices, Talk:Ivermectin. I'm happy to take a more assertive stance on pushing stuff over there (for instance I should have replied to several of the questions posted here there). As you note some conversation will still be needed at on user talks but this would be much more one on one discussion. To that end if you (or others) want to continue conversations about updating with me, we should create a section to do so there. If you wish to discuss my actions that would of course remain more appropriate for this page (or AN/ANI/ARBCOM). Best, Barkeep49 (talk) 22:29, 2 January 2020 (UTC) I am back from Ceylon; I found life in a country where pings are outlawed very relaxing. Do you want me to respond here, or do we want to keep discussion at WT:MEDMOS? My suggestion would be that we continue this particular aspect here, in the spirit on not splitting a conversation. SandyGeorgia (Talk) 12:41, 3 January 2020 (UTC) Barkeep49, I think your decision to permit editing of price information is a rational one, assuming the conflict is only about addition or subtraction (or moving to body), and assuming that updating the text might change it from "bad" to "good". That really isn't the case; it is just differently bad. Attempts to resolve some of the OR or false claims have been rejected by reversion and on several occasions caused protracted dispute. I only see that being more likely during an RFC. You say you are not keen for an embargo to "preserve bad information" but no amount of tinkering or updating data from 2014 to 2015 is going to make the information good. Nearly all of the prices are many years out-of-date, wrong and misleading. We've lived with that for years and really it is best if they are left be for a little while longer. -- Colin°Talk 13:12, 3 January 2020 (UTC) We may be in disagreement about the "rationality" of the decision, Colin; I have very big problems behind the logic in this decision, and am surprised no one has seen the logical flaw. I am waiting to hear if this is the right place to have that discussion, and to make a recommendation. SandyGeorgia (Talk) 13:18, 3 January 2020 (UTC) FYI, User:Colin/PriceEdits contains a computer-generated list of all price/cost insertion/deletion edits to 530 drug articles by any editor since 2015. It also contains my analysis of where editors have come into conflict. Plenty examples of why I note my concerns above. To take an example of the kind of "copyedit" being suggested by User:WhatamIdoing at the MEDMOS discussion, an editor changing "the wholesale price in the developing world is" to "The median buyer price according to the International Drug Price Indicator Guide was" sparked an edit war at Lactulose. -- Colin°Talk 10:40, 2 January 2020 (UTC)

I look forward to learning about the flaw that Sandy sees with allowing the updating of information. Best, Barkeep49 (talk) 15:51, 3 January 2020 (UTC)

Acknowledged, but I am off now for several hours. I will post once I am home (unless I get in an accident on snowy roads :( SandyGeorgia (Talk) 16:05, 3 January 2020 (UTC)

Barkeep49, I'm not sure this move is a good one, though I see why you did it. I am uncomfortable with complaints about an admin decision as to whether recent edits break AN/I closure (or similar edits will so in future) being made on a guideline talk page, rather than the talk page of that admin. Isn't that what you keep reminding us? I would prefer if your user talk page was a safe space to raise concerns about current edits and your moderation of the topic discussion, and we leave this page to discuss drug prices on articles and what an RFC might say. -- Colin°^Talk 16:10, 3 January 2020 (UTC)

I actually was hoping, too, that it would stay at Barkeep's page, but for a different reason (I had no plans to rewatch this page with all parties not even coming to the table ... I have done my share, and it was a waste of my holiday season, as things resulted). I'd be much happier if this separate discussion stayed where it was. Gotta go, now I will be late and driving on icy roads, whatever you all decide. SandyGeorgia (Talk) 16:15, 3 January 2020 (UTC)

Sorry to drag you back here and be safe. Best, Barkeep49 (talk) 16:35, 3 January 2020 (UTC)

Colin, way to fairly throw my own words back at me :) I think this is best for a couple reasons. I think discussion around these issues are tied into the larger discussion and so the ANI consensus needs to be respected - plus I have a pretty thick skin. Some number of editors who might be interested in this topic might not know to go to my talk page for hidden pertinent discussion. Additionally, I would love if other uninvolved sysops engaged in helping to moderate this dispute and putting this here makes it easier for them to weigh-in as necessary. Finally, I think it helps send the signal that were I to close a discussion on some other page saying it should be had here, that I'm applying the ANI to all editors - even myself. I am, of course, willing to admit when I make a mistake. If it proves that having these discussions here make things worse than my user talk well I'll reverse course again. Best, Barkeep49 (talk) 16:35, 3 January 2020 (UTC)

Responses to post-ANI feedback

I'm back; thanks for the well wishes. While I was driving, I tried to organize my thoughts to avoid my usual verbosity. It is unfortunate to clog this page with this feedback, but I understand Barkeep's reasoning for wanting to keep this here.

I would like to unpack my concerns in steps, so as not to drop a wall of text filled with my usual typos on this page, and for us to be able to calmly digest the different components.

1. Misunderstanding about my charges above
2. Questions/concerns about 3 x IRC
3. Logic of the decision applied
4. Has everyone been treated equally?
5. Where do we stand, and how does that affect where we go next?

Starting with No. 1, I felt completely betrayed yesterday. Usually I apologize for an outburst, but I hope no one feels I have something I need to apologize for, because I pushed back from the computer and hit the break button before I completely exploded. I always apologize for misunderstandings when my wording is not careful, so in that vein, I am sorry if Barkeep felt I was questioning his integrity or saying he had disrespected me or us or done something intentionally deceptive or malicious; that is not the situation.

But yesterday did not feel like this month was well spent. The ANI started on December 5, and today is January 3; we expended the entire holiday season on this, and we all typed 'til we were "blue in the face" with unfailing good faith and very little need for, as Barkeep said, people to walk things back. (I will discuss in part 4 whether some people should have been asked to walk things back, but my concern there is not aimed at Barkeep.) Those who participated here (which I will later distinguish from the MEDLEAD RFC) showed that they can address this dispute calmly and without behavioral issues, which is part of why I do not feel we are yet to the arb stage. As we have all expressed (and I believe sincerely meant), Barkeep has gone above and beyond the call of duty and has been exemplary in his treatment of all of us. Yet, through no malintent or negligence, we came yesterday to a most unsatisfactory juncture, for reasons I will unpack bit by bit here.

Basically, I feel our good faith was taken advantage of. I believe the problem yesterday happened because Barkeep has been shielded from full information of what he was walking in to, by necessity, because we are not allowed to "rehash old grievances". So, Barkeep could not have foreseen yesterday's reaction to something that, to many of us, was entirely foreseeable, because it precisely fits the pattern we have been dealing with. I'll unpack that in part 3.

When Barkeep posted to AN, I cheered him, because it has become obvious that he was working too hard. When he retracted his AN post, I worried. I find it a bit disgusting that there are probably twice as many people who participated at the ANI than have participated at the MEDLEAD RFC, and it is unconscionable that what I asked in the ANI (when I laid out my terms) that more eyes watch these events, and Barkeep closed by asking for more volunteers, he got none. I would like to revisit the ANI when I get to Part 5.

Kevin, thank you very much for stepping forward with reassuring information, but I never thought that Barkeep's conduct back-channel needed to be questioned; he has given no reason for doubts about his integrity or how he would approach other admins. My concerns had to do with him not knowing the full picture, and perhaps unwittingly falling into something he might be unaware of.

And that relates to Part 2. Barkeep, I have emailed you a description of things that have happened to me in the past on Wikipedia-- many years past-- as a female editor. To be able to move forward, I would like to ask if either you or Kevin know why the third admin on IRC has not self-identified as Kevin did. I hope you can view my question in the context of the private information I sent you, and understand that it would be reassuring to me if that admin would self-identify. If they won't, that will force me to ask the question in a less-than-desirable way. Meanwhile, I hope we'll all give a good re-read to the ANI thread, because I want to next raise some specifics about where we stand. That's all for now; Barkeep and Kevin, please let me know if it is likely that the third IRC admin will speak up. I'll continue afterwards with Part 3. Regards, SandyGeorgia (Talk) 22:34, 3 January 2020 (UTC)

Moving on to Part 3: logic of the decision applied (although we haven't heard from and don't know the logic of the third admin, or whether they have been previously in conflict with anyone here).

The ANI close put an "embargo on adding or removing pricing during this process". James broke the embargo, but was not sanctioned (we will deal with that in Part 4).

Barkeep said: Why was updating ok with me when I'd said even adding a tag was not? In my thinking it's different because it's not adding something new and because the embargo is designed to ameliorate the conflict not preserve bad information.

And Kevin (aka L235 said: My reasoning was that the spirit of verifiability doesn't allow us to keep old bad information if newer, equally- or better- sourced information is available -- it must be updated or removed, and the closure explicitly prohibits removing it, so updating it is the only option. Additionally, updating pricing doesn't seem to involve the same dispute as to whether prices should be included or not.

So it is easy to see that both Barkeep and Kevin are operating from a position of wanting to preserve the integrity and verifiability of information to our readers, while minimizing conflict. So far so good. Well intended, makes sense.

But here's the problem. We have reams of evidence above that all of the drug price information we have presented to our readers in these 500+ examples is bogus. We have computer-generated analysis that shows essentially all of the prices were inserted by one editor. I have been asking for a month, and we have not yet seen one single example of these database sources being used correctly. So, questions are:

1. How do we protect the integrity and verifiability of content for our readers by disallowing the tagging of bogus information for the benefit of our readers, while
2. Allowing the very person who inserted all of the data to begin with to go about correcting it, when we have no indication that editor has engaged to understand why the data is a problem?

How does that make sense? It doesn't help our readers, integrity of the content, or verifiability at all, and gives no one who wants to alert our readers that there is a verifiability problem with the data that ability, at the same time that the very person who made all the errors was granted an exception to breaching the ANI close. Perhaps a bot could go through and tag every instance in Colin's data with a dubious tag, while at the same time including a link to the talk page of the article, where the ANI embargo would be explained. James should not be touching this data, as we have no indication that he understands the errors he has made. But it needs to be tagged for the benefit of our readers: that is amply evidenced in lengthy discussions on this page.

Have the admins who made this decision understood that the issue is not just whether we allow pricing data in drug articles; it is about data that has been shown to be 100% bogus, and as of now, is a disservice to our readers, full of multiple kinds of errors, entered by one person only? SandyGeorgia (Talk) 19:27, 4 January 2020 (UTC)

So I better understand why you were hoping to keep this on my user talk because it's a little more relaxed to talk about an editor without naming that editor. I appreciate that you're trying to walk a fine line here. I'm waiting to respond more fully until you've completed all five parts. Best, Barkeep49 (talk) 21:29, 4 January 2020 (UTC)

Sample of price calculations

@Doc James: Would you like to reply to any of that? --Tryptofish (talk) 21:38, 4 January 2020 (UTC) struck --Tryptofish (talk) 21:52, 4 January 2020 (UTC)

User:Tryptofish what claims are you wanting me to address? If people want to discuss how to summarize this source I am happy to.[95] If they are saying that this source is unsummerizable or simple false well that is a different discussion. Doc James (talk · contribs · email) 02:44, 5 January 2020 (UTC)

@Doc James: Anything at all that you can provide useful information about, particularly if it helps in creating the best possible RfC we can produce, would be most welcome, in my opinion. As Sandy notes correctly below, however, it would be best to have such a discussion in other sections of this talk page, and not here.

My reason for pinging you had been, however, the comment about whether you do or do not appear to understand. I have concerns about that comment. Barkeep replied to me (the comment that is now below) that this would not be helpful to discuss here, so I struck the ping to you. When I logged in today, I was planning to raise the issue again at Barkeep's user talk page, but I see that Sandy has very kindly reached out to me, so I'm just going to pursue it with her at my own talk page. --Tryptofish (talk) 18:33, 5 January 2020 (UTC)

User:Tryptofish still not clear what you are asking about? Doc James (talk · contribs · email) 07:25, 6 January 2020 (UTC)

James, Trypto originally pinged you to my post above (19:27 on 4 Jan); that post ended up on this page from a moved discussion, and continuing discussion of some aspects of it does not belong on this page (behavior conduct issues rather than content/development of RFC), which is why Barkeep asked Trypto to strike the ping to you. To sum it up, we are working now to see if you understand the levels of problems in the price examples. I have had some concern that, since you have forbidden interaction with Colin (and you appear to read only posts that ping you), there exists a very real possibility that you aren't even aware of the concerns we are discussing. SandyGeorgia (Talk) 11:54, 6 January 2020 (UTC)

Yes I understand clearly that these prices are rough estimates. They are for one common dose at one point in time from one or more buyers or sellers in LMIC at the wholesale level. It is not the price for every dose and at all point in time or at the point of sale with all taxes etc included.

I also equally understand that medicine is not an absolute science but deals in shades of grey with most of what we know being approximates at best. That when medications are studied, the population included in these studies do not generally match the patient in front of me. Those in studies are generally younger, have way fewer comorbidities, are less often women, and many not be of the ethnic background of my patient. That simple because a p value is reached, does not mean a clinical benefit will occur, or that a lack of a p value being reach means the medication is useless etc, etc, etc. And yet we still include these details in Wikipedia without generally stipulating that these are not universal truths for all people at all points in time.

In fact it appears that what is being requested for prices of medications is something we do not demand of other areas of medicine. Yes it would be wonderful if medicine and economics were as precise as physics or math, but it is not. Providing rough estimates of benefit of a medication is similar to providing rough pricing / cost information of a medicine. Both are useful IMO and why I spend my free time adding them. Sure we disagree about the benefit of these rough estimates.

Now if people are interested in how we should make these rough estimates better I am happy to discuss this. If you feel providing any rough estimates for any where in the world is impossible well than I guess we have found our fundamental disagreement. Doc James (talk · contribs · email) 12:25, 6 January 2020 (UTC)

James All of that is helpful, and we know that you understand all of that, but we are trying to present something useful and accurate to our readers, who are not necessarily statisticians, physicians, pharmacists or economists. IMO we have not done that so far with drug prices.

I picked the simplest example from those I've seen so far, using a common medication (chlorthalidone), but the other examples we've discussed have bigger issues than the chlorthalidone example. I picked chlorthalidone because I've dealt with it so much as a Spanish-language interpreter in a free clinic, and because that example of how we represent cost is not plagued with as many SYNTH problems as we see in other drug articles. (Interpreting is tedious repetitive work, because working with dozens of different doctors, you end up having to tell different patients the same thing over and over, because you have to exactly repeat what the doctor says ... it would be much faster for me, the doctor, and the patient, if I just said, OK, do you want me to explain to them in my own words that they must cut their chlorthalidone tablet in half, but the standards and ethics of interpreting don't allow for that.) Chlorthalidone, in agreement with the Cochrane review, is commonly prescribed at 12.5 mg for hypertension, in my experience, without variation. We have presented one cost, based on 25 mg x 30, without considering different doses, what is the most common use, and for what indication we are giving that cost. Perhaps it would be more helpful if we continued all example discussion below in the new section you have started on DDD. SandyGeorgia (Talk) 13:17, 6 January 2020 (UTC)

Yes it is a rough cost. Sure some people take 12.5 and some take 50 mg. Do you want me to provide a price range for 12.5 to 50 mg? I can do that. I am happy to do that. It will still be accurate and useful if I do that. Doc James (talk · contribs · email) 13:27, 6 January 2020 (UTC)

James I cannot tell you what range to pick; that is original research/synthesis, and presenting information not supported by the source. What we should be looking to achieve (and the reason I picked this very simple example) is to present something useful and accurate to a common (layperson) reader of Wikipedia. We should not be using SYNTHESIS from multiple sources to do that, especially when our synthesis (in some examples), goes way beyond simple math, and includes a multitude of assumptions that are not transparent to our readers. Please have a look at the WP:CALC section of NOR, and let's continue this below in the new section you started. The other examples have many more problems than whether we are reporting 12.5 or 25 mg prices, and whether that is for hypertension or diabetes. SandyGeorgia (Talk) 13:38, 6 January 2020 (UTC)

You do not need to tell me what range to use. That is easy to find in the references. Multiplication is simple math. Doc James (talk · contribs · email) 13:54, 6 January 2020 (UTC)

The discussion is continued below; continuing this in two sections is unhelpful. Please understand that when you are synthesizing information from multiple sources to decide how to do that math, it is not simple math. SandyGeorgia (Talk) 14:02, 6 January 2020 (UTC)

Doc James, I will give you a simple example. How did you derive the specific information entered for chlorthalidone?

In the United States the wholesale cost is about US$13.50 a month.[96]

SandyGeorgia (Talk) 03:06, 5 January 2020 (UTC)

User:SandyGeorgia Sure. Cost per 25 mg tab of chlorthalidone is 0.45 USD per the ref. 25 mg is a commonly used dose[97] and the defined daily dose.[98] There are about 30 days in a month. 30 * 0.45 = 13.50

If people want the exact calculations for ever one happy to put them in as comments with refs to the DDD. Doc James (talk · contribs · email) 10:13, 5 January 2020 (UTC)

The above "calculation" makes it sound simple and source-based, but it isn't. Let's please avoid claims on this page that aren't source-based. It needs to be made clear that the source in the article neither defines a typical dose nor a typical indication, both of which are necessary to state a monthly cost of treatment if all one has is the average price of a 25mg pill. The Drugs.com link (which isn't used as a source for that price claim) also does not claim a "commonly used dose". Indeed as both Drugs.com and our article state, there are two uses of this drug: one for high blood pressure and one for what our lead calls "swelling" but is actually fluid retention (oedema). The first may be treated initially with a 25mg dose, which may be then risen to 50mg if insufficient response and up to a maximum of 100mg. The maintenance dose is 25 to 100 mg, with the link not providing any guidance as to where in that range is common. The edema indication is treated with 50 to 100 mg orally once a day, or 100 mg orally every other day; some patients may require 150 to 200 mg orally at these intervals. The BNF gives somewhat different guidance, perhaps reflecting differences in prescribing practice in the two healthcare systems. The article gives a monthly cost to treat, but neglects to mention the indication being treated. Wrt DDD, a measure designed solely for drug utilisation studies, it is most interesting that the paper linked above notes we used the World Health Organization DDD classification to analyse dose–response relationships. This classification implies that a dose of 25 mg of chlorthalidone is equivalent to a dose of 25 mg hydrochlorothiazide. This assumption is probably not valid. A recent cross-over trial suggest that chlortalidone is about 1.5–2 times as potent as hydrochlorothiazide with regard to antihypertensive efficacy. -- they admit using the DDD for a purpose that WHO explicitly warn against: "DDDs do not necessarily reflect therapeutically equivalent doses of different drugs and therefore cannot be assumed to represent daily doses that produce similar treatment outcomes for all products within an ATC category". We are similarly abusing DDD for our cost-to-treat claims, made all the worse since there is no published information about what exact indication a DDD was calculated for.

The question asked was "How did you derive the specific information entered for chlorthalidone"': this information was not derived from the source used, and indeed was not derived from any source offered. -- Colin°^Talk 19:05, 5 January 2020 (UTC)

James. OK, so far we are to gather that you chose to present an interpretation of a 25 mg dose for diabetes (apparently based on information from other sources). Why did you choose that dose and that condition, as opposed to another condition which may have a different dosage? And how is this not SYNTH? Why did you not choose the very common 12.5 mg dosage (even correctly cited at chlorthalidone on Wikipedia to a Cochrane review) for high blood pressure-- which requires the patient to split the tablet, since there is no such thing as a chlorthalidone tablet made in the recommended dosage for high blood pressure? SandyGeorgia (Talk) 19:37, 5 January 2020 (UTC)

Drugs.com says for hypertension "Manufacturer recommends initial dosage of 25 mg once daily; may be increased to 50 mg once daily. Some experts recommend dosage of 12.5–25 mg once daily based on efficacy and tolerance demonstrated in clinical studies."

Sure I picked a middle value. User talk:SandyGeorgia how do you propose summarizing? Doc James (talk · contribs · email) 02:40, 6 January 2020 (UTC)

James the original text you said was for diabetes, but now we are talking about high blood pressure, right? Does the source indicate which is the most common use? Also, you are citing drugs.com for recommended dosage for high blood pressure, but even our article better cites a higher quality Cochrane review in support of 12.5 mg (common dosage in my experience as well).

A meta-analysis of trials of chlortalidone for high blood pressure found that lower doses of chlortalidone (e.g., 12.5 mg daily in ALLHAT study) had maximal blood pressure lowering effect and that higher doses did not lower it more. (Musini VM, Nazer M, Bassett K, Wright JM (May 2014). "Blood pressure-lowering efficacy of monotherapy with thiazide diuretics for primary hypertension". Cochrane Database Syst Rev (5): CD003824. doi:10.1002/14651858.CD003824.pub2. PMID 24869750.)

I don't believe 25 mg is a "middle value" for any condition for which the drug is used; is it? The drug is used for multiple conditions, and at multiple dosages. Does the source used give you the information used to decide which dosage to report for which indication?

I cannot recommend a way to summarize the information from the database because doing so appears to involve original research ala synthesizing of information from multiple sources, all of which brings us to the various problems mentioned throughout this page.

How do you propose summarizing, considering these problems? In order to move forward with the RFC, we have been searching high and low for an example of policy-compliant use of the source (NOR, V, SYNTH). This is separate from the whole question of whether we even should be putting database prices in the leads of articles, so we are hoping you will provide us an example, so that the RFC can present the best possible information for consideration of the broader problem. SandyGeorgia (Talk) 03:22, 6 January 2020 (UTC)

We can say "In the United States the wholesale cost is about US$13.50 a month as of 2019 for a dose of 25 mg per day." Which requires no synthesis. I do not see an issue with using the DDD for an approximate.

If you look at page 19 of the Cochrane review the 25 mg dose lower BP by 13.6 well the 12.5 to 15 mg dose lowered BP by 10.1. So not exactly equivalent. We see a trend to greater benefit with 25mg. Doc James (talk · contribs · email) 07:33, 6 January 2020 (UTC)

James, wrt "which requires no synthesis", are you accepting that all the price claims where articles state a cost per unspecified dose, or per day or month for some unspecified indication, and the source lacks a single and explicitly therapeutic dose value, fail WP:SYNTH? As for DDD, WHO reject this usage, saying it is a misuse of a technical metric designed solely for medicine population utilisation research. If it was a valid use, the MSH Price Guide would have a column for "Cost per day". They don't and so we can't claim that either: that would be WP:OR. As for 12.5 vs 25mg, the fact that sources disagree suggest there is No One Dose. -- Colin°^Talk 10:57, 6 January 2020 (UTC)

There are common doses. And there are common dosage ranges. Yes we all agree there is no one dose. WHO does not reject cost per DDD. They mention it as one possible price metric. What do you think of "In the United States the wholesale cost is about US$13.50 a month as of 2019 for a dose of 25 mg per day."

So if the MSH Price Guide had a column for "Cost per day" you would be satisfied? What about cost per DDD? Doc James (talk · contribs · email) 11:03, 6 January 2020 (UTC)

James, if the sources said what we claim in the article text, we'd all be much happier, though there would still be questions of WP:DUE, etc. None of our sources give "Cost per DDD". You are performing WP:OR. -- Colin°^Talk 11:36, 6 January 2020 (UTC)

We have sources that give "cost per DDD"[99] Doc James (talk · contribs · email) 11:49, 6 January 2020 (UTC)

James, when you say We see a trend to greater benefit with 25mg, who is "we" and what do you mean? SandyGeorgia (Talk) 11:59, 6 January 2020 (UTC)

Sandy in the table of the Cochrane review. Doc James (talk · contribs · email) 12:05, 6 January 2020 (UTC)

James The Cochrane review is a meta-analysis looking at many trials. Do you disagree with the summary of the review that is at chlorthalidone now? A meta-analysis of trials of chlortalidone for high blood pressure found that lower doses of chlortalidone (e.g., 12.5 mg daily in ALLHAT study) had maximal blood pressure lowering effect and that higher doses did not lower it more. SandyGeorgia (Talk) 12:16, 6 January 2020 (UTC)

Actually Sandy I'm beginning to regret commenting on this. As an aside I found this and this saying "Mentioned in the NICE hypertension guidelines (CG127) at a dose of 12.5mg-25mg, but only available as a 50mg tablet.... Hygroton® tablets discontinued by manufacturer October 2013. New generic product available as 50mg tabx30=£88.04." and "chlortalidone has become very expensive (£88 per month) and therefore rarely used in the UK". The BNF confirms only 50mg tablet available in UK at £88 for 30. The price of "£88 per month" either assumes a 50mg dose or assumes that the unused portion of each tablet is discarded each day. I found a source suggesting the tablet is dispersible in water, which would permit a crude method of achieving a quarter-dose by taking only a quarter of the dispersed liquid, and discarding the remainder. That isn't optimal or economic. Anyway, it does indicate there is actually a notable cost factor influencing UK prescribing [which doesn't necessarily mean we have to quote a cost in GBP, merely note it is expensive in the UK so not commonly used], and that there may be reliable secondary sources discussing the cost issue in the UK, but also that the dose-cost calculation in the UK may not be straightforward due to only the 50mg tablet being available. It doesn't excuse doing original research or synthesis on the topic. -- Colin°^Talk 21:02, 5 January 2020 (UTC)

Yes that would be interesting to include.Doc James (talk · contribs · email) 02:40, 6 January 2020 (UTC)

Barkeep49, could you please move this discussion out of a section that we all agreed should not be here, to its own section where we can further discuss the example? SandyGeorgia (Talk) 14:38, 5 January 2020 (UTC)

Back to responses

I don't think that would be helpful. Barkeep49 (talk) 21:46, 4 January 2020 (UTC)

With that noted, and I do realize that there has already been discussion about whether to address these things here or in user space, I feel the need to say that I'm uncomfortable with having this kind of discussion here. To lay out a case against a particular editor, well, I'm just not comfortable about this. --Tryptofish (talk) 21:58, 4 January 2020 (UTC)

I also admit to some discomfort - hence my reply to Sandy above - but am hoping that parts 4 & 5 are forthcoming soon so that it can be judged in total. Barkeep49 (talk) 22:21, 4 January 2020 (UTC)

Wellllll ... I am over here cursing at now having totally lost Parts 4 and 5 to multiple edit conflicts :( :( I will start over, but yes, this is awkward. Do you still want it here? SandyGeorgia (Talk) 22:28, 4 January 2020 (UTC)

SandyGeorgia, oh no! It doesn't recover if you go back? Please go ahead and finish here at this point. If I need to walk back my decision to move it here I will. Barkeep49 (talk) 22:38, 4 January 2020 (UTC)

Perhaps you've noticed it takes me quite something these days to make even a simple post :) Essential tremor, eyesight, and no, I have never figured out why I so often lose the whole thing in edit conflict. Almost done re-composing in sandbox. Might not be as carefully worded. SandyGeorgia (Talk) 22:43, 4 January 2020 (UTC)

Part 4: Barkeep49 said:

In the time that I've been working on this conflict I've worked hard to treat all editors with respect. ... But I also remember acutely what it's like when you don't have the sysop flag. I wrote, in a line I had to remove from my ACE statement due to space constraints, "I remember what it’s like to feel put down not or otherwise dismissed because" I wasn't a sysop. We lose so many good editors for so many reasons and I'm sorry that the conditions here are such that we're going to (potentially) lose you. The places you've chosen to contribute in this dispute have been made better because of your contributions. I can only hope you decide Wikipedia remains worth it.

First, thank you Barkeep for the kindness in your response. As to whether Wikipedia remains worth it, I am first and foremost a medical editor, and next, was highly involved in the featured article process. Obviously, I would prefer to contribute medical FAs to Wikipedia, and I can better spend my time IRL if I can't do what I do best here. That's why I'm here, trying to resolve this conflict.

I do believe you have treated everyone equally, and with respect. I'm glad you raised the memory of what it feels like to be treated lesser when you don't have a sysop flag. In my case, I never wanted it, not only because it would be a distraction from contributing content, but because of one of my earliest experiences on Wikipedia. I was attacked by someone claiming to represent "we admins";[100] the position expressed there was completely fictitious, against every behavioral policy, and I had no recourse but to sit on it. A few months later, Raul appointed me FAC delegate, and a year later, that admin was desysopped. I understood then that a fact of Wikipedia's dispute resolution processes was that it takes a long arbcase and a lot of different instances of misuse of the tools to deal with admin abuse, so we regular editors had best simply accept and live with that reality.

In this instance, although you have treated everyone equally, and "adminning" the MEDLEAD RFC was not your remit (and no other admin came to your assistance), let's look at how non-sysops vs sysops have fared.

1. On this page, notice that the "remit to a single venue" is casually overlooked until I pointed it out. I am not sure a regular editor would assume so casually that they can break the ANI condition.
2. Look at the statements aimed at WAID and myself here; no response from anyone. In contrast, at the end of this section, notice how you are respected; the same poster is asked to dial back.
3. In this section, even while we are under sanctions, look at the alphabet soup thrown at me, after I tried to resolve an (admittedly) malformed RFC. AGF?

Yes, you have treated everyone equally and respectfully. But nonetheless, not everyone ends up being treated equally. I have a tough skin because of years of corralling cats at FAC, and having to stay above it and stay neutral in disputes no matter what was thrown at me. But the effect piles up, and then it is best to take a break for tea, and I'm sorry that occurred on your watch, which has been fair.

The reason it occurred when it did is because I so abhor the effects of backchanneling (something that was well understood during my tenure at FAC, and everyone knew if they cooked up support off-Wiki, I'd shut down their FAC ... and don't even think of emailing me about your FAC unless it is something like a serious COI, which did happen once, and I wrote the arbs). My issue is that if you had had the same conversation with those admins in public view, others would have had the opportunity to point out the items in Point 3, without us coming to this level of awkwardness after the fact. All of this thread could have been avoided; transparency works. I still feel awkward about putting this on this page, but understand why you wanted it here.

I believe Part 5 is now completely summarized back on your talk, but having lost a bunch of this post, I may need to go back and re-read everything to see what I have not addressed.

I think we have a way forward, in addition to the RFC, that doesn't require arb intervention at this point. I believe it has now been revealed that the premise that the ANI close was founded upon was faulty, and we can revisit. That does not mean we should not launch the RFC that we have worked so hard on. Best regards, SandyGeorgia (Talk) 22:57, 4 January 2020 (UTC)

Thanks for your thoughts and concerns Sandy. I'm going to respond to your statements in two pieces. The first will be at my user page and focus on my conduct and that of others. The second part is here and will focus on the core content issue (should updating pricing information be covered by the ANI embargo). I normally wouldn't split discussion like this, but in this case there are related conduct issues (relating to my conduct and the conduct of others) and content issues (related to an interpretation I made of an a community decision. As one part of that close makes clear that the question of pricing is related to a single forum, it's important that the content based discussion be held here, while this talk page remains an inappropriate venue for conduct decisions. This is an imperfect solution but all available solutions are imperfect and so we're stuck with trying to make the best of a difficult situation. Now onto my response.

You write "We have reams of evidence above that all of the drug price information we have presented to our readers in these 500+ examples is bogus...[removal of conduct related discussion addressed on my talk page]... I have been asking for a month, and we have not yet seen one single example of these database sources being used correctly." This is where I think we've been ill served that the people who feel that pricing information has been used correctly are not participating as actively as those who oppose it. Though I will note that at times both WAID and Trypto have presented modified versions that they feel is policy compliant. Those examples have not been satisfactory to you which is fine. Different Wikipedians can have different interpretations of our policies and guidelines. So I would suggest in fairness that your second sentence is really "we have not yet seen one single example of these database sources being used correctly in my expert opinion" (with the words in bold being my addition). This kind of situation is what the essay WP:POLSILENCE is talking about because everyone reading this should know that the silence here doesn't mean that everyone agrees. And at the end of all of this, that's why we're having an RfC to find out if people agree with you (and the interpretation of policy and guidelines that you and others who share your beliefs cite).

That RfC, hopefully, gets us a consensus that lets us start to adjust articles around drug pricing again. In the interim when asking to interpret the embargo I have been attempting to balance "What will keep this conflict from escalating?" with "What will serve our readers?". So adding disputed tags serves our readers but did not, in my evaluation, do enough to serve our readers to outweigh the potential that had to escalate the conflict. Similarly adding to a draft (which obviously won't be seen by most readers) does not offer enough benefit to our readers to outweigh the potential to escalate the conflict. In this case I made the judgement that it could potentially help the conflict (by possibly showing a formulation that some editors would find appealing) while also helping our readers. Of course it could also exacerbate the conflict. That's why I decided this was the time I needed others thinking. Through a mistake filled route we got here (which I've previously addressed and accepted responsibility for) we do have a second uninvolved sysop (L235) offering their related reasoning that reached the same conclusion. I have, for reasons I explained in more depth in the conduct response on my user talk page, not felt that any history was appropriate to consider in light of the ANI's consensus that there would be "no rehashing of grievances.". What I read you writing here is that you weigh the competing priorities of service to readers and conflict escalation differently. Which I understand and have and will continue to consider but also does not at this point change my interpretation of that balance. Best, Barkeep49 (talk) 04:07, 5 January 2020 (UTC)

I believe WAID's example were of a hypothetical "if-then" format; perhaps she will correct me. If Trypto has an example, then it would be most helpful if he would fill it in at User:Tryptofish/Drug prices RfC draft 2 so we can all understand it. And perhaps Doc James will answer the questions about how the data in at least one article was derived. SandyGeorgia (Talk) 06:31, 5 January 2020 (UTC)

Barkeep49, I may be wrong but I think WhatamIdoing offered an example of revising the text to eliminate the WP:NOR problem but accepted that then it introduced a WP:DUE problem. Indeed, I think their point at the time was that DUE was the underlying problem. I have not actually seen anyone (other than James) offer article text + source that they themselves believe is fully policy compliant for the purpose of a lead in a drug article, but if I'm wrong I would like to see it. On another note, I feel all this analysis over the ANI restrictions and whether they were broken and fairness of treatment, is really getting us into off-topic timesink territory. Further, when we launch an RFC, I really think WT:MEDMOS could do with all this ANI section being archived away. None of it is relevant to what our article text should say and what our guidelines should say. -- Colin°^Talk 10:18, 5 January 2020 (UTC)

Following up, I am fairly certain that the words in my expert opinion have never come from my mouth, and I really hope never been typed by my fingers. In real life, I would not be likely to promote myself as an expert even in areas where I am. On Wikipedia, we should deal in facts, not opinions, and I was explicitly addressing the facts that have surfaced after a month of reviewing these articles looking for one good sample. I believe my words covered the facts to date: I have been asking for a month, and we have not yet seen one single example of these database sources being used correctly. That doesn't mean there may not be one; yes, we are hamstrung because of silence of those supporting this data source. Re WP:POLSILENCE, I am not convinced that essay applies here, and I wouldn't even use it at the MEDLEAD RFC, where it might apply-- it's an essay. More significantly and to matters at hand, if we can get the chlorthalidone example moved to its own section, we can continue to explore now one example. SandyGeorgia (Talk) 17:55, 5 January 2020 (UTC)

Barkeep49 Trypto has clarified above, and his understanding is the same as mine. We do not have a policy-compliant example of these database sources being used (yet, we are working to get James to explain his derivation), and Trypto does not have one either, and there is not one on this page. We need to get older portions of this page archived (perhaps to a separately numbered archive of this page as I mentioned before); no mere mortal can work this way. We are not tracking in here what people have actually said, and what I said above about having not a single sample yet appears to be correct. SandyGeorgia (Talk) 20:34, 5 January 2020 (UTC)

I strongly suggest we archive the older sections at the top of this talk page to Wikipedia talk:Manual of Style/Medicine-related articles/Archive 10RFC, link that archive to the top of the remaining sections, separate out the chlorthalidone discussion with James above from the discussion that does not belong on this page, archive the rest of that discussion, and create a clearer work space here. SandyGeorgia (Talk) 20:38, 5 January 2020 (UTC)

We definitely do need to clear-up our workspace (I am having trouble following it and I can't imagine trying to jump into this fresh). I might not get around to do this for a day or so though. Barkeep49 (talk) 00:33, 6 January 2020 (UTC)

I think that Wikipedia talk:Manual of Style/Medicine-related articles/Archive Price RFC might be a more typical name. However, I don't care what it is, so long as some stuff gets moved into it soon! WhatamIdoing (talk) 03:54, 6 January 2020 (UTC)

I have always assumed that when SandyGeorgia talked about not having an "example of these database sources being used", that "these database sources" was a key qualifier. Database entries can't justify the importance of their own contents to the world, so the importance of including that data point needs to come from a different source. WhatamIdoing (talk) 03:39, 6 January 2020 (UTC)

WhatamIdoing Seeing the clarity in your final draft RFC has me hanging my head as I am now realizing how much we have not been always understanding each other on this page; no wonder its 800KB! I was stuck on-- if we can do it for Epipen, and we know we can, we need James to show us a policy-compliant example of what he is doing. I was concerned with how to deal with SYNTH, NOR etc, while you stayed focused on, how can we generate MEDMOS text. I suppose my repeated requests for a sample made no sense at all :) Very well done. (And I saw a post somewhere up there about everyone knowing your views on recent (five-year) sources ... I missed that ... could you fill me in on my talk? SandyGeorgia (Talk) 03:49, 6 January 2020 (UTC)

No need to take it elsewhere: Read (or remember any version of) WP:MEDDATE, and then just imagine what anyone involved in writing MEDRS would think about proposals to remove good content about a rare disease solely because the cited source was six years old. Wikipedia:Nobody reads the directions, but we love to guess their contents based on the shortcut and what someone told me several years ago about a completely different situation. WhatamIdoing (talk) 03:54, 6 January 2020 (UTC)

Ah, ha ... like me always having to update my sources at TS, when I know the older papers-- written in the heyday of TS research, when there was money it it and top minds wanted to be in the field-- are so much better quality than many of the newer ones. SandyGeorgia (Talk) 04:18, 6 January 2020 (UTC)

Collapsing not that important comment as suggested by Barkeep49. Nil Einne (talk) 14:17, 6 January 2020 (UTC)

Comment I don't really know where to leave this since it mostly concerns the whole page including the latest thread (at this time) but I don't want to make a new section so I'll just leave it here. I've been spending way too much time on Wikipedia so don't want to get drawn into this page although I've already read too much than I intended to but I wanted to make a quick comment as an outside non admin observer from what I read. While this is the talk page of an MOS page and so I understand unrelated stuff may come up, may I suggest given how things are going at the moment, that editor's try and keep the focus on how to improve the MOS page. This will include discussion on the RfC but such discussion should focus on the RfC wording. By this stage, I'm not sure it's particularly useful for editors to try to explain their PoV to each other too much since IMO it's clear it isn't helping much. In other words, comments should be something like "from my PoV, I think we need to add X to the RfC" or "from my PoV, the this part of the RfC should be reworded to". I appreciate that after you've done that, you may need to offer some explanation of why you feel so, and this may involve discussing your PoV, but I do feel there's perhaps a bit too much discussion which doesn't seem to relate to the RfC wording or other proposes changes to the MOS page. I appreciate the ANI directed all pricing related discussion here, but I don't particularly see much point for most discussion like the merits of DDD until the underlying issues are (hopefully) resolved via the RfC and I strongly suspect that was the intention of the ANI participants. Yes I appreciate there may be some irony here given my recent posts, and I'm personally often involved in diversions of a thread or discussion, but it does seem to me that in this particular case, there is need for focus, especially by those who care a great deal about the outcome. At the very least, it may be worth voluntarily collapsing such discussions. Also, I would urge anyone interested, to participate in forming the RfC. I appreciate this is a long and confusing process, but I don't see any attempt to exclude or ignore participants. So I don't think the community will take too kindly to anyone who clearly had the opportunity to help draft the RfC but either didn't take part or didn't participate enough to make a meaningful change, who later complains about the RfC wording. I'm going to AGF that no one will be foolish enough to think they can get away with saying after the RfC, okay great, but I can do X because the RfC never actually asked about it, so there was no consensus. As a final comment although my view counts for little, IMO the embargo should apply to everyone equally. This means Doc James can update or "fix" information, but so can everyone else. If an editor makes a change and others object because they feel it is making the article worse, then it should voluntarily be reverted. If editors just feel it isn't improving anything, then it can stay since whoever made the change clearly feel it's improving things. I appreciate this creates an unfair situation since those who feel the info needs to be removed because it's inaccurate etc, can't do anything. But those who feel it is useful, like Doc James probably often can. And I understand how frustrating it must be for those who feel the info we have is damaging, But it seems the best temporary solution. By letting Doc James etc fix identified problems, we lessen the risk editor's may feel 'well clearly there are problems, but I maybe they can be resolvable". Doc James etc have had some opportunity to demonstrate that it is achievable and how, and the community can evaluate the result. This gets back to what I said earlier namely there should be a focus on the RfC. While I understand the view that another day is too long, it has already been very long that another 2 months (I'm assuming we'll have an RfC result by then unless there's the dreaded 'no consensus'), is IMO not going to greatly change any damage caused by having the information. Nil Einne (talk) 14:17, 6 January 2020 (UTC) Nil Einne, regarding the focus of this page, we all agree, and have taken your comments on board several days ago. The section you are posting in now was (most unfortunately) moved here from a talk page, after I needed calming tea from one incident, and I think we all wish it could have been moved back, but we are where we are, and moving forward. Conduct issues are separate from RFC issues. I'm sorry you had to read through all of that, which is wholly off-topic on this page, but it ended up here as a result of good faith on the part of the admin who put it here. SandyGeorgia (Talk) 14:46, 6 January 2020 (UTC)

DS

I have a procedural/administrative question. I see that there is a DS notice at the RfC draft page, based on this being related to WP:MOS. It had never occurred to me that we are working under a DS situation here. Are we? --Tryptofish (talk) 20:20, 4 January 2020 (UTC)

I put it there so respondents wouldn't fall afoul-- yes, we are. The same DS notice was added at the talk page of the MEDLEAD RFC by Barkeep (didn't stop people from misbehaving, though :) SandyGeorgia (Talk) 20:48, 4 January 2020 (UTC)

@Barkeep49:, sorry to saddle you with anything more, but you or another admin need to do the Wikipedia:Arbitration Committee/Discretionary sanctions#Awareness for this talk page. I think that includes an edit notice here. --Tryptofish (talk) 20:59, 4 January 2020 (UTC)

Was I not supposed to place that template at the new RFC? I am not aware of the intricacies; I placed it to follow suit on what was done at the other RFC, so we wouldn't forget to let people know when they come to the RFC ... SandyGeorgia (Talk) 21:08, 4 January 2020 (UTC)

No problem there. Anyone may place that template (although I think it goes at the top, not the bottom). --Tryptofish (talk) 21:10, 4 January 2020 (UTC)

Ah ... I first put it at the top, but then thought we would scare people off, and they wouldn't participate ... so I removed it to where discussion would begin. Well, not to worry, I'm sure Barkeep will fix it before launch at any rate. I was just afraid we'd forget. SandyGeorgia (Talk) 21:15, 4 January 2020 (UTC)

(edit conflict)I'm happy to fill out DS related paperwork but when I went to put the edit notice here it was already present. I've been handing out individual DS alerts as people join in the conversation here (I some how had missed Trypto until just now and have a couple others who've joined in the last week who I don't think have notices). As of now there are no page restrictions for this page (e.g. 1RR) so I didn't think an edit notice was necessary. Best, Barkeep49 (talk) 21:21, 4 January 2020 (UTC)

Yeah, I kind of got a laugh out of getting the alert to me right after I posted here. And I just now saw the notice at the top of this page, so I'm dope-slapping myself for not noticing it earlier. However, I know that there is a recent ArbCom rule that an edit notice, which is something that appears in the edit window whenever one makes an edit, and not simply a top-of-page template, which is what we have here, is required. Only template editors and admins can do that. And DS do apply to things like personal attacks, regardless of individual page restrictions. --Tryptofish (talk) 21:26, 4 January 2020 (UTC)

Tryptofish, yeah if I, or another uninvolved sysop, enabled some sort of page restrictions like 1RR an edit notice is required. However there are no such restrictions yet levied on this page. So there is a general notice that it falls in the scope of DS and there is specific editor awareness through DS alert but no edit notice is needed. Best, Barkeep49 (talk) 21:44, 4 January 2020 (UTC)

OK. That's why they pay you the big bucks, and not me! (Not to mention that mop!) --Tryptofish (talk) 21:46, 4 January 2020 (UTC)

IMO it feels superfluous at best, and I'm concerned about it discouraging participation. But I wouldn't remove it without asking at one of the ArbCom noticeboards. Although the page is in the MOS "namespace", the contents have very little to do with the MOS. WhatamIdoing (talk) 03:48, 6 January 2020 (UTC)

@WhatamIdoing:, I'm pretty sure you can move it to the talk page, as was done at MEDLEAD RFC. SandyGeorgia (Talk) 12:38, 6 January 2020 (UTC)

Defined daily dose (DDD)

There has been a fair bit of discussion on whether or not DDD is appropriate for rough estimates of medication prices. We have a number of sources which support this use. Specifically the government of Canada says the DDD can be used to provide "a rough idea of the daily cost of utilizing a drug in a specific formulation" and "provides a rough idea of the cost differential between the two formulations of the same drug". We are using it for the first purpose. Doc James (talk · contribs · email) 11:19, 6 January 2020 (UTC)

James, I think our priority right now should be to launch an RFC. There's a whole section at WHO on DDD and specifically on its use and abuse. It very much argues against the use we are doing, and why their sole focus on DDD is for utilisation. And yes at a crude population level, it could give a rough idea of the cost of utilising that drug in a specific formulation. That's up to researchers to judge if it meets their needs, not Wikipedians. What it doesn't represent, is the dose that an individual patient might take and thus cost them per day or month. Repeatedly MSH and WHO have to remind us it is not a therapeutic dose nor does it represent average prescribed doses (which themselves vary from country to country and year to year) . Wrt your second example, yes WHO give that as an example of how it could be used to compare the cost of e.g. 5mg tablet vs 10mg/5ml syrup. But, if you think about it for a moment, all you are doing with that is agreeing on an arbitrary dose to compare two formulations: the actual dose need not be representative of any therapeutic value. It's just a number.

Furthermore, there are no sources saying what indication the DDD was calculated for or what choice they made if their own source lacked a clear "maintenance dose" value (they sometimes pick initial and sometimes max). When we have a drug for multiple conditions (epilepsy, neuropathic pain, mental health disorders), the DDD is totally useless because the cost to treat really depends on what indication you are treating, and you don't know that. DDD is a red herring. Our only source that offers it (MSH) warns about its careful use and does not themselves use it to give a cost-per-day, which should be a clear warning sign that this is inappropriate original research. The other sources (Drugs.com, BNF, NADAC) do not give a DDD or indeed any one dose, so any attempt to use them to give a cost per day/month/treatment is both original research and synthesis of sources. -- Colin°^Talk 11:50, 6 January 2020 (UTC)

Yes the DDD only provides a "rough idea"[101] of the daily cost. Yes it does not provide "detailed" description of the cost. That is why we use the term "about". Doc James (talk · contribs · email) 12:14, 6 January 2020 (UTC)

This source converts the MSH data into price per year with "Carvedilol is listed on the Management Sciences for Health International Drug Price Indicator Guide with average price per tablet of $0.20 or $144 per year for twice daily treatment"[102] Doc James (talk · contribs · email) 12:21, 6 January 2020 (UTC)

James, it is so "rough" because it is for population studies. The kind where someone says we might spend £1.5 million on a drug, but it probably doesn't change the point if the actual cost is £1 million or £2 million. Wikipedia is giving a price in dollars and cents for one patient, and then waving the word "approximately" about as if that absolves all sins. As for the paper from 2011, it is wonderful what Google can turn up. I tried to find the record they quote. Carvedilol in 2011 gives three doses. The 6.25mg tablet has a median (only) supplier price of 0.0414. At the DDD of 37.5, is six a day x 365 = $90. The 12.5mg tablet has no suppliers and a median buyer price of 0.1669. That's three a day x 365 = $182. The 25mg tablet has no suppliers and a median buyer price of 0.2041 which is pretty close to the "average price per tablet of $0.20" they mention. But wait, 25mg doesn't go into 37.5mg, and they mention "twice daily treatment" so I guess they mean 50mg per day. That price is twice a day x 365 which is $149. Not quite the $144 they give but close.

What does this tell us? The paper mentions "per tablet... twice daily" and yet there is no tablet dose that is 18.75mg. This is because DDD is not actually a therapeutic dose that any patient might take. Their maths mostly work at 50mg. Looking at Drugs.com we see the 6.25, 12.5 and 25mg tablets "twice a day" being the initial and then tritrated "if tolerated" up to the maximum, but lots of other dose options too. That really isn't the paper to convince anyone that DDD is a useful measure, when they don't use the DDD of 37.5mg but instead 50mg. Further, we see that if we did try to use the DDD, we get yearly prices of $90, $149 and $182 depending on which tablet size we chose. All three tablet sizes make sense for individual patients and indications and stages of treatment. This is why we don't allow original research: the numbers are effectively random. -- Colin°^Talk 13:25, 6 January 2020 (UTC)

The argument here is similar to saying, "doses of medications are random". Please note they are not. There are well accepted dosage ranges. Do we need a RfC to ask "are dosages of medications random" as that is the argument you are making? Doc James (talk · contribs · email) 13:32, 6 January 2020 (UTC)

No, I'm saying that original research produces random numbers. The source gives three pill sizes and none of them are 37.5mg or 18.75mg. You say we should use DDD and then you cite a paper that looked at the DDD of 37.5mg and went ??? that's not gonna work and picked 50mg instead. Decisions decisions decisions and each time a different result. If the researchers had searched in 2010 rather than 2011, they'd only have the 6.25mg tablet, and nobody is taking six a day unless they have no other choice. -- Colin°^Talk 13:59, 6 January 2020 (UTC)

On the big picture, we have (mis)spent a lot of bandwidth on this talk page trying to sort out the various problems (SYNTH, NOR, WEIGHT) in price text in our drug articles. We went down that path when we (I?) were (was?) seeking one good example of drug pricing information from these databases that did not have SYNTH problems, and did not find one. We can save a good deal of time by having you actively engaged in those discussion, James, and I'm not sure forbidding pings will help us move forward. Do you read all posts here, or do we have to ping you to each post? I ask because I really hate this pingie-thingie myself; when I come to a page I participate in and follow, I pull up a diff of everything since I last read, and the extra pings are just an irritation. Do we need to ping you, and if we are trying to sort out the price data in our drug articles, how can we assure you see the discussions if Colin is forbidden from pinging you? SandyGeorgia (Talk) 12:25, 6 January 2020 (UTC)

I guess the question is do we have a fundamental disagreement? Do you believe it is possible to provide a rough estimate of the cost of a medication in LMIC or other region of the world? This can be broken down into two parts:

1) Do you believe it is possible to provide a typical dosage range for a medication in adults used for a specific purpose?

2) Do you believe it is possible to provide a rough price for an amount of medication in LMIC?

The popular press manages to determine the cost for a course of treatment "Sovaldi treatment cost $1,000 a pill, or $84,000 over 12 weeks." The Guardian managed to determine the typical dose, the typical duration of treatment, and the rough cost per dose than do the math. [103]. What you call SYNTH and NOR is simple WP:CALC. WEIGHT is clear per sources such as Doctors Without Borders.[104]

The next question than becomes one of knowledge parity. The popular press generally just writes for wealthy people in the developed world. Do those in LMIC deserve to have pricing information for medications they may care about? My position is yes, and we have excellent sources such as MSH that provides these details. Doc James (talk · contribs · email) 12:55, 6 January 2020 (UTC)

I am thinking these could be two useful questions for the RfC. User:WhatamIdoing User:Tryptofish wondering your thoughts? If either of these are false than we would all agree that it is impossible to list a price for a medication for an area.

Other questions could be is multiplying a dosage range for a purpose by the rough cost estimate by a time period WP:CALC or WP:SYNTH and are prices of medications WP:DUE.Doc James (talk · contribs · email) 13:17, 6 January 2020 (UTC)

James, these aren't yes/no questions that apply in all cases. We can find drugs with one standard dose and we can find drugs with many suppliers in the MSH. But in the general case, no you can't. There are multiple indications, multiple dose ranges, multiple patient ages, weights and liver function and existing medications.

What matters isn't so much right now these abstract questions, which are complicated to answer, but actual sources and actual texts in actual articles, which despite months of discussion have not changed. If you believe these things are possible, and possible generally for most drugs (over 500 have these prices), then argue the case when the RFC is posted. There is a reason why WHO/HAI have only 14 core global medicines and survey at most 50 (at specific strengths and formulations for specific indications and patient age and specific treatment duration or daily dose). The MSH database is way too sparse to be a reliable source for 500+ medicines. It is officially not a reliable source for international reference price if there are not many suppliers. That's WHO/HAI official policy. Which trumps any Wikipedian opinion. -- Colin°^Talk 13:34, 6 January 2020 (UTC)

James, it is helpful to have you fully engaged in understanding that these are complex questions even for those of us who understand the sources and understand SYNTH. The problems we have now in more than 500 articles go way beyond what an average reader, or even RFC respondent, can understand. Diverting energy now towards addressing those problems is a distraction from the RFC. But, we must continue to make sure you are fully engaged with us in sorting out these problems. We have this split now between the sample price discussion above of chlorthalidone, and this section; let's continue here. Do you now see that what we have presented to our readers for chlorthalidone, in the lead of an article, is neither useful nor accurate nor in accordance with NOR ? Sorting out that we have a big problem in 500+ articles is a very different matter than deciding via RFC whether drug prices from databases should even be in our articles at all. SandyGeorgia (Talk) 13:55, 6 January 2020 (UTC)

After our discussion what I see is that what we have in our article on chlorthalidone is a perfectly reasonable estimate of the price per month of the medication in question. Looking at the references yes 12.5 mg can be used, 25 mg can be used, and 50 mg can be used. Could the price range from 6.75 to 27 USD? Sure. I prescribe medications that range in price from pennies to 10,000s per dose (a million fold difference). A 2 or 4 fold difference fits well within the range of "about".

Additionally I believe our readers are smart enough to realize this. Doc James (talk · contribs · email) 14:14, 6 January 2020 (UTC)

I chose chlorthalidone as a starting example because it is the one of the simplest I have seen in these discussions; the text we have presented on other drugs has far bigger problems. It is your opinion that our readers can sort it out. It is my opinion (and others) that what we have here is a problem not only of LEAD, NOTPRICE and WEIGHT, but a problem of SYNTH. Do you see why we need a separate RFC on that matter before we move on to the wider RFC? How can we ask Wikipedia editors whether price information should be included, when we do not even have price information that we all agree conforms with other policy ? SandyGeorgia (Talk) 14:31, 6 January 2020 (UTC)

James, mathematically, you cannot say "the wholesale cost is about US$13.50 a month" if you feel the price is only accurate to within 2x or 4x approximation and you don't care if 12.5mg, 25mg or 50mg tablets are used. Our readers trust what we write and really no amount of "about" or "approximately" covers us if we give a price to four significant figures. It is one thing if our sources use that language, but it isn't a sticking plaster for dodgy maths. They may look at treatment X and treatment Y and conclude that treatment Y is 2x or 4x more expensive than X whereas in fact the difference in price is down to the random chance of original research. Btw, at Carvedilol we give a "wholesale cost per dose". What does "per dose" mean? The dictionary tells me it is how much you take at one moment of time. So a DDD of 50mg but taken twice a day would produce a 25mg dose, but in a once a day sustained-release tablet would produce a 50mg dose. I don't think there's any consensus that we should provide prices in dollars and pence and then excuse our original research random results by saying the prices are only meant to be accurate to two orders of magnitude and our readers are bright enough to know that. But you can try that claim at the RFC. -- Colin°^Talk 14:35, 6 January 2020 (UTC)

Generally there is one main indication and regardless typically most if not all indications generally use similar doses. Most medication have a fairly narrow dosage range. Some of the sellers offer the medication in more than LMIC 100 countries. Qualifiers applies to all evidence within medicine and most of the time we have no idea how the qualifies affect claims of benefit as they have just not been studied. But these still not prevent use from providing an overview. Yah sure MSH is not as good when there are fewer supplies. Same as a meta analysis is not as accurate when their are fewer RCTs. Doc James (talk · contribs · email) 13:51, 6 January 2020 (UTC)

When our math is doubly wrong (12.5 vs 25 mg for chlorthalidone), we cannot say we are within a "fairly narrow dose range", and that is only one very simple example. SandyGeorgia (Talk) 13:57, 6 January 2020 (UTC)

Are you saying 12.5 to 25 mg is a wide dosage range?

Now back to the question at hand. Do you believe that there is an accepted dose range for a specific indication in a typical adult? We can go with this if you do not want to use DDD. Doc James (talk · contribs · email) 14:09, 6 January 2020 (UTC)

I am saying we don't have sources to back up the information we are giving our readers without using SYNTH, and our personal opinions, what we have experienced or seen as physicians or people interpreting for physicians, should not be coming in to play at all. We do not have sources that back up the information we have presented unless we do synthesis. And we do not have sources that, according to DUE WEIGHT, tell us what to present in the example of chlorthalidone. The problems in other articles are worse. This is probably why we should be respecting WP:NOTPRICE and not presenting these prices at all based on database sources, but that is my opinion.

What would be helpful to see is whether you have any example, for any drug, that discusses drug cost in a way that does not breach, IMO, WP:SYNTH. Then we could more accurately discuss WP:WEIGHT and WP:LEAD issues. SandyGeorgia (Talk) 14:22, 6 January 2020 (UTC)

What I am saying is we do have sources that provided accepted ranges for medications plus we have sources for DDD. It is thus a simple WP:CALC to convert price per dose to price per day for a typical dose or dose range which is more useful than just the price per amount by itself. Doc James (talk · contribs · email) 14:27, 6 January 2020 (UTC)

So do you see that we have a policy disagreement as to whether this is simple math without synthesis that needs to be put forward in an RFC, because we got ZERO feedback on the matter when we posted a question to the NOR noticeboard? And that we need to sort that out before we can solve the bigger question? We have not, throughout these discussions, seen one straightforward example of drug price information from the sources used in over 500 articles that we can all agree does not involve SYNTH. SandyGeorgia (Talk) 14:35, 6 January 2020 (UTC)

We have "As of 2016 a 12-week course of treatment costs about US$84,000 in the United States, US$53,000 in the United Kingdom, US$45,000 in Canada, and about US$500 in India.Hill A, Simmons B, Gotham D, Fortunak J (January 2016). "Rapid reductions in prices for generic sofosbuvir and daclatasvir to treat hepatitis C". Journal of Virus Eradication. 2 (1): 28–31. PMC 4946692. PMID 27482432." Not sure if you consider this SYNTH aswell.

But yes lets put the SYNTH versus CALC discussion to a RfC. Doc James (talk · contribs · email) 14:53, 6 January 2020 (UTC)

it isn't a simple calculation to convert. Firstly the sources don't give a price per "dose" either. They give a price for one tablet size or another. Even if we thought DDD was acceptable (which WHO, who invented it, don't) it doesn't necessarily divide into specific tablet sizes (see above for 37.5 DDD) so no there is not a straightforward calculation. The drug may be once a day or twice a day or some other option, and this may depend on what kind of tablet you take. These are all things a doctor will decide at prescription time. The most you can do, James, is convert a price per day to a price per month, assuming it is obvious the indication requires long-term treatment, of course. -- Colin°^Talk 14:39, 6 January 2020 (UTC)

Yes we known that you think that WHO does not think that it is acceptable. But it does not appear that WHO thinks that it is unacceptable. Doc James (talk · contribs · email) 14:53, 6 January 2020 (UTC)

I think Nil Einne (above, collapsed) has a point. Most of this involves one party making a claim and then another party disagreeing. We can do that in the RFC. I think this should wait till the RFC, when other voices can chip in with support or rejection and hopefully a consensus form. There's zero evidence this discussion is heading towards any consensus, that's why we are having the RFC, to get other voices. -- Colin°^Talk 14:55, 6 January 2020 (UTC)

RfC are we allowed to multiple the "typical dosage range per day of a medication in adults as used for a specific purpose" by the "price per dose" to get the cost per day for a specific purpose? Doc James (talk · contribs · email) 14:58, 6 January 2020 (UTC)

Example article text + source please, which gives either of those things, singular. -- Colin°^Talk 15:15, 6 January 2020 (UTC)

Reboot

Allright, we are spinning our wheels. (Not complaining, since this is much better than the alternative, which was silence.) Let's start over.

James, on this page we have been discussing a SYNTH problem. You believe this is simple math, others do not. You believe using other sources supports the math you are using to add text cited only to a database. Let Colin pick a typical example, and you write text that incorporates all of those other sources you are using to support your math, and shows how you have used them. Then we can talk SYNTH vs. CALC. And from there may emerge an example to be used for an RFC. SandyGeorgia (Talk) 15:07, 6 January 2020 (UTC)

I dont' want to be accused of cherry picking. James cites Carvedilol above. Try that. -- Colin°^Talk 15:15, 6 January 2020 (UTC)