Wikipedia talk:WikiProject Medicine

Welcome to the WikiProject Medicine talk page. If you have comments or believe something can be improved, feel free to post. Also feel free to introduce yourself if you plan on becoming an active editor!

We do not provide medical advice; please see a health professional.

• Unsure about something? Make sure to look at our style and source guidelines.
• Please don't shout, remain civil, be respectful to all, and assume good faith.
• Put new text under old text. Click here to start a new topic.
• Please sign and date your posts by typing four tildes (~~~~).
• Threads older than 10 days are automatically archived.
• Please see Wikipedia:WikiProject_Medicine/Newsletter/Mailing_list

List of archives

Merged Archive WikiProject Preclinical medicine: 1 Merged Archive WikiProject Clinical medicine: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 WikiProject Medicine Archives: Index, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171

Introducing ScienceSource

The ScienceSource project, funded by a WMF grant to work on medical referencing, was started in June and is now launching its first participatory activity, its "focus list" of biomedical articles, hosted on Wikidata.

In the story so far, work has gone on here on Wikipedia:WikiProjectMedicine/Referencing case studies for MEDRS, and on Wikidata relating to Medical Subject Headings (diseases). There is a wikiversity:Wikiversity:ScienceSource mentoring page under development for those who want broader background. We are using lightweight videos for that.

The focus list, information at d:Wikidata:ScienceSource focus list which is WD:SSFL on Wikidata, is the first pass at which (open access, biomedical) articles ScienceSource should look to download, in future months. We want those with expertise to work with us on compiling an excellent medical bibliography. Once the list is set up, it will be possible to query and display it in multiple ways, so slice 'n' dice will be on the cards.

It is really not so troublesome to add a single Wikidata statement, once you have the relevant item about the article. And there is a tool to translate a DOI into the item number. Just about every relevant DOI used here in a reference is likely to be present on an item on Wikidata, since recent work has gone on in this area. (For lists, there is technical support available, so do ask.)

There has been discussion of topics around the project in recent issues of Facto Post (back numbers), which by convention is not delivered here. You can sign up for this mass message to be delivered on your User talk page. I'm always glad to hear from people on my own talk page. Charles Matthews (talk) 09:54, 10 July 2018 (UTC)

I followed several of those links and, sorry, I don't understand what this means or what you want people to do? Bondegezou (talk) 10:01, 10 July 2018 (UTC)

Would be useful to have all journal articles used within Wikipedia tagged by whether or not they are a "|type= review" versus not. This of course does not related to what is proposed above.

Would also be useful to have a bot run to add missing pmids / dois if only one is present.

This looks like a Wikidata project rather than a Wikipedia one. Doc James (talk · contribs · email) 10:13, 10 July 2018 (UTC)

@Doc James: Isn't the missing pmid/doi infill on WP still being contributions done by User:Citation bot? I long ago suggested that the PublicationType field on pubmed should be harvested and mapped into our citations' |type= parameter, using values such as "Systematic Review", "Meta-analysis", or just "Review". There was resistance to the idea, based on concerns over data quality at the time, IIRC. LeadSongDog come howl! 16:09, 10 July 2018 (UTC)

@LeadSongDog: Thanks, that's interesting history and I wasn't aware of it. The objection can be met by storing the PublicationType field on Wikidata, on the item for the article, and referencing it to PubMed (so implying "for what it's worth"). Charles Matthews (talk) 09:39, 11 July 2018 (UTC)

Thanks for the responses. I have just added the three introductory videos (60 seconds) to the Wikiversity page. These are designed to give people some overall idea of the project.

In terms participants in this WikiProject would understand, the big problem ScienceSource addresses is the one discussed a couple of years ago (see Wikipedia talk:WikiProject Medicine/Archive 85, #24 on the ToC). We are not working on that yet, though it would be good to discuss how we see a new approach. We shall be downloading articles to our own wiki, which is being configured as we speak. Those articles are going to be "good for referencing", on the face of it. The algorithm will check if those articles actually "pass MEDRS". Other software will help locate useful factual content in the articles. Putting it all together, reliably sourced medical facts will be obtained.

One month in, we are still "in the foothills". The focus list project is to get a preliminary bibliography together. That is what I'm proposing today, for participation.

Yes, we have to start with data. Whether something is a review or not of course is very much on topic for the project, and I was discussing it on Saturday. There is an issue of scale: it would theoretically possible to find out (for what it is worth) for each article now on Wikidata and from PubMed, whether PubMed reckons it is a "review article". Right now that would require searching 17.5 million articles. If we have a list of candidates of (say) 10K suggestions, we have a more manageable problem. We'd still need to understand better what "review" means. Charles Matthews (talk) 10:26, 10 July 2018 (UTC)

Work like doi:10.1186/2046-4053-4-5 and related seems highly relevant to the proposals. Automated selection of papers for systematic reviews is becoming commoner, and there is also increasing interest in automated data extraction from papers. That said, I am dubious about the challenges of obtaining "reliably sourced medical facts" at the end. Bondegezou (talk) 10:36, 10 July 2018 (UTC)

OK, the grant process took over three months to complete, through half-a-dozen stages, and much of what was said then can be read on m:Grants:Project/ScienceSource and its talk page. I don't want to rehash that here, for obvious reasons. We said that we'd be trying to engage the medical editors and others in what we are doing, and getting a good baseline for papers is where it starts. (By the way, the company has been involved in work around systematic reviews.)

The idea of getting people to add some small numbers of papers to a list is actually consciously modelled on m:The Wikipedia Library/1Lib1Ref, and we took advice from someone involved in that effort. Charles Matthews (talk) 12:40, 10 July 2018 (UTC)

I just read Wikipedia:WikiProjectMedicine/Referencing case studies for MEDRS and find all that fairly clueful. The hardest thing in all of this, in my view, is identifying "high quality sources". When I go through pubmed looking for reviews, I don't necessarily take the newest, non-predatory one. I also look for "signs" that the piece is really an effort to summarize accepted knowledge in the field vs someone making an argument that might be not really mainstream in the field. One sign of that, is to the extent to which the authors are citing their own papers or are sampling widely from the field. Another is pretty subjective, but seeing the extent to which people are buying woo (see the discussion about PMC 5438513 in this archived discussion here, from last summer; I guess you could look for woo-buying keywords like "guarantee" or the like to filter those out. I almost never cite Frontiers journals because many issues are guest edited, and while articles in those issues are interesting you can bet that people are making arguments rather than really giving the state of play. Unfortunately some Cochrane reviews are the product of advocacy; see for example this discussion here about the HepC review, here about Cochrane's alt med group; here on Psych drugs; see also Tom Jefferson (epidemiologist) for his views on what kind of evidence should "count" for vaccines, which is way, way out there.

I think the judgement that goes into selecting sources is as difficult and important, as what information is extracted.

Then there is the whole issue of giving WEIGHT to various findings. You can see the content at Oseltamivir#Medical_use where clinical guidelines and Cochrane reviews directly contradict each other, and you can see the negotiations on the talk page where we worked out how to handle that.

These are challenges; I am not sure how your project can address them... Jytdog (talk) 19:46, 10 July 2018 (UTC)

Thanks for the comments. This is a 12 month project, and we have eight or nine months now to tune what we do with MEDRS. You make some reasonable points about the actual practice. I expect an approach of salami-slicing the difficulties, rather than one big advance. The starting level is publication date/publisher/type of publication. We can gather good data on those.

After that it will get somewhat messier, from a data point of view. We can address "stale or not" considerations for past-best-before-date reviews, with a list of exemptions. We can also address the "pecking order" issue of some reviews, e.g. Cochrane, being considered ace of trumps in the particular area: Cochrane reviews are at least well documented here, and I have been doing some preliminary work about that.

Then, we shall set off writing an algorithm in SPARQL, which (in contrast I think to previous proposals) is quite well adapted to problems where the available data is disparate. A SPARQL query that looks at least quite like what MEDRS is trying to say is at least a formal representation of a purported solution. In other words, fixing it up would look like debugging, while trying to get shades of meaning in the hard cases from MEDRS might descend into "semantics". Actually arguing about semantics is perfectly fine in its place: my comment on "review" above was intended to say that. (The ontology of "review" has been studied formally, and I expect to get to that area in due course.)

As for machines making good "judicial" decisions, I doubt we'll crack that as an issue of principle. Actually NPOV here admits multiple points of view under some kinds of RS criteria. Where the machine says "yes" and there is a consensus here that the answer should be "no", we do have a debugging situation, and the rationale would be very interesting input into our work.

Anyway, thanks too for the kind words about the case studies page. It is obviously a first pass, only, and to illustrate how we'd try to apply some of the MEDRS discussions here. Charles Matthews (talk) 08:45, 11 July 2018 (UTC)

So, is this a rule-based approach rather than using machine learning? Bondegezou (talk) 10:17, 11 July 2018 (UTC)

Not machine learning. I know that "algorithm" often does mean machine learning these days, but we mean it in the Computer Science 101 sense, Turing machines ... When I say SPARQL, it is partly because it is the query language I understand, and partly because it is a natural way to draw in facts from Wikidata. We will be using human checking of the facts (relation extraction), by the way, and combining it with algorithmic processing of the publication metadata. Charles Matthews (talk) 10:43, 11 July 2018 (UTC)

Thanks for the details. Machine learning techniques using the full text of papers are coming on in leaps and bounds, as per doi:10.1186/2046-4053-4-5 I mentioned previously. Generally this is in a decision support role, e.g. the use of machine learning to prioritise what papers a human should look at for a review, or using the machine learning answer as a second reviewer. I would have thought that sort of approach would be valuable in a Wikipedia context, but this time selecting reviews that Wikipedians should be looking at. However, these are perhaps ideas for another project. Bondegezou (talk) 10:57, 11 July 2018 (UTC)

I'm aware of these ideas: in fact one of the authors in the paper you cite is in the group at UCL with which my company has worked. Here, though, we are working on a problem rather directly related to Wikipedia editing. Charles Matthews (talk) 13:05, 11 July 2018 (UTC)

I'd like to echo the concerns about treating Cochrane reviews as the (singular) ideal source. Although not perfect, Cochrane is the best... for what it does... which isn't everything. You might a Cochrane review for ==Treatment==, ==Screening==, ==Diagnosis==, and ==Prevention==, but you usually want a textbook for ==Classification==, ==Causes==, ==Pathophysiology== and ==Symptoms==, and a plain narrative review for ==Research directions==. WhatamIdoing (talk) 17:28, 11 July 2018 (UTC)

What kind of help is being requested?

User:Charles Matthews what kind of help is ContentMine seeking from members of this WikiProject, concretely? Would you please describe the kinds of tasks you are asking us to do? Jytdog (talk) 17:16, 11 July 2018 (UTC)

At this point, we are asking for bibliographical help: we need to find around 30K "suitable" open access titles to download and process, and the chosen approach is, firstly, tagging on Wikidata.

Later on, we'll need co-operation on our own wiki, at sciencesource.wmflabs.org, in the fact-checking process that will be carried out via an annotation system (technically similar to the one now on Europe PMC). We'll also need to understand the "expert" process of source criticism, that we are trying to mechanise. Charles Matthews (talk) 17:53, 11 July 2018 (UTC)

With respect to high quality secondary sources that are open, these are few. We have USPSTF reviews, we have AHRQ reviews, and we have PLOS medicine meta analysis.

Most literature reviews are not OA, as the journal generally pays the authors to write them.

There are also some nice overviews by the NIH and CDC. Doc James (talk · contribs · email) 20:18, 16 July 2018 (UTC)

Intended use of the resulting tool by the community

User:Charles Matthews ....three questions:

What exactly will this tool do - will it just analyze sources and report, or do you intend it to actually edit articles, or to suggest content and sourcing?

Do you imagine this as a WP:BOT or Extension or WP:Gadget or WP:Scripts or something else that will reside on WMF servers, or is this something else, that would reside somewhere else?

How do you expect it to be used and by what kind of WP editor; what kind of skills would be needed to use it? Thanks. Jytdog (talk) 17:16, 11 July 2018 (UTC)

Not a tool as such. It will be a wiki in its own right, where work goes on as described as above. The outputs from the project will be referenced statements on Wikidata. The real task for the ScienceSource project is to build a community there. Once there is a community, it will behave in a way antipodal to a bot: it will raise issues and think of other purposes for the system. We really hope.

There will be an algorithm, that is being refined, that evaluates sources. It will be pretty trivial to adapt to checking existing sourcing across Wikipedia, for example. Where there are gaps in referencing, it would be rather more work to locate a decent secondary source that might serve as a reference: but that is a known other problem, getting machine-readable topics associated to articles, and we'd have done some of the work by the initial narrowing of our "corpus".

Given that the wiki is a Wikibase site, i.e. is like Wikidata in look-and-feel, I think Wikidatans may comprise a proportion of editors. The annotation system will resemble other ones that are popular. I realise medical editors may initially be a hard sell for participation, given that they have plenty to do here. The work will be "upstream" of what goes on in daily Wikipedia life. As they said back in the Renaissance, ad fontes. The scientific sources of clinical medicine – we really are trying to shine some daylight on them. As for UX, we have a deal with the WMF involving consulting their people, so that "skills" don't become the limiting factor. Charles Matthews (talk) 18:06, 11 July 2018 (UTC)

Intended use by the company

User:Charles Matthews would you please clarify what value this project and the grant brings to the company, and how the company intends to use the tool and any data/know how it gathers while building it, to increase the value of the company or to provide new ways to serve its paying customers?

You are asking editors to volunteer to help you; please describe the value that our help would bring to the company and its efforts to remain viable as a commercial enterprise. Thanks. Jytdog (talk) 17:16, 11 July 2018 (UTC)

Obviously getting such a grant is a boost to any company in a similar position, given that (a) Wikipedia is a household name, and (b) Wikimedia is a vastly influential player in the open sector. The data will be open, and openly licensed. The software will be open source and posted to GitHub or equivalent.

ContentMine applies for both grants and contract work, and cannot do otherwise. These issues were hashed over on m:Grants talk:Project/ScienceSource#Questions about applicant and alternative funding sources early on in the grant application. I can add a bit to that, now. The wiki, we hope, can be a platform for further projects on the scientific literature. There is really no reason it needs to be a one-trick pony, but of course if the community doesn't like a new direction it won't hang around. The annotation technology is innovative in the Wikimedia context, and seems to have potential to contribute also in the development of "structured data on Commons" and edutech. We'll have to see, but Wikibase sites (Wikidata-like wikis) are a coming thing, and our version may have legs. Technical innovation is like that, and one can't really know whether the technology is going to have a "killer app".

To get back to community. I'm a Wikisource person, and the basic idea of downloading open access papers there has not taken off. The grant, for me, allows Wikimedia to have a good look at text-mining, a major technique of contemporary tech, in a setting where people can see what it is and isn't. The added value of volunteer work will be basically verification that facts said to be stated in papers are actually there when you go carefully through the language. Any third party will be able to harvest those verifications just as much as ContentMine. Charles Matthews (talk) 18:28, 11 July 2018 (UTC)

Thanks for your replies here. Thanks for linking to the questions about the company side at the grant page, where you wrote It defines itself as not-for-profit... however the listing at CompanyHouse is unambiguous that this is a business. One of the things that went very wrong with the company doing medical videos is that there was some weak signalling that it was thinking about becoming a nonprofit; it never was and it never did. ContentMine is a for profit entity; please don't fuzzify this.

It is immensely valuable to the movement, in my view, that the company intends to make the tool and the data available to all comers for all purposes. I am hearing that.

But the answers above were not clear about the value to the company outside of getting the grant itself. Can you please answer that clearly? Thx — Preceding unsigned comment added by Jytdog (talk • contribs) 19:09, 11 July 2018 (UTC)

(You need to sign.) The founder in the old thread gave the "Shuttleworth" answer to being a non-profit. In the UK, the non-profit does not exist in the same way as in the USA. You are a charity, which is a much more restricted kind of entity, or you are a business of some sort. For example, I was the first hire of Wikimedia UK in 2010. At that point they were a type of company (limited by guarantee) and were aspiring to become a charity. They got there in the end, but there were real and serious questions whether they should qualify under English&Welsh law.

So, being a business in terms of registration is what ContentMine is.

I shall do my best to find a further answer to your question, since this is clearly a matter that concerns you. Perhaps you want my take on the "business case", from ContentMine's point of view, for applying for the ScienceSource grant in January of this year. It all started back in August, when there was a strategy gathering over a weekend in Cambridge. I brought up annotations, about which I'd been talking with Zazpot, and which had been mentioned as an approach to text-mining by Peter Murray-Rust. At that point we were brainstorming about "revolutionising literature search", about how the company had technology in two directions, one being ElasticSearch run on new papers on Europe PMC, and how it needed to find a way into medicine and education because these affect basically everyone.

The MEDRS idea came up when I was on the phone to RexxS, months later. It ticked a couple of boxes, medicine clearly, and greater participation in text-mining, which was the Achilles heel of WikiFactMine, the previous project of the company. We talked in December to the WMF, and on the call community participation came up again, from the Foundation's side, as one thing they'd really like to see in a tech project.

So the business case was that: MEDRS was a clear target that would interest many people; internally ElasticSearch probably was something of a tech dead end, and had been a bit of a stretch in WFM; and there was an algorithmic slant that fitted rather well for a Cambridge company that had made one tech hit (the fatameh tool by T Arrow). I won't give you more blow-by-blow, but we had encouraging noises from Foundation people, qualified by the need to draw conclusions and "learnings" from the previous grant. Well, I wrote the learnings, frank comments I'd made were thrown around during the grant process, as you'd expect, and in the end we made it through the process. From the point of view of the company, this largely open debate (plus two interviews) has allowed for a fresh start.

I'm not that surprised to be asked again about the kind of issue you raise, and indeed I know rather more about the commercial position than I did at the start of the year. In terms of ContentMine, I don't see what to add, though. It is certainly not the average start-up. Charles Matthews (talk) 20:02, 11 July 2018 (UTC)

The concerns I am raising here consumed this board in an extremely ugly way not very long ago. Please read Wikipedia_talk:WikiProject_Medicine/Archive_110, which is only one place where the matter was discussed. I am raising the kinds of questions that people who were very, very upset in that discussion were raising, so that they are raised now, at the beginnning. I am trying to avoid us going through that again.Jytdog (talk) 20:12, 11 July 2018 (UTC)

OK, I can see you are arguing from a precedent. But not the exact relevance of that precedent to anything within my control.

You might assume my good faith, and appreciate that volunteer engagement in the project I'm introducing has been the constant theme of ContentMine's dealings with the WMF throughout. Indeed the reporting at m:Grants:Project/ScienceSource/Timeline is dominated by a need to report in detail on all the many efforts ScienceSource has said it will make to engage volunteers.

We were offered the full grant we applied for, in the middle of May. but conditional in particular on submitting a detailed publicity plan to the WMF. The metrics in that (submitted) plan are as given in the long second table. (Full disclosure: the grant-runner has not yet signed off on the plan.)

Perhaps this will have some clarifying value for you. If the precedent you see is of commercial promotion on Wikipedia, and you see an analogy, there is this: the Foundation has made the grant to ContentMine, in their kind of phrasing, while trusting the company on tech, but regarding its track record as weak on self-promotion. In short, the driver of something that may be bothering you is the WMF. Charles Matthews (talk) 20:45, 11 July 2018 (UTC)

Nothing is bothering me and I am not arguing anything. I am trying to help get a full disclosure on the table. That full disclosure would have been a useful starting point for discussion to get further buy-in from the editing community.

You do not understand where I am coming from with regard to companies doing work in Wikipedia generally, nor with respect to the issues here.

I think people from companies have a place in WP, but their presence and activity have to be managed very carefully, with very clear disclosure and full buy-in from the editing community. I was not one of those, who was very upset over the Osmosis thing. I listened to those who were.

Based on your approach to this discussion, I will not be participating in this project.

I recommend that folks here steer very, very clear of this project. Jytdog (talk) 21:05, 11 July 2018 (UTC)

I'm naturally sorry to hear that. You may not, in turn, understand where I come from on COI. I was deeply involved in drafting WP:COI in 2006, and commented at the time that the issue was going to be of increasing importance to Wikipedia. Charles Matthews (talk) 21:12, 11 July 2018 (UTC)

I don't care where you are coming from nor what you have done in the past. Your approach to this discussion is all I needed to turn and walk away. Jytdog (talk) 21:37, 11 July 2018 (UTC)

In the cold light of early morning now, it seems to me that I can do little good in returning to this thread; but on the other hand that I need to. I should at least note that my reply just above, mentioning COI, was not in fact to the text immediately above it, where the language on "full disclosure" now stands: "full disclosure" was introduced in one of the series of edits marked "ce" that can be seen in the page history, and my reply was to a previous version of what Jytdog wrote. I take it that from full disclosure we should be reading this as "full disclosure, the acknowledgement of possible conflicts of interest in one's work", but applied to the company. Yes, it could be a problem if the company's "intended use" of the grant had some conflict of interest with the missions of Wikimedia (WMF or movement), or Wikipedia. Companies can be led to behave in ways that "increase the value of the company", and so on, in the terms in which the question was first posed.

It has been noted before that the project has little to do with Wikipedia editing, and in Wikimedia terms is centred round Wikidata. I frankly don't see that ContentMine has any conflict of interest in what it intends to do in relation to Wikidata.

It is ironic to me, at least, that this is a tale of two guidelines, running from MEDRS, where I'm actually just learning the ropes, to COI, where the distinction between potential and actual COI in Wikipedia terms is burned on my Wikipedian heart. Charles Matthews (talk) 04:22, 12 July 2018 (UTC)

Nope. If you don't understand something you should ask.

If you had taken the time to read the Osmosis discussions (which you obviously have not) you would have seen that one of the things that upset people here the most, was the way that the company was leveraging the relationship and the work for its own benefit in ways that were not fully explained up front. that is one of the things that was upsetting people.

That is what this section was intended to get stated now. I was asking you to "fully disclose" how the company would benefit from this relationship and how it would use the products, such that those kinds of questions would not arise only after the fact.

The response only makes it more clear to me that I have no desire to be involved with your project, and intensifies my warnings to others to steer clear of this. Jytdog (talk) 18:01, 12 July 2018 (UTC)

I'm confused Jytdog as to what is not being disclosed. What exactly is full disclosure to you? Expecting a perfect summary of each possible way an entity may benefit from a specific action is naïve, and I find Charles Matthews answers to be exemplary. Even if we demanded a review of each possible way something may benefit the entity — we could never get a satisfactory answer, because giving such an answer is impossible as it is not possible to know what the potential benefits may be. The fact that he acknowledges COI is enough, and per our guidelines (WP:MEDCOI) it does not bar him from making contributions.

I think it is totally unreasonable to suggest anyone read about the Osmosis case, and it would be nothing but a massive waste of time. That whole debacle was a tempest in a teapot — and the amount of text is enough to keep someone occupied for days straight just reading back-and-forth from a heated and sometimes quite irrational debate. There were some potential issues, but nothing near what it was turned into — and even then — the Osmosis case is so dissimilar to this case as to be entirely irrelevant.

However seeing as you seem to have identified some problems, would you please opine on what you think is "leveraging work of the community too far in favour of a company", and whether there are any specific responses that are problematic? Saying that someone hasn't taken time to read about every possible past case or that disclosure is not "full" from the get-go — as well as saying that someone "should ask", when they are engaging in discussion — is not constructive. Carl Fredrik ^talk 19:16, 12 July 2018 (UTC)

You are wanting rehearse the Osmosis thing and I am not going there. You (and everyone else) will of course react to this proposal and to these responses how ever you see fit. Jytdog (talk) 19:33, 12 July 2018 (UTC)

Points from me, and thanks to Carl Fredrik for giving a third opinion:

• I'll take up the logo thing, since at very least it is possibly misleading.
• I reject the implication made of an oppressive form of argumentation, which seems odd given the way I linked to the related issues raised by John Cummings.
• Some of what is said above is hurtful, but I'm content just to move on now.
• The Osmosis reasoning seems to me guilt by association, and while I didn't know about it as a matter of background, and I don't mind frank discussion about my own actions, I'm not happy with the way it has been used here as a slur against ContentMine. Charles Matthews (talk) 03:47, 13 July 2018 (UTC)

You are still not hearing the issue. Osmosis and the people here working with them, were all going along happily in good faith. Harsh questions were raised later, that were all the more harsh because' they were being asked later. There is no "guilt by association". What there is, is a parallel situation that has high potential to end up in the same place. months from now. I don't want to see that happen. You don't want to learn from the past; that is your deal, and all the more reason to steer clear of this project. Jytdog (talk) 19:13, 13 July 2018 (UTC)

That is your take. You have asserted but made no case for a "parallel", and I have objected strongly to your attempt to draw it. You continue to discourage participation, but I hope people will take your comments to be alarmist and unfounded. You have had detailed replies on the concrete points you have raised. Charles Matthews (talk) 07:27, 14 July 2018 (UTC)

I do understand, that you do not get it. You also seem to be under the impression that Osmosis was somehow "bad". They are just a company, doing what companies do. The problem was how things were handled, which is what led to the blow up. The parallels are obvious. Osmosis was creating openly licensed videos with health content; you are creating an openly licensed database with refs about health and you've made some hand-wavy notions that this could be useful to us here one day. The relationship with Wikipedia was great for Osmosis' marketing; your first response in this section emphasized the benefit of the association. Osmosis used the videos they posted here elsewhere (youtube). They had their branding prominently on the videos (here and there). These last two things - subsequent use, and branding, are things I remain unclear on with your proposal. It is pretty obvious to me that there is huge value to a curated database of 30K MEDRS refs (yes they are open, but nobody will understand the database as well as ContentMine). How else will the company use that? On the "wiki" you mention, is that going to be ContentMine-branded? These were the kinds of issues that really upset people - especially the branding here in WP (that part doesn't seem to be that relevant here, not now at least). If I were in your shoes I would have read the archived Osmoais page about 10 times to make sure I understood what went wrong and so I could avoid it. (if you do read it, you will see that i was more upset with the people who were upset, than i was with the actual videos. It was however clear to me that the relationship between Osmosis and the community was not handled well, by anyone (not them, and not the people here working with them) Hence my questions and hence my dismay at your response and approach. btw you might also want to read this from a recent Arbcom case about COI in WP. Jytdog (talk) 08:03, 16 July 2018 (UTC)

Re the ArbCom case, then, as summarised in one of the initial statements: "paid editing and the potential of paid advocacy meatpuppetry by an administrator and SPI clerk as well as the potential misuse of the admin toolset to take actions he was specifically paid to do". I circulate a mass message here, Facto Post, using admin privileges. A newsletter for the WikiFactMine project was a specific request from the WMF grantrunner. On "paid editing", Wikimedians in Residence naturally do that, but within the terms of use of the site. I don't do any advocacy editing here on enWP in or related to the article space.

I'm engaging here because of definite commitment in m:Grants talk:Project/ScienceSource#Comment 4, which has become in part the "case studies" page you commented on above, and the intention to bring further issues here. We are building an expert system and WT:MED is where the relevant experts hang out. These associations are natural. Some of your remarks would have helped, as arguments, under Comments 10, 11 and 13.

On the scope of the project: yes, we want to build a "better mousetrap" and have people come see. Considered as a research project, which this is in part, that is what you'd expect?

I do have to catch a plane to Cape Town in a few hours. Parsing out your concerns further, there is COI. My conscience is clear, and I did have three years experience on ArbCom. I'm a meatpuppet for no one. I'm coauthor of How Wikipedia Works, but have never mentioned it on my user page here, because I assumed someone would shout "hypocrite" some day.

On some sort of "hidden agenda", there is a very public agenda to do more with the downloading techniques. This side has been floated both in Facto Post and in the first of our video set ("content management system" for OA). It would be timely and would probably require both a new architecture and a community of willing tech-savvy people. The logo on the wiki would be the one above, with the CM logo integrated. If anyone objects, really, we can change the logo on Commons. That is as far as branding goes, for me. It is fair comment that ContentMine would gain expertise on MEDRS and its implementation: the grant proposal states that we want to help recruit medical editors ("Such a corpus will assist Wikimedians in writing and referencing medical content, to a high standard, and closely linked with Wikidata's science and metadata content.") We can hardly do that without understanding. That understanding has value: see the lead section of systematic review, for example. There is no agenda here that could be effectively hidden.

Charles Matthews (talk) 09:20, 16 July 2018 (UTC)

So instead of addressing how the company would use the fruits of volunteer labor and branding here, with us, you brag. The purpose of linking to that particular section was that the subject admin had the insight (after the fact) to realize that virtue and experience are not some kind of a get-out-of-jail-free card with respect to doing the work of managing a COI. Which means working through that stuff here, not comtinually pointing us elsewhere. You entirely missed the point. I give up trying to explain. Jytdog (talk) 10:32, 16 July 2018 (UTC)

Actually if you mean COI in the enWP sense, the relevant here would be WP:COIN, I think. If your implication is that the ScienceSource project as run on its own wiki would be exploitative, which is what your language suggests, we can indeed address that here, but then I don't see the exact relevance of the case.

You are doing your best to scare people off, but rather than bragging, I believe we were agreeing that the volunteer work would be of value. Once the wiki is configured, which is not something within my control, I would think that those who came to the wiki might very well raise those points or similar ones, which I'd be glad to address. Better than simply filling up more of WT:MED.

As CFCF pointed out above, setting some sort of "disclosure" standard that is impossible to meet is one way to be somewhat unreasonable. As is some sort of "conflict of interest" standard that allows in all sorts of hypothetical commercial uses that might possibly arise, as if this was a multiverse exercise in trying to come up with obscure scenarios. Which does indeed bear a resemblance to types of business thinking, in relation to blue-sky research. But I don't think I'm missing the point. Charles Matthews (talk) 11:07, 16 July 2018 (UTC)

Nope. You don't understand what I am trying to communicate. Jytdog (talk) 22:01, 16 July 2018 (UTC)

Note about the WMF vs en-WP vs WP:MED

User:Charles Matthews I see that the company is already citing the WMF wikimedia movmement logo on its partners page. Please be aware (if you are not) that the WMF is not the same as the English Wikipedia (or any other WMF project) and that each project governs itself. Please be careful that the company avoids saying things like it is "partnering with Wikipedia" or "partnering with WikiProject Medicine in Wikipedia" unless you get absolutely clear buy-in from the editing community. (something like an RfC perhaps) We just went through a very ugly thing (which I am not trying to revive, everybody!) where a company that was uploading open-licensed medical videos to WP was... overly aggressive in claiming a relationship with Wikipedia and this with this wikiproject in its promotional materials, which was part of what led to a severe backlash here (which is still recent and somewhat hot). OK? (I am posting this here, so it is very public, and very unambiguous) Jytdog (talk) 17:16, 11 July 2018 (UTC) (fix per comment below Jytdog (talk) 23:37, 11 July 2018 (UTC))

That's the Wikimedia movement m:logo, not the foundation's logo. The foundation's logo has a single, uniform color (usually 100% black, not multi-colored) and says "Wikimedia Foundation" underneath it. WhatamIdoing (talk) 18:11, 11 July 2018 (UTC)

Jytdog, thanks for the heads-up anyway. I'm not management, I am a 15-year Wikipedian, I'll try to keep them straight. Charles Matthews (talk) 18:30, 11 July 2018 (UTC)

Not a 15-year-old Wikipedian, otherwise you're really not wearing well, Charles, --RexxS (talk) 19:06, 11 July 2018 (UTC)

If it's ageist jokes now, you dinosaur, I'm going to ask whether you were suited up in that Thai cave. Charles Matthews (talk) 20:51, 11 July 2018 (UTC)

User:Charles Matthews, away above you wrote my company; just above here you wrote I'm not management, I am a 15-year Wikipedian. The lack of clarity and hat-switching is very unhelpful and is the kind of thing that will cause major problems later. Jytdog (talk) 18:06, 12 July 2018 (UTC)

Surely not. "My company" = "the company I work for", not "the company I own", certainly. My position is defined as Wikimedian in Residence. Roles in the team were defined at m:Grants talk:Project/ScienceSource#Comment 6 and hold with the slight revision that the "work package owner" parts of job titles are not being used. Cesar Gomez is the manager. Charles Matthews (talk) 09:00, 13 July 2018 (UTC)

Please be careful how you describe the company and your relationship to it. Jytdog (talk) 19:15, 13 July 2018 (UTC)

yes, 'clarity' in your relationship to it is best--Ozzie10aaaa (talk) 10:54, 26 July 2018 (UTC)

Project page

User:Charles Matthews please move Wikipedia:WikiProjectMedicine/Referencing case studies for MEDRS out of WP:MED space until you gain consensus here that this WikiProject wants to be associated with ScienceSource or ContentMine, to the extent of hosting its project page.

This is exactly the kind of association that I tried to advise you against claiming above. Thanks. Jytdog (talk) 15:21, 26 July 2018 (UTC)

Pathognomonic NOT pathognomicity is the word that should be defined.

I have heard and used the first word lots of times, NEVER the second. It may technically be a word, but makes no sense not to have the common word defined, but just this useless variant.

I agree the previous article/definition was terrible, but should we throw out the baby with the bath water? Resurrect the proper word and lets have a go at it again, OK?

Someone more competent at this than me needs to enter the word to get us started (I'm looking at you oh great editor in the sky!) — Preceding unsigned comment added by Mcollister (talk • contribs) 17:57, 18 July 2018 (UTC)

• Pathognomonic

Fixed, ping Mcollister. Carl Fredrik ^talk 18:26, 18 July 2018 (UTC)

• What just happened here? Natureium (talk) 18:54, 18 July 2018 (UTC)

The article was moved in January of this year to Pathognomicity. That would in theory be the noun form, but it is incorrect per WP:COMMONNAME. Carl Fredrik ^talk 19:39, 18 July 2018 (UTC)

I see that WP:NOUN is still in the article titles policy, and presumably it applies here, too, just like it does for all the other subjects that are more commonly seen in their adjectival forms. WhatamIdoing (talk) 20:23, 18 July 2018 (UTC)

WP:COMMONNAME > WP:NOUN — That's even made explicit… … … Carl Fredrik ^talk 20:32, 18 July 2018 (UTC)

Where do you see that? It can't be in COMMONNAME, which doesn't mention adjectives at all, but which says in two different paragraphs that the common name is not the only criterion that should be considered. WhatamIdoing (talk) 04:52, 19 July 2018 (UTC)

WP:NOUN reads: "Nouns and noun phrases are normally preferred". The adjective form doesn't even exist, it gets 0 hits in google N-gram [1]. We don't even have a dictionary telling us that the word exists — at all. Carl Fredrik ^talk 06:26, 19 July 2018 (UTC)

WP:COMMONAME reads: "it generally prefers the name that is most commonly used", which undercuts your claim that it was "made explicit" that COMMONNAME overrides NOUN.

Now, if the word doesn't really exist, that's a serious argument against using it as the title. But that's a separate question. WhatamIdoing (talk) 18:01, 19 July 2018 (UTC)

(edit conflict) However, it would be correct per WP:NOUN: "Use nouns: Nouns and noun phrases are normally preferred over titles using other parts of speech ... Adjective and verb forms (e.g. elegant, integrate) should redirect to articles titled with the corresponding noun (Elegance, Integration). In this case, though, if there's a massive disparity between the occurrence of pathognomicity and pathognomonic, my instinct would be to go with the commonest name. It might also be worth pointing out that these are articles, not definitions; we have Wiktionary (which is happy with adjectives) for that. --RexxS (talk) 20:26, 18 July 2018 (UTC)

I routinely get frustrated by cases like this. People resort to wp:COMMONNAME as if it was the only important criterion under wp:AT. Nonsense. It is only one of five wp:CRITERIA, and has no more import than the other four: naturalness, precision, conciseness, and consistency. Please take the time to consider all five, not just one. LeadSongDog come howl! 22:33, 18 July 2018 (UTC)

I did, and it falls under the adjective form according to all WP:Criteria. The verb form doesn't even exist, it fetches 25 real results On Google — none of which are definitions. One would at least assume it existed in a dictionary.

We don't make up words for the sake of adhering to a policy. This should be a non-debate. Carl Fredrik ^talk 06:26, 19 July 2018 (UTC)

We don't create articles to provide a wp:DICDEF either. Perhaps this should be reframed as List of pathognomic signs and/or List of pathognomic symptoms, or (per wp:USEENGLISH) List of disease-specific signs and List of disease-specific symptoms LeadSongDog come howl! 17:14, 19 July 2018 (UTC)

It seems that the correct noun form is pathognomy, which is unfortunately spelled exactly the same as a different kind of Pathognomy (a thing from the phrenology era that is still talked about in the performing arts). It's mentioned in at least the 2002 edition of Stedman's. WhatamIdoing (talk) 18:08, 19 July 2018 (UTC)

If that's the best title, should we move the current article to Pathognomy (psychology) or the article being discussed to Pathognomy (medicine)? I think Pathognomonic signs without the "list of" would also work. I think "disease-specific signs" is less clear, and the intended audience of this article isn't going to be lay people. Natureium (talk) 18:23, 19 July 2018 (UTC)

Except when it's symptoms, not signs. Carl Fredrik ^talk 18:56, 19 July 2018 (UTC)

P.S. While I would prefer pathognomonic, I have no qualms with pathognomy (medicine), just not pathognomicity. Carl Fredrik ^talk 09:24, 20 July 2018 (UTC)

Is is surely pathognomy and not pathognomony? Natureium (talk) 17:23, 20 July 2018 (UTC)

It is https://en.oxforddictionaries.com/definition/pathognomy WhatamIdoing (talk) 19:15, 20 July 2018 (UTC)

Health effects of green tea

Green tea( 3 appearances)

• Green tea (edit | talk | history | protect | delete | links | watch | logs | views)

There's some discussion about how to present these, which could use more input. I have a couple of concerns:
— First, a heavy reliance on Chinese papers which seem to have a more sensational view of the health impact of green tea (a mainstay of traditional chinese medicine) than non-Chinese sources.
— Secondly, a heavy reliance on reporting of what green tea consumption is "linked to" (i.e. correlations). For example we say "Daily consumption of green tea is significantly correlated with a lower risk of death from any cause". Presenting such material in a health effects section gives the impression that there is a cause-and-effect here – we see this kind of writing in newspaper reporting of health topics a lot, and is not something I think we should be replicating, since lay readers may find it confusing.
Alexbrn (talk) 17:59, 19 July 2018 (UTC)

I think it would be possible to say that there is a correlation, and also say that correlation does not prove causation. I don't think that it is necessary to remove well-sourced information in such a case. WhatamIdoing (talk) 18:12, 19 July 2018 (UTC)

Thank you WAID, that's precisely what I said on the talk page. I'm also adding some non-Chinese studies that cite the Chinese studies that concern Alex since these review articles from well-respected journals seem to think they're legitimate. We reflect what the sources say. TylerDurden8823 (talk) 22:27, 19 July 2018 (UTC)

NIH and NICE search(first 2 results)--Ozzie10aaaa (talk) 00:13, 20 July 2018 (UTC)

No problem using decent sources, but I wonder how a statement like "a high amount of green tea consumption may be associated with a lower risk of liver cancer in Asian women" actually conveys any knowledge. Wikipedia is WP:NOTEVERYTHING. How are we going to add WAIDs suggested text stating the correlation does not mean causation? Alexbrn (talk) 16:16, 20 July 2018 (UTC)

The statements are already heavily qualified. I have no idea what you mean about statements not conveying knowledge and I fail to see how that's a helpful comment. I don't think WP:NOTEVERYTHING applies to the article's current content at all. It seems like if it's not one thing then it's another... TylerDurden8823 (talk) 16:43, 20 July 2018 (UTC)

Alex, you seem to have left off the beginning of that sentence, which says "A 2015 meta-analysis of nine prospective cohort studies concluded that a high amount of green tea consumption.... The conclusion of a recent meta-analysis sounds like "knowledge" to me. Is it the final word? No, there will probably be clearer information in the future, e.g., if some bright researcher combines information such as that found in PMID 26028072 with those results to wonder whether the difference is HBV, or a different one wonders whether high consumption of green tea by women means limited alcohol consumption, whereas for men perhaps it means limited water consumption but still a cancer-causing amount of alcohol. But even though we don't have the final word, we do have some information: a "MEDRS-style" source says that high consumption is associated with less liver cancer in Asian women. We can say that. WhatamIdoing (talk) 17:07, 20 July 2018 (UTC)

Well, yes - I left it off because we usually omit all that gubbins about "A systematic reviews said ..." &c. and merely assert what is known, as a matter of good style - and the source concludes the findings are merely indicative, and is full of caveats which we do not mention. And how would you "say that correlation does not prove causation" in these cases anyway, even once the source was represented honestly? Alexbrn (talk) 17:30, 20 July 2018 (UTC) — Preceding unsigned comment added by 109.175.105.69 (talk)

Well, I'd probably write something like "Drinking more than five cups of green tea each day has been associated with a slightly lower risk of liver cancer among women living in Japan, China, and Singapore, but not among men; however, it is not know whether this small benefit is due to the green tea or to other factors."

If you wanted to dwell on this subject, you could use https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5411268/ (which directly discusses the currently cited study) to add the relevant context: this slight benefit is unlikely to be found in people at low risk of HBV and HCV infection (the relevant factor appears to be "living in these countries", not "being Asian"), and that screening, treatment, and vaccination against those two viruses would produce a far more significant reduction in liver cancer than the slight effect of drinking a lot of green tea. WhatamIdoing (talk) 19:41, 20 July 2018 (UTC)

Alex, I strongly disagree with the notion that omitting "all that gubbins" is "better style". I think it's important to mention the type of study being cited. There are plenty of Wikipedia pages where this occurs. As I said earlier, these statements are already heavily qualified. Furthermore, your messages are mixed whether you want things shorter or longer. Are you suggesting more nuanced descriptions or more concise ones? Commit to a side and be specific. The sources'respective conclusions are represented honestly. If you think they aren't, provide a specific example of how they aren't. I have no objection to WAID's suggestion above. TylerDurden8823 (talk) 19:54, 20 July 2018 (UTC)

Omitting "all that gubbins" is nothing to do with style. It's a consequence of WP:ASF - we state uncontested facts; and we attribute opinions where differing opinions are equally valid. Mixing those up is usually the hallmark of a POV pusher who has either attributed something unnecessarily to try to make it look more impressive, or has not attributed in order to ignore a contradictory valid view. --RexxS (talk) 22:41, 20 July 2018 (UTC)

Rexx, I really hope you're not suggesting I'm a "POV pusher". If you are, I'll remind you of WP:FOC and WP:5P4. The fact remains that, as written, the article is stating facts about the conclusions these studies reached. We don't say (explicitly or otherwise) that causality is established because it hasn't been nor has it been ruled out. The presented information is consistent with the policy you cite (WP:ASF). TylerDurden8823 (talk) 22:55, 20 July 2018 (UTC)

What is the matter with you? Of course I was not aiming at you or anyone in particular. I was merely relating an observation of what I've seen all too often in controversial medical articles (and I'm pretty sure I'm not the only one who's observed that). I have no idea who made the mistake of attributing when a simple statement was sufficient, and I don't really care. I'll remind you of WP:AGF in return.

Now, looking at the article it presently says "In a 2015 meta-analysis of such observational studies, an increase in one cup of green tea per day was correlated with a 5% lower risk of death from cardiovascular causes." Either:

1. that conclusion is uncontested, in which case the "In a 2015 meta-analysis of such observational studies" attribution breaches WP:ASF – "When a statement is a fact (e.g. information that is accepted as true and about which there is no serious dispute), it should be asserted using Wikipedia's own voice without in-text attribution"; or
2. that conclusion is disputed or subjective, in which case the contrary view needs to be presented, per WP:ASF – "When a statement is an opinion (e.g. a matter which is subject to serious dispute or commonly considered to be subjective), it should be attributed in the text to the person or group who holds the opinion."

Make your mind up which it is, but don't tell me "The presented information is consistent with the policy you cite (WP:ASF)", when it patently isn't. --RexxS (talk) 23:20, 20 July 2018 (UTC)

@RexxS:, take a breath and write nicely. I don't appreciate the condescension (it's uncivil) and there's no place for uncivil tones here on Wikipedia. There is absolutely nothing wrong with me. Your comment sure sounded like it was being aimed at me (I suggest rereading it). If it wasn't aimed at me, then I recommend more caution with your wording going forward. I've worked on plenty of controversial medical articles and seen POV pushers too but I think it's crucial to be mindful of how what we say can be interpreted. I'll remind you once again to focus on content instead of me.

As for ASF, it says: "When a statement is a fact (e.g. information that is accepted as true and about which there is no serious dispute), it should be asserted using Wikipedia's own voice without in-text attribution." The problem here seems to be perspective. I have made up my mind. They're opinions/findings of these articles because they're clearly considered controversial (more so for the cancer subsection). They're not uncontested facts and there's no debate about that. Historically, if I tried to trim the attributions, I get push back that I'm trying to suggest that causality is established or saying this in Wikipedia's voice. If I add them back and attribute them to the studies, then it's labeled as clunky. I would request Alex make up his mind instead since it seems there's no method of satisfying his wishes short of eliminating whole sections and oversimplifying the article. The contrary view is presented (the first sentence of the very subsection under discussion outright states there is no definitive evidence of green tea preventing or treating cancer, Rex) and the statements about the studies are (again) heavily qualified. So, I have historically tried to be consistent with ASF but it feels like my hands have been tied because I get vigorous objection no matter which way I put it. TylerDurden8823 (talk) 23:40, 20 July 2018 (UTC)

Don't patronise me. I've spent far too long trying to defend this project from civil POV-pushers to be impressed by your faux advice. If you're reading tone into my posts, that's your problem, not mine. Let me repeat myself, as it seems you didn't hear me the first time: Of course I was not aiming at you or anyone in particular. Did you get it this time?

Back to the content. You've decided, have you, that an increase in one cup of green tea per day was correlated with a 5% lower risk of death from cardiovascular causes." is an opinion, not a fact? Okay, then cite the source that dissents from that opinion, and attribute each of them in the text to the person or group who holds the opinion. At present, it's neither one thing or the other. This isn't the cancer section; it's the cardiovascular section, and either we have good evidence for green tea's ability to prevent CVD, or we haven't. All I see at present is a lot of innuendo: "a correlation between"; "correlated with a 5% lower risk of death"; "may be correlated with a reduced risk"; "appears to lower systolic and diastolic blood pressures". To be honest, I'd rather we said nothing than present that sort of hedging. Are you happy with that section? --RexxS (talk) 00:13, 21 July 2018 (UTC)

RexxS, I think you're setting up a sharp contrast between "uncontested" and "disputed", when reality is fuzzier than that. I've done a little poking around, and it appears that the following are fair statements for the current mainstream beliefs on the subject:

1. The 2015 paper came to certain conclusions (i.e., the ones described in the Green tea article); nobody contests that.
2. There is a high likelihood that the conclusions of the 2015 paper are an accurate description of objective reality, especially to the extent that you understand the conclusions narrowly (e.g., only in the specific populations, only for very high consumption sustained over years, etc.).
3. The benefit reported in the paper likely applies only to people with HBV or HCV infections. (Tyler, note that this means it's not being "of Asian descent" that matters; what matters is "anybody, of any race or ethnicity, who is living in China and Japan, where these infection rates are high").
4. The benefit reported in the paper is much lower than the benefit of preventing or treating HBV/HCV infections.

Now to talk about whether we can call that "uncontested": Smart people know about replication problems, so a single paper isn't persuasive to them, even if that paper will later be proved to be an accurate description of objective reality. The fact that it is accurate doesn't mean that people currently believe that it is accurate. As a result, the claim isn't disputed, but it's not truly accepted at the moment either. People are waiting to see what the next paper says, without taking sides, and I'm hoping that the next round of research specifically considers viral status as a key factor. (And because of point #4, even if healthcare professionals believe it, it's not likely to change their daily practice, because they have better options.)

Moving specifically to the social reality on Wikipedia, no matter what the substance, if it was ever used in TCM for any purpose, then some editors will be even more skeptical of it having any beneficial health effect whatsoever. If we were talking about some brand-new, purely synthetic molecule, and we had a meta-analysis by someone who didn't work for the manufacturer, we'd just report the results as they are: A huge dose has a small effect in this population. But since TCM has doubtless recommended a cup of every kind of tea imaginable at some point for something, including just being tired or thirsty, there is a lot of social pressure here to show off our anti-altmed biases by ostentatiously hedging the statement: This one study found that a huge dose has a small effect in this one population, but, hey, there are probably other studies in other populations that didn't find this small effect. We can't engage in social signaling if we plainly state what appears to be a tentatively accepted fact, namely that drinking a ton of green tea every single day has a small benefit for a single kind of cancer in a specific population.

And, in this case, JzG has just blanked all mention of it, leaving the section to rely on older sources that say that green tea in general (not just high amounts) probably does something for cancer in general (not just liver cancer in one population) but nobody's really sure, on the grounds the 2015 study is, in his personal opinion, at high risk of bias. I assume that this is because the authors are employed by a hospital China. I might agree that their study was at high risk of being the victim of publication bias if it had concluded green tea was carcinogenic, but a meta-analysis is much easier to independently replicate than a primary study, so its conclusions are very likely as accurate and unbiased as the published literature allows. WhatamIdoing (talk) 06:16, 22 July 2018 (UTC)

• We need to be very careful in using any Chinese-authored papers on anything at all. The evidence clearly shows that they have massive systemic bias. I would be comfortable with excluding any Chinese study from any article on Chinese medical practices unless it is corroborated by regionally independent review studies. Guy (Help!) 08:49, 21 July 2018 (UTC)

yep--Ozzie10aaaa (talk) 10:36, 21 July 2018 (UTC)

Do what you want. I've lost interest. TylerDurden8823 (talk) 08:29, 22 July 2018 (UTC)

Water allergy

Just FYI, a recent thread: Talk:Water#H2O_molecule_as_an_allergen? Brandmeister^talk 19:20, 19 July 2018 (UTC)

commented--Ozzie10aaaa (talk) 23:45, 20 July 2018 (UTC)

closed, reference desk question. Jytdog (talk) 14:16, 21 July 2018 (UTC)

bestpractice.bmj.com

Can anyone who has access to this website send me everything on treatment/management – or just all of the content on narcolepsy if it's available in one pdf file – from the following link? https://bestpractice.bmj.com/topics/en-us/428 My email is commented out in the source of this section. Seppi333 (Insert 2¢) 23:35, 20 July 2018 (UTC)

sent. Jytdog (talk) 14:15, 21 July 2018 (UTC)

Domestic violence article -- RfC on NPOV

Opinions are needed on the following matter: Talk:Domestic violence#Does the article lend undue weight to women as victims and/or their use of self-defense as a reason for domestic violence?? A permalink for it is here. Flyer22 Reborn (talk) 12:10, 21 July 2018 (UTC)

• give opinion(gave mine)--Ozzie10aaaa (talk) 10:51, 22 July 2018 (UTC)

Korean melon tweet from "Wikipedia" account

I happened to notice this Wikipedia tweet: "The seeds of the Korean melon have been investigated for use in controlling diabetes." https://twitter.com/Wikipedia/status/1020854694099353602 However this content was removed in February as non MEDRS here from what I can see. JenOttawa (talk) 02:51, 22 July 2018 (UTC)

Yeah... I noticed that on Facebook. Seppi333 (Insert 2¢) 10:43, 22 July 2018 (UTC)

@Zefr: That's your area – you can deal with it. Seppi333 (Insert 2¢) 11:20, 22 July 2018 (UTC)

@Zefr: I left them a note on twitter.JenOttawa (talk) 12:06, 22 July 2018 (UTC)

No response on twitter [2]. @Doc James:, what do you think? JenOttawa (talk) 12:34, 28 July 2018 (UTC)

Redirect Psychiatry task force talk page here?

I dropped off a "please watchlist this problem article" request over at Wikipedia talk:WikiProject Medicine/Psychiatry task force. In looking the page over, I now see that there's been no interactive discussion there by taskforce participants since May 2017. It might make more sense to redirect its talk page to WT:WikiProject Medicine. This is pretty common for taskforce/workgroup talk pages when they turn moribund.  — SMcCandlish ☏ ¢ 😼  11:19, 22 July 2018 (UTC)

Yes, would make sense. Do we have any active taskforces? Doc James (talk · contribs · email) 11:43, 24 July 2018 (UTC)

Wikipedia:Articles for deletion/David Alan Ditsworth

This AFD could use the input of some knowledgeable editors (especially if they have access to the sources listed in the article...) Thanks! --Randykitty (talk) 20:55, 25 July 2018 (UTC)

• give opinion(gave mine)--Ozzie10aaaa (talk) 10:56, 26 July 2018 (UTC)

Mini-tablet (drug)

Mini-tablet (drug) was recently moved from user to main space and I came across it while doing WP:NPP. The user has no user or talk page and this article is their only true contribution. I don't feel equipped to patrol this for NPP but hoped someone here might be able to weigh in. Best, Barkeep49 (talk) 05:51, 26 July 2018 (UTC)

• currently at Draft:Mini-tablet_(drug)--Ozzie10aaaa (talk) 10:59, 26 July 2018 (UTC)

Do we need this article? It seems like it could all be covered at tablet (pharmacy). Natureium (talk) 14:06, 26 July 2018 (UTC)

It'd be a pretty easy merge target, if the length won't overwhelm the existing one. (It could also be kept as a separate article, since it has a separate history. These small pills are on the market mostly because governments demanded that some drugs be accessible to children and people with disabilities, and not just because someone decided to give a name to each size of pill.) Whatever we do, it should probably also get some links from articles about pediatrics and disabilities.

User:Barkeep49, you said that this was "promotional". I can't figure out what your concern is, which I suspect means that most editors won't be able to figure out what you're concerned about. If a second look confirms your opinion, then you should leave a note on the talk page to identify specific problems (a couple of examples should be enough). Then experienced editors can form a consensus about whether they agree with your concerns. As it stands, I don't know if your concern is more like "It name-checks the main non-profit organization that supported the EU requirements" or "It says that this pill style, which was designed to have certain advantages, actually has some advantages" or something else. WhatamIdoing (talk) 16:29, 26 July 2018 (UTC)

@WhatamIdoing: I moved it to draft and included the promotional note in the log after discussion with someone who does some MEDS related editing and also NPP. In looking over the article again this morning that was probably not a fair comment to have made as it doesn't seem to be pushing any given product. Best, Barkeep49 (talk) 16:55, 26 July 2018 (UTC)

Thanks for the rapid update. I agree with you that the article needs some work – something's a little off, and it doesn't feel like an encyclopedia article. WhatamIdoing (talk) 19:23, 26 July 2018 (UTC)

Image was copied and pasted from a source. Some of the text was copied form here verbatum. https://books.google.com/books?id=xgHABAAAQBAJ&pg=PA194&lpg=PA194 Based on these concerns I have deleted the article in question. Doc James (talk · contribs · email) 12:46, 27 July 2018 (UTC)

Discussion of sources

• Talk:Toxoplasmosis#Primary_sources_and_popular_press

Others thoughts? Doc James (talk · contribs · email) 11:10, 26 July 2018 (UTC)

It's churnalism and complete garbage: "... of those students tested, those who were infected were 1.4 times more likely to be business majors". Then they draw conclusions about risk based on that, and they expect readers to take them seriously? They've clearly been watching too many Tom Cruise films. --RexxS (talk) 14:05, 26 July 2018 (UTC)

Polychromia

I came across this article in the process of going through short orphan articles, and I don't really have enough expertise to assess the usefulness of this article (even with sources). Is it a legitimate term that needs a separate article? Would it be better off merged elsewhere, or soft redirected to wiktionary as a dicdef? I can do any necessary legwork; I only need to be pointed in the right direction. Thanks in advance. ♠PMC♠ (talk) 16:55, 27 July 2018 (UTC)

• Baran, Robert; Berker, David A. R. de; Holzberg, Mark; Thomas, Luc (2012). Baran and Dawber's Diseases of the Nails and their Management. Hoboken: John Wiley & Sons. ISBN 9781118286708. Retrieved 27 July 2018.--Ozzie10aaaa (talk) 17:35, 27 July 2018 (UTC)
• There is only one mention of the term in all the books and articles indexed by ClinicalKey, and it's not in a context having anything to do with skin. "Blood films are not specific, and show typical features of hemolysis including prominent reticulocytosis and polychromia, but the red cell morphology may be normal." -Blood and Bone Marrow Pathology, Second Edition. Pubmed only brings up an article about fish. Natureium (talk) 17:40, 27 July 2018 (UTC)
  • I wonder if that blood-cell ref is a typo, and that polychromasia was intended.
    I can confirm the skin-related sense in a few sources, but the article's sole sentence is misleading. It's more like "a multi-colored mole is probably skin cancer" than "a person's whole body could look like a calico cat". WhatamIdoing (talk) 18:21, 27 July 2018 (UTC)
    • I think it is, because polychromasia is used four other times in the chapter. Natureium (talk) 19:22, 27 July 2018 (UTC)
• So is it worth having its own article about? Or should it be merged and redirected somewhere like...I dunno. The article on moles, maybe? ♠PMC♠ (talk) 21:11, 28 July 2018 (UTC)

I think merging is a good idea, but I don't know if that's the best topic or what a better topic would be. Natureium (talk) 00:37, 29 July 2018 (UTC)

Follicular phase

Hi guys. I've just come across Follicular phase which is a real mess from a referencing point of view, lots of unreliable source tags and refs stuck in at random. Could someone who knows the area, and with the ability to check out some of the references, take a look? TIA Le Deluge (talk) 09:37, 28 July 2018 (UTC)

• Weschler, Toni (2015). Taking Charge of Your Fertility: The Definitive Guide to Natural Birth Control, Pregnancy Achievement, and Reproductive Health. HarperCollins. ISBN 9780062409911. Retrieved 28 July 2018.(indicated 2002)...removed this source and a few other edits--Ozzie10aaaa (talk) 20:49, 28 July 2018 (UTC)
  • Why? I followed User:Lyrl's work back in the day, and I got the impression that this is one of the best reliable sources on the subject of menstrual cycles, which is all the more important because it is aimed at a lay audience. Why would you just remove it, especially to leave (still accurate) content completely unsourced? WhatamIdoing (talk) 20:40, 29 July 2018 (UTC)

per Wikipedia:Identifying reliable sources (medicine)#Books Medical textbooks published by academic publishers are often excellent secondary sources...however should you be of a different opinion, please revert--Ozzie10aaaa (talk) 21:20, 29 July 2018 (UTC)

Aquarium therapy

I just became aware that this page exists. It should get some scrutiny for health claims and sourcing. --Tryptofish (talk) 17:47, 28 July 2018 (UTC)

• article [3]needs to be assessedhas been assessed, and Zefr[4] was a big help, thank you--Ozzie10aaaa (talk) 18:59, 28 July 2018 (UTC)
  • The issue has been resolved via a local consensus to merge and redirect. Thanks everyone who responded. --Tryptofish (talk) 23:31, 28 July 2018 (UTC)

Hindawi

Borderline predatory, and now this...Please use cautiously if at all... Jytdog (talk) 22:52, 28 July 2018 (UTC)

Only borderline? Natureium (talk) 00:35, 29 July 2018 (UTC)

The Lancet published Wakefield, which is a very similar story. The problem in part is that journals want to publish controversial stuff. Doc James (talk · contribs · email) 20:19, 29 July 2018 (UTC)

Osteopathic medicine redirects

There are several redirects to Comparison of MD and DO in the United States which I think probably ought to be deleted or retargeted, on the basis that we ought to avoid having redirects from general, geographically-unspecific terms to U.S.-centric articles. I've retargeted Allopathic and osteopathic, Allopathic osteopathic and Osteopathic allopathic to alternative medicine, which strikes me as the best target for those titles, and suggested the same course of action for Allopathic and osteopathic medicine, which is currently at RfD. There are several others I'm unsure of what to do with with though:

• Comparison of allopathic and osteopathic continuing medical education
• Comparison of allopathic and osteopathic medical schools
• Comparison of allopathic and osteopathic medicine
• Comparison of allopathic and osteopathic physicians
• Comparison of orthodox and osteopathic medicine
• Comparison of osteopathic and allopathic medicine
• Differences between allopathic and osteopathic medicine
• Differences between osteopathic and allopathic medicine

Each of these has the double problem of suggesting to the reader that the target contains a comparison or list of differences (making alternative medicine an unsuitable target), and pointing from a nonspecific phrase to a specifically American article. There is also Separation in medicine, which could refer to any number of things. I'm asking here rather than (immediately) taking these to RfD because (1) I'm aware that that alternative medicine-related articles have been the source of much antagonism in the past; and (2) it's a topic about which I'm utterly clueless – what, for example, is the difference between osteopathy and osteopathic medicine? As such I wouldn't feel entirely comfortable intervening when there could be any number of complexities I don't know about. Any thoughts would be much appreciated. – Arms & Hearts (talk) 22:46, 29 July 2018 (UTC)

I think either they should be kept as redirects to the Comparison of MD and DO in the United States because they are titled "comparison" or deleted because there's no good article, and these seem unlikely search terms and article links. Natureium (talk) 23:49, 29 July 2018 (UTC)

Intermittent fasting reviewing

I have just updated Intermittent fasting with up-to-date info and reviews. The changes are substantial and I was told this is a hot topic, so it would be very helpful if someone knowledgeable on the topic could review it :-) Thank you very much in advance! --Signimu (talk) 00:51, 30 July 2018 (UTC)

UK Older People’s Quality of Life questionnaire (OPQOL)

Is anyone familiar with this questionnaire from the UK? Here's another link from a provider of questionnaires I didn't see it on List of patient-reported quality of life surveys, does it make sense to add it? I wasn't sure if it's a widely used survey. -Furicorn (talk) 02:29, 30 July 2018 (UTC)

It's hard to find good secondary sources, but checks of Google Scholar and Pubmed show that it has been used in a substantial number of peer-reviewed publications, so I would say that it belongs in the list (the OPQOL in general, not necessarily the UK-specific form). Looie496 (talk) 14:32, 30 July 2018 (UTC)

Leptomeningeal cancer, Neoplastic meningitis and Meningeal carcinomatosis

Meningeal carcinomatosis: tumor cell clusters in the subarachnoid space in a brain biopsy

Should these be one, two, or three articles?

I was led by a comment on a talkpage to add "Leptomeningeal disease" to the lead of Leptomeningeal cancer (it's used in the article text and some of the sources), and to create a redirect from it and a dab page entry at LMD. I then realised that the other listed synonyms of Leptomeningeal cancer didn't have redirects, so set about creating them... and found that Neoplastic meningitis and Meningeal carcinomatosis existed as separate articles, with leptomeningeal carcinomatosis redirecting to the former. So three of the terms listed in the lead of the LC article are covered elsewhere.

I'm not a medic, just a wikignome. Could someone knowledgeable in the field please have a look at these articles and redirects? (Oh and carcinomatous meningitis redirects to Neoplastic meningitis, though not mentioned there as a synonym, although this source gives it as a synonym of Leptomeningeal disease.) PamD 08:16, 30 July 2018 (UTC)

More info: all three articles have existed since 2007 and earlier, though Leptomeningeal cancer was a single-sentence stub until massively expanded in April this year by an editor in her first and only two edits. Educational assignment perhaps? PamD 08:31, 30 July 2018 (UTC)

should remain as separate articles...IMO--Ozzie10aaaa (talk) 10:39, 30 July 2018 (UTC)