[[file:Jasmine-flower-clipart2.gif|40 px|left]]the above article could use some edits,[https://en.wikipedia.org/wiki/User_talk:Kevin_Gorman]..... thank you--[[User:Ozzie10aaaa|Ozzie10aaaa]] ([[User talk:Ozzie10aaaa|talk]]) 00:51, 30 July 2016 (UTC)
[[file:Jasmine-flower-clipart2.gif|40 px|left]]the above article could use some edits,[https://en.wikipedia.org/wiki/User_talk:Kevin_Gorman]..... thank you--[[User:Ozzie10aaaa|Ozzie10aaaa]] ([[User talk:Ozzie10aaaa|talk]]) 00:51, 30 July 2016 (UTC)
{{-}}
:"Could use some edits" is not helpful. What specifically are you requesting? ‑ [[User:Iridescent|Iridescent]] 12:10, 30 July 2016 (UTC)
:"Could use some edits" is not helpful. What specifically are you requesting? ‑ [[User:Iridescent|Iridescent]] 12:10, 30 July 2016 (UTC)
::Any improvements to the article. It's something we do here at WPMED: if someone we know dies, then we try to improve an article related to that person. Since this editor was an organ donor, articles on organ donation are a good match.
::I've worked on those articles in the past, although it's been a long time since I looked at them. Most of them have a lot of information about the sociolegal issues of convincing people to be donors. It might be interesting to see whether we could turn up some sources about how the surgeries happen, and expand the articles with that technical information. It sounds like his heart was transplanted, so perhaps that's an area to focus on. [[User:WhatamIdoing|WhatamIdoing]] ([[User talk:WhatamIdoing|talk]]) 12:47, 30 July 2016 (UTC)
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.
Together with Wikipedia:WikiProject X we are a bunch of editors who have been redesigning the WikiProject Medicine page, in part because it was a jumble of things built up over 10 years. We've tried to strip out the essentials, while leaving anything that still is useful — trying to make the most often used parts more accessible.
One major reason why this was done is to make it easier for new editors to understand what our project is and to make it easier to join. The beta is available at Wikipedia:WikiProject Medicine/sandbox and includes new links and new tools as well as a rework of the old.
Before this is pushed live, I want to know if you have any questions or comments or any features you feel are missing. Remember, this page should welcome new users, and it is possible to keep all the older infrastructure on another page, so that seasoned editors can find what they need to! Carl Fredrik 💌📧08:33, 24 June 2016 (UTC)[reply]
Perhaps its me, but I want to know where the content box for the various topics under discussion is going to go? Somewhere near the top, I hope? -Roxy the dog™woof12:56, 24 June 2016 (UTC)[reply]
Looks good, although I'm not too enthusiastic about the banner. I don't think most of the images will be meaningful to most people - either they require specialist knowledge to identify/appreciate (why show a keratometer instead of something immediately recognizable like a microscope?), or their primary contribution appears to be aesthetics (what's added by a picture of some syringes?) I also don't think it's representative of the full range of medical topics - I think it's overemphasizing hospital and laboratory work, and underemphasizing the social aspects of medicine among other things. There's nothing immediately recognizable as relating to infectious disease, nutrition, clinical research, and so forth. That said, some of these things probably don't have any good pictures available, so it might be difficult to address this. Sunrise(talk)13:21, 24 June 2016 (UTC)[reply]
This is a first draft for a banner, and I am willing to redesign it per the request of users. Sunrise — throw me some good free images that I can include that you think are better. Also remember, there are more images if you expand the banner up to a width resolution of 2560px. Carl Fredrik 💌📧14:06, 24 June 2016 (UTC)[reply]
JakobSteenberg — This will be solved before the page goes live — the problem is that list is auto-generated by number of overlapping articles. So I chose to add the section on the side with the images — currently I haven't finished how to overlay the text and link on top the image — but I will. The same is true for WP:NEURO, WP:VET, WP:ANIMALANAT. Carl Fredrik 💌📧09:52, 25 June 2016 (UTC)[reply]
First thought: Didn't we just redesign the project pages a year or two ago?
Second thought: I think you need to spend more time looking at it using different combinations of screen size and font size. For example, you're using {{col-begin}} to create fixed column numbers, and that means that it's hard to read on narrow screens. Try zooming in about six times and see what that looks like. WhatamIdoing (talk) 01:30, 25 June 2016 (UTC)[reply]
1.) No, we really didn't we added lots of stuff, but it really isn't more useful. I don't know anyone that actually uses the page, and for new users the page is actually absolutely horrendous — we're scaring away new potential contributors — all we managed to do was bloat the page.
2.) The site currently doesn't work well in the mobile viewer, I'm thinking about how to solve that, but it is a bit difficult. But this is true for the page as is. Carl Fredrik 💌📧09:55, 25 June 2016 (UTC)[reply]
On behalf of WikiProject X, I would like to thank Carl Fredrik for his leadership in the re-design. Please let me know if you have any questions regarding the layout or any of the WikiProject X tools, most of which are on display at Wikipedia:WikiProject Medicine/Tools. As far as being viewable on mobile goes, we are currently developing a MediaWiki extension that will replace the template-based approach of WikiProject X with a proper WikiProject interface; this will allow the WikiProject layout to adapt to mobile interfaces. Harej (talk) 18:55, 30 June 2016 (UTC)[reply]
I stumbled upon Epigenetics of neurodegenerative diseases. Not sure what to do with it. A lot of work which unfortunately seems to duplicate information placed under individual disorders. I am tempted to AfD it but would appreciate if someone more knowledgeable could assess its value. — kashmiriTALK20:51, 30 June 2016 (UTC)[reply]
The broad question is whether the topic itself needs a separate article. At least as regards spinal muscular atrophy, the part on therapies is based mostly on primary sources, outdated, unjustly selective, and simply misleading, and I'd rather see it go. I see no point to duplicate the contents of spinal muscular atrophy. I suspect the same might be as regards other disorders listed in that student work. — kashmiriTALK10:02, 1 July 2016 (UTC)[reply]
I had a look at it, and it needs a lot of work. It does feel like a student wrote a paper and decided to post it on Wikipedia. However, I think it's a valid topic and should be kept. (If not, then a bold WP:MERGE is a better choice than AFD.)
Picking up sections from other articles is actually a Good Thing™, as it provides specific examples and makes it possible for the reader to get all the information in one place. I'm inclined to keep that feature, even though the specific contents of most sections need to be replaced. WhatamIdoing (talk) 15:29, 1 July 2016 (UTC)[reply]
So, I've gone ahead and updated the WP:MED landing page. It should be easier to navigate, and include less text than the earlier versions (hopefully scaring away fewer new editors).
I'm interested in hearing about any concerns and bugs you may find. The page is an extensive and specialized redesign of the Wikipedia:WikiProject:X-interface - which means it will see improved mobile support in the future.
I'm also interested in hearing about the performance of the page, does it load quicker, slower, or no difference? Do the images load slowly?
Anyway, the Research and News sections need some updating, but I've left that for everyone to pitch in, it should be straight-forward enough for anyone to change - ask me or Harej if you have any questions about editing in the modules!
P.S. You may need to sign up again, we will be rolling over all the previous sign-ups, but if you want to use any of the new functionality with images and telling a little bit about yourself you should go ahead and do it within the coming weeks. Best, Carl Fredrik 💌📧14:01, 1 July 2016 (UTC)[reply]
This sounds like a good time to archive the names of inactive participants. (Also, is anyone actually using that list? If we're not using that list, do we need to keep it?) WhatamIdoing (talk) 15:16, 1 July 2016 (UTC)[reply]
The new system automatically archives inactive users after 2 months (it also tags them as active if they start editing again). The lists are good for mass-messaging, but one of the problems is that members of our project edit on different Wikipedias, and some might be active elsewhere. I'll see if Harej has any input on this. Carl Fredrik 💌📧19:16, 1 July 2016 (UTC)[reply]
You have license-related problems with the icons. Unless it's public domain, then there must be a link to the File: page. By default (but not strictly as a requirement), when you click on the image, it takes you to the File: page for the image. If you want to change that behavior, then you have to provide an alternative method of accessing the File: page. WhatamIdoing (talk) 05:48, 2 July 2016 (UTC)[reply]
Yeah, they should be PD, or PD-shapes (letters arent covered by US copyright). The only one that is sort of iffy might be the events one, it is CC-BY and a little complex so could do with a replacement. I will see if I find a replacement, otherwise I will link it at the bottom of the page.Carl Fredrik 💌📧08:00, 2 July 2016 (UTC)[reply]
We seem to be missing {{:Wikipedia:WikiProject Medicine/Article alerts}} on the new page. I.e., this:
Seppi333, the reason this isn't on the main page is that we need to welcome new readers. Very few would actually use the page, and for all these types of uses there is the new WP:MED/Tools which can be expanded for seasoned editors. It also includes a number of new tools that we haven't had before, and having all of them on the main page would have made it slow and somewhat scary for new users. Carl Fredrik 💌📧15:17, 2 July 2016 (UTC)[reply]
Thanks for your work on this page. Having checked under Tools-Tasks or -More, then under Research and Resources, I was looking for a hierarchy of how we as editors could apply criteria for upholding MEDRS in a checklist-way similar to the pyramids of WP:MEDASSESS. Some factors include 1) Cochrane review, 2) meta-analysis, 3) topic review of significant clinical trials, 4) journal quality/impact factor, 5) dealing with alternative medicine, 6) separating primary research (usually rejected) from acceptable in vitro pharmacology, 7) assessing reports from medical organizations, news, etc. Basically, this would be a simplified checklist of descending priorities for MEDRS quality, with gadget(s) where useful such as for reporting an impact factor. Would be interested in your thoughts and solutions for such a checklist. --Zefr (talk) 15:08, 2 July 2016 (UTC)[reply]
Zefr -- The main reason why this page was redone was to make it easier for new editors, and useful for older ones. I think your idea would be very useful for the main page - as long as it is simple and straight-forward. I've been thinking about different guides that could help new users, and as I see it the major problem is when editors:
Search for sources
Most editors don't know where to look for information, or which sources have the best information. The reason I haven't worked on this yet is that different topics have different requirements on level of evidence - so it might be difficult to draft any general guidelines. Carl Fredrik 💌📧15:36, 2 July 2016 (UTC)[reply]
Draft hierarchy for a MEDRS credibility checklist: Descending order of quality; the higher the ranking, the better the quality, with an overall goal of summarizing scientific consensus.
Sources below item 11 are not sufficient for MEDRS.
Alright, I'll take a bite out of the bait you are offering. I am currently a nursing student and have lots of medical textbooks I am using and studying. I have a good bit of information in my medical textbook about alternative therapies - what then? The medical textbook is meticulously edited and includes references to systematic reviews and evidence-based medicine: "Review: music really does decrease the time in recovering from surgery." Opinions? Best Regards,
This was a widely publicized literature review and meta-analysis discussed in the public media last year. It was conducted at a reputable institution and published in a high-impact, peer-reviewed journal, The Lancet, then in your textbook. It is probably safe to say that music in the operating or recovery room (perhaps for the staff, but not for the patient) is not in common practice. Yet there are some alternative medical practices, like meditation to lower heart rate and blood pressure, that have shown positive effects, although generally have not been studied rigorously by RCTs as would apply to a new drug. WP and MEDRS sources are not intended to be about everything, WP:NOTEVERYTHING, WP:NOTTEXTBOOK. Thanks and good luck! --Zefr (talk) 00:48, 5 July 2016 (UTC)[reply]
Your proposed hierarchy isn't consistent with your own editing history Zefr. You've removed MEDRS-quality reviews and systematic reviews of RCTs from medical articles before. Moreover, #8–#11 are not consistent with MEDRS. Seppi333 (Insert 2¢) 00:55, 5 July 2016 (UTC)[reply]
If you haven't removed a good source before, then you're probably not editing enough. :-)
Otherwise excellent sources should be removed under many circumstances, e.g., misrepresentation of the source, needless/duplicate citations, having even better sources at hand, and UNDUE emphasis on a source. WhatamIdoing (talk) 05:59, 5 July 2016 (UTC)[reply]
IMO a decision tree is the wrong model. The "picture" is more like WP:CANVAS: you always want more 'good' than 'bad'. If you can get all 'good', then that's the best practice. However, if the best possible sources (*cough* rare diseases *cough*) aren't "perfect", then you should use the best sources that you can find. You shouldn't say, "Gee, there's no high-quality scientific evidence about ribose-5-phosphate isomerase deficiency (prevalence: exactly one patient), so I guess that article should be blank". You should instead say, "What's the best I can do for the article – and for any statement that is already present, is the source adequate (NB: adequate, not perfect, as in "enemy of the good")?" WhatamIdoing (talk) 06:04, 5 July 2016 (UTC)[reply]
* First of all: is it a systematic review/meta analysis, or did it just choose to include a new hip phrase in its title? Does it explain its search methodology; inclusion and exclusion criteria; and are these valid? Is it free from overt bias? If the answer to any of these questions is no, then it isn't a good source. For more information see guidelines such as AMSTAR, GRADE, PRISMA, QUORUM etc. † A good question is one that can actually be answered. A bad or unanswerable question can for example be "is treatment X good?". Good is a subjective measure, and can't be answered by these methods. A bad question can also try to explain multivariable effects with single-variable causes; or it could be unreasonably large or broad - such as "what are possible long-term and short-term health effects and economic effects of condition X, and what are their causes" the answer may be spreading itself so thin as to be unusable (very large and rigorous health technology assessments are an exception - but run into costs of millions of dollars, take several years and put very high demands on those who perform them to be unbiased). ‡ Otherwise reputable sources may be ill-fitted for writing encyclopedic content about health. Cohort studies, case reports, expert opinion, blog posts or company sites are just some examples of sources that may have legitimate uses, but don't belong on Wikipedia per WP:MEDASSESS. Additionally sources that seem to abide by these criteria but are blacklisted should not be used — the blacklist includes journals and publishers which are known for accepting fraudulent or self-published sources. A similar concept is predatory publishing.
I threw this together in a few minutes, and I'd like some of your thoughts? Keep in mind it is a draft, and I created it from the top of my head - that is I didn't take into account any of your suggestions. That means it should not go out like this, so please come with suggestions! The *asterisk* is meant to point to some external guidelines for assessing the quality of a systematic review (which seemed to in-depth for this chart), but as I see it the biggest problem we've had is that things that call themselves systematic reviews aren't. In case you're getting a black box at the bottom right it is supposed to read "MEDRS threshold". Carl Fredrik💌📧10:55, 15 July 2016 (UTC)[reply]
Thanks, CF. I like this decision organization which serves dually a quick-view of criteria for experienced editors meanwhile guiding less experienced editors how to screen sources. Under the top criteria, however, we need another qualification filter (personally, I would employ journal impact factor plus scan "cutoff" disqualifiers and "blacklist" items) to further test credibility of sources contributing to the second-level blue box decisions. For completeness of the "Other" branch, perhaps there could be a link to synthesize MEDRS "disqualifiers" like those under "Cutoff" above and the Blacklist below. Will keep pondering... --Zefr (talk) 13:52, 15 July 2016 (UTC)[reply]
Zefr, I added some clarification about what other means. It makes the image far less general, and makes the scale from low to high only concern MEDRS, not real usefulness (conflating cohort studies and company sites, which isn't really right). Carl Fredrik💌📧14:16, 15 July 2016 (UTC)[reply]
To keep the decision diagram as clean and precise as possible, perhaps link the "other" disqualifiers to an external page. --Zefr (talk) 14:42, 15 July 2016 (UTC)[reply]
Okay, yeah that might be a good idea. Also, before you interject WhatamIdoing, this only means to cover studies, not other types of sources. I'm inclined to consider that if a rare disease: only has ever been mentioned in case reports or non-review papers; not brought up in Orphanet or any other reputable source - then it shouldn't be on Wikipedia. Anyone can make up a rare disease, and that doesn't mean it actually exists or is notable. (Also note that apart from you I have done quite a lot of work and advocacy for rare diseases on Wikipedia, so please don't see my position as anti-rare disease-coverage. Carl Fredrik 💌📧15:02, 15 July 2016 (UTC)[reply]
Anyone can "make up" a rare disease, but not just anyone can get that rare disease accepted by the medical community. You can get recognition (e.g., from Orphanet) even when there are no papers that you are counting as "high-quality". WhatamIdoing (talk) 21:29, 15 July 2016 (UTC)[reply]
This paper[1] and comment[2] would seem highly relevant. The paper, while intentionally provocative, is a thoughtful appraisal of the shortcomings of primary clinical studies. Boghog (talk) 18:13, 15 July 2016 (UTC)[reply]
References
^Ioannidis, John P. A. (21 June 2016). "Why Most Clinical Research Is Not Useful". PLOS Medicine. 13 (6): e1002049. doi:10.1371/journal.pmed.1002049. {{cite journal}}: Unknown parameter |name-list-format= ignored (|name-list-style= suggested) (help)CS1 maint: unflagged free DOI (link)
Carl, I do appreciate the intention behind this. However, it's just not workable in practice. Here are some things for you to think about:
A source is only WP:Reliable insofar as that source is compared with a specific bit of material on Wikipedia.
Example: PMID 24610050 (Cochrane review) says that menstrual suppression via any of the three named methods is helpful for endometriosis. Your chart suggests that this source – assuming that I personally judge it to be "sound" and to "ask an answerable question" – is a high-quality source, full stop. Here's what I want you to remember when you write this: This source, which you've declared to be a 'high-quality' source, is absolutely, unacceptably, impossibly low-quality for the 99.9999% of other health-related statements in Wikipedia.
On the other side, PMID 22728166 (mid-size double-blind randomized, placebo-controlled trial) says that taking Vitamin C and/or Vitamin E significantly reduces self-reported pain from endometriosis. Your chart suggests that this source is low quality, full stop. But this source would actually be an acceptable quality source for a passing comment that says a 2013 study found that these vitamins reduced self-reported pain. Such a statement might not be WP:DUE (whether it's due would depend upon other factors, including the subject of the article and whether the conclusion is contradicted by other sources), but the source is reliable for that statement.
When you're looking for sources, you want to consider what's best (e.g., along the lines that you recommend here, at least within the practical limit of being unable to use sources that you can't get your hands on). But when you're evaluating an existing source, you should consider what's acceptable (i.e., you should accept a source that is less-than-best but which meets the minimum standards for the statement).
Editors should not be trying to determine whether an apparently reliable is "sound". Really. 99% of editors are simply incapable of doing this. What happens in practice is that editors decide whether the approve of the POV. We have seen this happen in hundreds of disputes. People make unsound claims that a paper is unsound (especially around toxicology and environmental issues) basically because they don't like the conclusions from a paper, and therefore the paper must be "wrong".
Editors should not reject sources because they're "biased". In practice, "bias" means "I have personal objections to this paper's funding source/main subject for the journal/author's profession/inclusion criteria/etc." This is exactly what editors must not do, right? We must not reject papers solely because the research was sponsored by the drug manufacturer, or because the author is a surgeon, or because I'm mad that the study counted fruit snacks in the same category as gummi bears.
Basically, IMO this isn't going to help.
What would probably help is starting with the basic issues of RSCONTEXT and about how to figure out what's DUE.
If you want something that's simple enough for newbies, then tell them to use textbooks. Or give them a link to PubMed that has pre-filled limitations of not only "reviews" and "last five years", but also "core clinical journals", and then tell them that if there's nothing available in either textbooks or in the core clinical journals, then they can ask us for help in finding the best sources that exist. WhatamIdoing (talk) 19:17, 15 July 2016 (UTC)[reply]
@WhatamIdoing: summarizes points in the last 2 paragraphs that can add filters for the middle blue boxes of CF's decision tree diagram: WP:RSCONTEXT, WP:DUE and WP:MEDASSESS qualifiers should include reference to PubMed recent reviews and core clinical journals with high impact factors. I suggest revision of the middle blue boxes with attention to these points which do then provide a useful guide for triaging MEDRS quality. It won't be perfect and all-inclusive, but it can be effective. --Zefr (talk) 20:33, 15 July 2016 (UTC)[reply]
I'm also unhappy about labeling any single quality as "MEDRS" (as in "MEDRS threshold" in this drawing), or even as "high-quality". So if it's a meta-analysis that I think is sound, then it's "MEDRS"? What if it's 50 years old, by a discredited author, and in a junky journal?
And is a narrative review actually "lower quality" than a meta-anaylsis? Well, that depends entirely on exactly what statement you're trying to support. Meta-analyses are excellent sources for scientific evidence about things that can be usefully measured with statistics (e.g., treatment efficacy or prevalence of side effects). If you're writing about anything else (e.g., patient acceptance of treatments or alternate names), then you're better off with that allegedly "lower quality" narrative review.
Here's a simple example: Pouring chemo drugs straight into the abdomen increases survival for ovarian cancer. The best possible source (NB: not "only acceptable source") for that is going to be a meta-analysis. But a lot of women refuse this outright or quit soon after it's started. The best possible source for a statement that "most women quit this treatment because it's brutal" is not a meta-analysis. The best possible source for what is, after all, a deliberate and purely individual decision to value quality of life now over a small increase in long-term survival is a narrative review or a book.
This chart isn't really about "MEDRS". It's about how to evaluate objective scientific evidence, for those times when you are writing about scientific evidence. It's MEDASSESS, ignoring the other five major criteria that MEDRS lists under == Basic advice ==. Seen in that limited light, it's not a bad explanation of how to apply the second paragraph of MEDASSESS. But there's so much more to MEDRS than MEDASSESS. WhatamIdoing (talk) 21:29, 15 July 2016 (UTC)[reply]
Let's keep in mind that CF's chart and the goal of this exercise should be to employ a quick visual guide and checklist for supporting statements, like a literature triage that will often require additional review and source substitution by other editors. The MEDRS threshold gives a triage tag as a decision point for whether the source is used or rejected. Keep going, CF. --Zefr (talk) 22:07, 15 July 2016 (UTC)[reply]
That's the problem: you've just "rejected" sources that talk about anything that isn't based on pure statistical analysis. There simply are no meta-analyses on the One True™ name for a given artery, and this diagram labels the actual best sources for that kind of material as being universally "lower quality" for all content about health. WhatamIdoing (talk) 02:26, 16 July 2016 (UTC)[reply]
Okay, let my try to refute your points WhatamIdoing. I think we're getting at this from different view-points.
This first flowchart was meant to explain evidence quality, not be a general overview of MEDRS (something which I agree isn't feasible to put in a flow-chart). Relevance is related to quality, but that wasn't what I was getting at. A source is most assuredly not "high-quality" for all statements, it needs to be relevant, and everything written in a meta-analysis is not high quality per se, if for example the background section cites a primary source without employing meta-analytical methods then that isn't meta-analysis quality for the statement. I've tried to clarify this, and will post a second draft below.
I took your suggestion of using WP:CANVAS, and the last section in my new draft addressees your point.
That 99% of editors are incapable doesn't mean they aren't going to do it anyway. Instead of saying that they shouldn't, and then tackling the fact that they do anyway — we can at least give them the tools to do it properly. If we have clear language here it is very simple to refute such "unsound claims of papers being unsound" — by asking which criteria it fails. "Either get me another reliable source that says this is biased, or tell me which criteria it fails on."
That statement sort of undermines the idea that editors can do anything. There is an example, just below this of a source claiming to be a systematic review, where the only thing that is systematic is the lack of any systematic review protocols. I've never come across a case where editors don't try to assess bias in a source when arguing — and if we ignore this we only open the field towards cherry-picking the sources you want, then performing some pseudo-argument for why they are better — or actually more commonly, having no-one argue because too few people edit an article. We need to recognize that bias in sources is an issue, and not something that we can sweep under the rug and pretend we are immune to on Wikipedia.
The best source to use in order to claim that women drop out of such treatment is a meta-analysis, which can say how many they were — and if a survey was simultaneously performed the most common reasons for doing so. That this isn't always done doesn't mean it isn't the best source — but we need to accept the "best possible sources". If the best we can reasonably get is a narrative review then we can accept that. If a systematic review or meta-analysis looks at the question rigorously and comes to a different conclusion then it trumps the narrative review.
To the last comment, which is furthest up — about making up rare diseases: well then all the more reason not to accept the papers themselves when we can use Orphanet which adds an extra layer of analysis on top.
A relevant source is one where the topic is the same as for the statement you use it for. E.g. — if you're writing about methotrexate treatment for rheumatoid arthritis: the best possible source in one covering treatment options in general, or even specifically methotrexate. The second best is a source that is about a different treatment, but states your topic in passing: "DMARDs in RA have benefits over methotrexate, which may cause leukopenia". Less relevant sources may be a paper on psoratic arthritis that mentions methotrexate as a treatment for RA. Entirely unrelated or marginally related sources should never be used.
Quality assessment is performed according to the section on assessing evidence quality, but in essence means that higher quality sources trump lower quality sources — where the lowest quality sources should be entirely avoided.
Medical research can move very quickly, with evidence changing from one year to the next. Major organizations often recommend reviews of the evidence be no older than 2 years in order to be deemed "best evidence" (providing they haven't been trumped by newer authoritative evidence) . Wikipedia employs a cutoff at 5 years, where anything older than that only being used in exceptional cases, such as rare diseases. Sources older than 10 years should not be used (not including certain related fields such as anatomy, where the body of knowledge moves far slower).
Always seek the best source possible. It is unreasonable to expect editors to always use the best sources, but they should always try. Issues finding or using the best sources may be: pay-walls (see the Resource Exchange for help on accessing pay-walled content); difficulties in determining which source is best; or it may just be unreasonably time-consuming to go through all possible sources. When you have several contenders for best possible source which don't agree — make sure to give all relevant viewpoints, remembering to avoid false balance.
Try to avoid bias when selecting sources. Bias can arise from your search criteria and search tools, and in which sources you have access to (see for example FUTON bias).
Proposed flowcharts for MEDASSESS
Old draft
* First of all: is it a systematic review/meta analysis, or did it just choose to include a new hip phrase in its title? Does it explain its search methodology; inclusion and exclusion criteria; and are these valid? Is it free from overt bias? If the answer to any of these questions is no, then it isn't a good source. For more information see guidelines such as AMSTAR, GRADE, PRISMA, QUORUM etc. † A good question is one that can actually be answered. A bad or unanswerable question can for example be "is treatment X good?". Good is a subjective measure, and can't be answered by these methods. A bad question can also try to explain multivariable effects with single-variable causes; or it could be unreasonably large or broad - such as "what are possible long-term and short-term health effects and economic effects of condition X, and what are their causes" the answer may be spreading itself so thin as to be unusable (very large and rigorous health technology assessments are an exception - but run into costs of millions of dollars, take several years and put very high demands on those who perform them to be unbiased). ‡ Otherwise reputable sources may be ill-fitted for writing encyclopedic content about health. Cohort studies, case reports, expert opinion, blog posts or company sites are just some examples of sources that may have legitimate uses, but don't belong on Wikipedia per WP:MEDASSESS. Additionally sources that seem to abide by these criteria but are blacklisted should not be used — the blacklist includes journals and publishers which are known for accepting fraudulent or self-published sources. A similar concept is predatory publishing.
* First of all: is it a systematic review/meta analysis, or did it just choose to include a new hip phrase in its title? A number of methods exist to assess this. † A good question is one that can actually be answered. A bad or unanswerable question is one that can not produce an actionable finding. ‡ Not all sources belong in an encyclopedia: cohort studies, case reports, expert opinion, blog posts, company sites or blacklisted sources are just some examples of sources that may have legitimate uses, but don't belong on Wikipedia.
Systematic reviews and meta-analyses
A systematic review or meta analysis explains its search methodology; its inclusion and exclusion criteria; as deemed valid by a number of protocolls. For more information see guidelines such as AMSTAR, GRADE, PRISMA, QUORUM etc.
A good question is one that can actually be answered. A bad or unanswerable question is one that can not produce an actionable finding. However that a source fails to find strong enough evidence for something doesn't make it bad (see known unknowns). For an example of a bad question: "is treatment X good?". Good is a subjective measure, and can't be answered by these methods. A bad question can also try to explain multivariable effects with single-variable causes; or it could be unreasonably large or broad - such as "what are possible long-term and short-term health effects and economic effects of condition X, and what are their causes" the answer may be spreading itself so thin as to be unusable (very large and rigorous health technology assessments are an exception - but run into costs of millions of dollars, take several years and put very high demands on those who perform them to be unbiased).
All study sources
Sources should be free from overt bias. Claims of bias should always be sourced to secondary reliable sources, according to the same criteria as other sources. That a source has potential bias is not always enough to omit it, if it is especially notable it is proper to include it with a qualifying statement explaining why it is biased (with references).
Not all sources belong on Wikipedia. Otherwise reputable sources may be ill-fitted for writing encyclopedic content about health. Cohort studies, case reports, expert opinion, blog posts or company sites are just some examples of sources that may have legitimate uses, but don't belong on Wikipedia per WP:MEDASSESS. Additionally sources that seem to abide by these criteria but are blacklisted should not be used — the blacklist includes journals and publishers which are known for accepting fraudulent or self-published sources. A similar concept is predatory publishing.
Books can often be a great source when starting an article, giving a good overview of the subject.
High quality text-books can be a good source to start an article, and often include general overviews of a field or subject. However, books generally move slower than journal sources, and are at an average 7 years behind the current state of art.[1] This makes using up-to-date books even more important, and they should preferably only be used for less controversial statements. A high quality journal article trumps most book sources.
Guidelines are important on Wikipedia because they present recommended practice.
Guidelines are important on Wikipedia because they present recommended practice. They're not necessarily the best sources for statements of fact, but are nearly always relevant when they convey information of recommended treatment and clinical practice. These sources may be biased, but individual editors should not try to assess this themselves. If reliable sources express concerns of bias this can be included in an article, but never omit a source or add extra qualifiers to a statement based on original claims or low quality sources claiming bias.
Guidelines sometimes clash with one another, for example: the World Health Organization vs. American Heart Associations on salt intake. In such cases it is important to present both sides, and to add information referencing the underlying science. Guidelines aren't always based on sound science, for example the WHO position on breastfeeding in the developed world — but this is never a reason to omit a guideline! Instead reference the scientific literature explaining how it may differ from the guidelines — remembering to avoid WP:original research. Unless the source itself critiques the guideline you should avoid words and phrases such as "however", "this is not supported by", etc.
*Health technology assessments or HTAs are the gold standard when it comes to assessing evidence quality and take into account various aspects such as effect, risks, economic costs, and ethical concerns. They seldom make recommendations, but instead explain most effective treatments, potential hazards and sparse knowledge. Their name is somewhat of a misnomer as they do not need to concern "technology" as perceived by the public.
Comments
as I said before it seems a good idea to me, and these charts are even better( I think (hope) some of WAID's concerns might be taken care of now...?), great effort/work Carl!--Ozzie10aaaa (talk) 12:13, 16 July 2016 (UTC)[reply]
Progressing nicely, CF! In the "referencing a guideline" chart, I suggest a better term than "health technology assessment" and request clarity. To some, a "technology" means a pre-approved drug, device or diagnostic, as for biotechnology. A note about sources mentioned in WP:MEDORG: all of these science-based bodies (appearing to be "national") have international policy influence, and so may be legitimately called the highest quality of reference guideline to source. Included among them should be EFSAscientific opinions which affect food and nutrient labeling and marketing in the EU. Because EFSA scientific reviews are so thorough, there is cross-referencing among international organizations, such as for the supposed health claims of probiotics. --Zefr (talk) 15:36, 16 July 2016 (UTC)[reply]
Hmm, the problem is that they are called "Health technology assessments" — but I agree it is a horribly confusing name, I just can't really think of a better term. The alternative I can think of is to use HTA and to clarify in text — but being rare few editors will come across these and it feels sort of unnecessary, assuming that such as "dry" section will be read at all, and not only bore readers enough so they don't read the rest.
An alternative to international is multi-national — but this is also problematic because the FDA, EMA, or CDC pull a lot of weight internationally, despite being either EU or US-agencies. But these are sort of one offs, and the ECDC or CCDC might on the surface seem to be similar in size and scope, but just don't pull the same weight. So I'm afraid we'll just end up making things more confusing rather than less. I'll await some more comments before making any changes there. Carl Fredrik 💌📧16:06, 16 July 2016 (UTC)[reply]
I'm very uncomfortable with the questions "Does it explain its search methodology; inclusion and exclusion criteria; and are these valid? Is it free from overt bias?" Who are we expecting to make the judgement on validity and freedom from overt bias? If the answer is Wikipedia editors (i.e. by local consensus on the article talk page), then we might as well hand over all of the CAM, fringe and controversial topics to the POV-pushers, because they will be more than happy to do that sort of analysis in order to reject sources supporting mainstream views and insert sources supporting the fringe view. Just look back at the talk page archives of Abortion, Acupuncture, Ayurveda for examples, and that's just the 'A's - don't get me started on the rest of the alphabet. The talk page for Bloodletting starts with a complaint that the article is too negative; and it doesn't get any better by the time you've reached the article Wobenzym ("the focus of over 200 clinical studies conducted by both the German Wobenzym manufacturer, Mucos Pharma, and third-party researchers." - sourced to Wobenzym.com). I'm sorry to be so negative on those questions, but to keep our encyclopedia honest we need the bulwark of WP:MEDASSESS's "Assessing evidence quality" means editors should determine quality of the type of study. Respect the levels of evidence: Do not reject a high-quality type of study (e.g., a meta-analysis) ... because of personal objections to the inclusion criteria, references, funding sources, or conclusions. --RexxS (talk) 17:02, 16 July 2016 (UTC)[reply]
What makes it work, at least in my mind, is that we here provide links to some real guidelines on how to assess evidence quality. As I read the current guideline, it allows for qualified objections, just not personal objections. I think it would actually be easier to deny such objections if we actually explain what you need to qualify a valid objection. The way it works today is that these discussions arise anyway, and pointing to the line that says no personal objections and the shill just says "well, it isn't personal", — and we have to explain that for objections to be valid they need to abide by certain criteria. This has to be done over and over again.
As I mentioned above to WAID, I think we need to take the fact that there is bias by the horns — and instead of shying away from it we should explain what it is — and what can be deemed legitimate bias. There are great guidelines about bias that we can refer to, and these get rid of stupid ideas like stating that you can't cite a paper by a surgeon, in a heart-beat.
I agree that it's a nuisance having to repeatedly tell the shills and "true believers"™ that they aren't qualified to perform their amateur analyses on respectable sources. I also agree that it's inconvenient to have to constantly point out to the vested interests that the WHO is allowed to be biased against banning abortion or that Ernst is actually qualified to write papers on CAM, but I'd rather do that than give them carte blanche to reject those sources on their personal say-so. In my humble opinion, the only qualified criticism of sources falling into the high-quality "type" is from other published, reliable sources independent of the subject. And that's what we need to be saying. --RexxS (talk) 17:47, 16 July 2016 (UTC)[reply]
The diagram that supports the inclusion of large RCTs is at odds with MEDRS – "Primary sources should generally not be used for medical content". RCTs of any size are not MEDRS-quality because they're primary sources.
[TLDR: A larger RCT is not necessarily a better RCT] Moreover, if two RCTs have exactly the same statistical design, except one has a sample size of 1 billion people and the other has a sample size of 100, they'll only differ in the precision of the estimator (e.g., for an arbitrary population parameter with a true value of 4, the associated parameter estimates on these samples could be something like 4.000000000000001 ± .000000000000001 vs 3.95 ± .5 for the larger/smaller study respectively). In other words, all else equal, massive RCTs are not "truer" or more accurate than smaller RCTs - they're just more precise. This is really only notable when an effect size is very small (e.g., a population parameter's true value is very close to the null hypothesis value) or there is a lot of noise in the data (i.e., large standard error for an estimator with a non-zero effect size), since larger sample sizes shrink the standard error of the estimate and therefore reduce the likelihood of failing to achieve statistical significance in these cases. As a statistician, I'm much more concerned about statistical bias in a study since this can effectively invalidate all the findings/statistics-based conclusions in a paper; sources of bias are not always reported and are sometimes not readily apparent/easy to identify. In the vast majority of cases (including all commonly employed statistical models in medicine that I know about - any exceptions would involve a biased yet consistent estimator, which is asymptotically unbiased), a study's sample size has absolutely no effect on the bias of an estimator. If the aforementioned RCTs had different statistical designs, the massive RCT might be biased with a highly misleading but very precise estimator (e.g., suppose that this RCT attempts to estimate the efficacy of an oral antibiotic against a gastrointestinal bacterial pathogen but somehow ends up giving the individuals in the treatment group a capsule containing the pathogen instead of the antibiotic – both the magnitude and direction of the parameter estimator would differ from the true value of the population parameter that is being estimated), while the smaller RCT might be unbiased (i.e., it actually gives the treatment group the antibiotic) with a less precise estimator. Assuming everything else was done correctly and in the same manner in these studies, the RCT with the sample size of 1 billion would be a flaming pile of shit, while the RCT with a sample size of 100 would yield statistically valid results. Seppi333 (Insert 2¢) 17:54, 16 July 2016 (UTC)[reply]
I know it goes against the current interpretation of MEDRS, and its probably the most controversial thing I've got in here, and also the first thing I would get rid of. I still contend that if there is ever to be made an exception for a primary study (as is allowed exceptionally) it would be for a high-quality, large RCT. The things you mention are why I added the little gray "high quality" rather than just "very large". And this is why I'd originally planned on not making the flow-chart too MEDRS specific, but more general when it comes to evidence quality. Now I agree, few editors here are in reality capable of determining if an RCT is high quality or not — and we shouldn't expect them to. So, just to be clear I kept it in for accuracy, rather than actual usefulness on Wikipedia — but I'll get rid of it. Carl Fredrik 💌📧18:05, 16 July 2016 (UTC)[reply]
That's probably best - or at least point it below MEDRS' normal threshold. The reason should be that if there exists a very large high-quality RCT on a particular topic, then it would be extraordinary if it wasn't analysed in a review within a reasonably short period. In general, it's better to wait for the secondary source than to allow the thin end of the wedge with primaries, IMHO. --RexxS (talk) 19:20, 16 July 2016 (UTC)[reply]
Quick reply: Carl, before I finish reading this, can you please find a way to separate the concepts of "evidence quality" and "source quality"? MEDASSESS is about the quality of scientific evidence. It is not the sum total of "source quality". "High-quality evidence" is not a synonym for "high-quality source". This is important because "evidence quality" is irrelevant for much of what an article needs to cover. WhatamIdoing (talk) 19:37, 17 July 2016 (UTC)[reply]
WhatamIdoing — I made this change a couple of days ago. I've never considered MEDASSESS to be about "evidence quality" in a general sense, so I think we agree, I just never assumed it covered that in the first place. Carl Fredrik 💌📧19:59, 17 July 2016 (UTC)[reply]
If you actually made this change a couple of days ago, then why does the column that says "Quality" at the top specifically refer editors to MEDASSESS (i.e., the section of MEDRS that is labeled "Assess evidence quality" and contains two large pyramids that rank the levels of scientific evidence in medical papers), and ignore other sections of MEDRS that also impinge upon source quality, e.g., WP:MEDINDY and WP:MEDANIMAL? WhatamIdoing (talk) 05:45, 18 July 2016 (UTC)[reply]
CF: I'm still feeling reluctant about HTA as the highest level of evidence in the draft chart for guidelines under WP:MEDORG. The HTA article is about "technologies" specifically, such as telemedicine within MAST. A position statement based on systematic reviews as Level I for evidence-based medicine would likely be the highest standard for a clinical practice guideline which would be adopted and held unchanged for 5 years internationally if sufficiently strong. As for the other charts and text you have drafted, what tasks remain needing collaboration and what is the process for implementing the consolidated changes for MEDRS? --Zefr (talk) 14:58, 20 July 2016 (UTC)[reply]
These are great charts. I like them as they are. I am comfortable with content not being perfect, but instead good enough. Because of this I recognize the criticism here as valid but still think these charts are fine to display and use as they are.
About reading the charts - I am a little confused that they read right to left as "most acceptable" to "least acceptable". I would have expected that a chart which was communicating a progression would have the lowest measured values on the left and the highest measured values on the right, but these charts all reverse that. Artistic taste is no barrier to accepting these but if the art is ever considered, my vote would be for "least acceptable" on the left and "most acceptable" on the right. Blue Rasberry (talk)15:30, 26 July 2016 (UTC)[reply]
That's an interesting point, Lane, and it highlights the differences in different people's cognitive processes. I think that from the point of view of categorisation, it is logical to proceed from least to most as we move from left to right - at least for those of us who read habitually from right to left.
However, from an instructional point of view, we want to present desired outcomes first, as those imprint preferentially over ones presented later. When teaching, we always give correct examples first, and only later show wrong examples, if at all.
So for me, I'd prefer to see the best quality sources placed on the left (as now), because those are probably read first and those are the ones we want folks to use wherever possible. When we translate into Arabic (or any other rtl language), I'd recommend mirroring the chart for those same reasons. Incidentally, there's an argument for reducing the distraction of the poorer sources by using less saturated colours and less contrast, and conversely increasing them for the best sources (and/or using a bolder font), so that the eye is drawn preferentially to the part of the chart that we want the reader to start at. --RexxS (talk) 16:02, 26 July 2016 (UTC)[reply]
Is it ok to make medical claims like this without MEDRS as long as it's attributed and not in WP's voice? David Gerard and I didn't necessarily agree, so I said I'd get input from WT:MED. It says:
^Pickersgill, Gina. "Dr Richard Bandler On Healing - A Special Interview - by Gina Pickersgill". NLP Life Training. The Best You Corporation. Archived from the original on 1 March 2012. Retrieved 8 August 2013. GINA: I have seen you demonstrate a technique that some people refer to as Dr. Bandler's Beauty treatment? Please tell us about that. RICHARD [BANDLER]: Basically what happened is that I noticed that when I hypnotically regress people repeatedly they looked younger. So I started first thinking, well isn't there a way to maintain that. I noticed when I hypnotically regressed people to before the age of 5, who currently wore glasses, didn't need them to see. So I started leaving people's eyes young and growing the rest of them up to the present and it would change the prescription of their glasses radically to the point where they could see better. And done enough times, some of them could see without glasses. So I went a little step further, and did a DHE (Design Human Engineering™) treatment where we set up a mechanism in the back of their mind that repeatedly age regresses them hypnotically; when they sleep, when they blink, all kinds of things and in a state of time distortion. And it can take years off the way people look, it also ups their energy level and in some cases the bi product (sic) has been they recovered spontaneously from very serious diseases. Because they were aged regressed to where before the disease started. Now I cannot prove that but I've seen it enough times that I'm impressed with it.
^In a seminar, Bandler & Grinder (1981, p. 166) claimed that a single session of NLP combined with hypnosis could eliminate certain eyesight problems such as myopia and cure the common cold (op.cit., p. 174)...(Also, op.cit., p. 169) Bandler and Grinder believed that, by combining NLP with hypnotic regression, one not only cured a problem, but became amnesic for the fact that it even existed at all. Thus, after a session of "therapy," a smoker denied smoking before, even when family and friends insisted otherwise, becoming unable to account for such evidence as nicotine stains. Grinder, John.; Richard Bandler; Connirae Andreas (ed.) (1981). Trance-Formations: Neuro-Linguistic Programming and the Structure of Hypnosis. Moab, UT: Real People Press. ISBN 0-911226-23-0.
^Bandler, Richard (1993). Time for a Change. Meta Pubns. p. vii. ISBN9780916990282. In single sessions, they can accelerate learning, neutralize phobias, enhance creativity, improve relationships, eliminate allergies, and lead firewalks without roasting toes. NLP achieves the goal of its inception. We have ways to do what only a genius could have done a decade ago.
^Bandler, Richard (2008). What is NLP? (Promotional video). NLP Life. Retrieved 1 June 2013. We can reliably get rid of a phobia in ten minutes -- every single time.
^Bandler, Richard; John (1985). Andreas, Connirae (ed.). Trance-formations. Real People Press. pp. 166–8, 173–4. ISBN0911226222.
^Grinder, John; Bostic St. Clair, Carmen (2001). "Chapter 4: Personal Antecedents of NLP". Whispering In The Wind. J & C Enterprises. ISBN0971722307.
The lead overall makes it clear that this method is bunk, but that line makes me uncomfortable since it clearly can't be supported with MEDRS, because I think it pops out at the reader since a lot of the terms are wikilinked and there are an excessive number of citations, not to mention the lengthy embedded quotes (that were so long I had to collapse it). Maybe the quotes were added to emphasize the lack of evidence, but the problem is, people who believe this stuff don't see that. Thoughts? —PermStrump(talk)19:50, 9 July 2016 (UTC)[reply]
I don't see anything wrong with documenting the founders' ludicrous claims as being things that they claim, given NLP's status as not just a pseudoscience, but an exemplary pseudoscience, is made clear in the article. We can add something noting that these claims are utterly disproven, but the fact that they make these terrible claims is IMO actually important - I think we need to document notable pseudoscientific claims accurately - David Gerard (talk) 22:01, 9 July 2016 (UTC)[reply]
I definitely get that you weren't promoting those claims. :) I think for someone who is prone to believe it though, it could be easy to miss the context in the surrounding paragraphs, because of all of the citations and blue links in that sentence. —PermStrump(talk)23:23, 9 July 2016 (UTC)[reply]
Technically, this is a BLP claim (i.e., "These BLPs actually said something this nonsensical on the record") rather than a medical one.
If you're worried about the appearance of the bluelinks, then they're mostly unnecessary. If you turn "myopia" into "nearsightedness", then you can probably unlink at least half of them now. The links are only useful if people don't know what it is.
These claims could be presented even more extremely ("X has claimed that he can cure all phobias in all people in only 10 minutes"), perhaps even followed by his own words that he "cannot prove" his claims. But rather than trying to balance this single sentence, it would probably make more sense to consider where it is within the article and what the context is. Writing "Proponents say that X, which doesn't work, does something, even though it doesn't" is not good encyclopedic writing style. WhatamIdoing (talk) 05:53, 10 July 2016 (UTC)[reply]
It's a more general question of how to present bogus and dangerous claims in notable pseudosciences. But I don't think straight-up removing the actual details of the claims, as you did, can possibly be the right answer - this stuff is bogus and dangerous, but needs proper coverage, to the level of saying what its bogus and dangerous claims actually are - David Gerard (talk) 05:59, 10 July 2016 (UTC)[reply]
I see a couple of problems here. First these treatment claims are detailed only in the article lede and not in the body - but ledes are meant to summarize bodies. Secondly, and more importantly, the soures are weak primaries (an interview and seminar transcript). How is it WP:DUE for Wikipedia to include this content if it has not been picked up by reasonable secondary sources? Alexbrn (talk) 06:18, 10 July 2016 (UTC)[reply]
That level of detail problems is an artifact of edit wars with advocates, as detailed exhaustively in the talk page (where it should be discussed if you want such edits to stick). I would actually say think about the general problem at the moment, making the article more coherent later - David Gerard (talk) 08:04, 10 July 2016 (UTC)[reply]
The "only in the lead" issue is easy to solve: Change the lead to something like "They claim to cure a variety of psychological and physical conditions" and move the specific names to an appropriate section of the article.
I didn't see any non-self-published sources in my spot-check of the list. On the other hand, "here's what the quacks themselves claim" is a reasonable use of things published by the quacks. In an article like this, a proponent should see a description of what seems important to them, too, not just the things that seem important to critics. WhatamIdoing (talk) 16:25, 10 July 2016 (UTC)[reply]
@Dodger67: "Cell stretching" is a sort of loose term for inducing mechanical stress on cells in culture to study the effects - you can buy lab equipment to do this - but the draft is written about a treatment applied to real people, not a lab experiment. I don't know if this is notable quackery, but it is quackery for sure, and the draft as written is unacceptable and full of nonsense. (My favorite: "...as it takes less than 6 weeks for this collagen to get produced, its four levels of structuration aren't properly organized".) The sources that are legitimate are describing biochemical/biomechanical effects in a research context and are not talking about treatments or therapies at all. The sources that do relate to this "treatment" are garbage. Opabinia regalis (talk) 22:42, 26 July 2016 (UTC)[reply]
Not sure what to make of this article--it's over 51k bytes and was just created in a single edit by a new user. Other eyes would be appreciated. Everymorning(talk)19:20, 19 July 2016 (UTC)[reply]
So much of these articles are misleading like "The blood of a healthy newborn is always sterile than the adult blood because of the presence of a P-glycoprotein functional barrier between the mother and the fetus in the placenta, also known as the blood-placental barrier"[5]
I guess the word is "healthy". No different than a "healthy" adults blood though as soon as an adults blood is infectious it is no longer healthy.
What we appear to have is boosterism / spamming. Doc James (talk · contribs · email) 23:22, 19 July 2016 (UTC)[reply]
While I think your actions are well-justified in this instance, I disagree with your statement that "as soon as an adults blood is infectious it is no longer healthy" if by "infectious" you mean that it contains bacteria or viruses that could be transmitted from one person to another. Bacteremia is fairly common (e.g. this review), and viremia is even more common (about half of healthy adults chronically carry at least one anellovirus e.g. PMID 26959653). — soupvector (talk) 03:48, 20 July 2016 (UTC)[reply]
For example, on the page you linked it's hard to see the "contact" link because it's surrounded by advertisements. Unpleasant to say the least. — soupvector (talk) 21:49, 20 July 2016 (UTC)[reply]
Basically it is a proprietary wrapper for a google CSE, with extra ads. I'm not too thrilled about these types of search engines when they don't even give any information on how they choose to include or exclude different sites. Carl Fredrik 💌📧22:03, 20 July 2016 (UTC)[reply]
I would appreciate your thoughts on import text from these publications and external sources more generally into medical articles on Wikipedia and how I could improve the process and documentation I have made. In addition ways to make it easier for people to use the UNESCO publications which are not available under an open license (either older publications or co-published works) as references.
Thanks @Doc James:, do you you see any reasons why text from open license UNESCO publications couldn't be copied into Wikipedia articles? I've created a workflow to do it, I just want to understand if there are any other barriers other than the workflow and the suitability of the text. --John Cummings (talk) 10:15, 22 July 2016 (UTC)[reply]
There are problems with "the suitability of the text" in terms of using long pieces of text directly, not least the considerable length of these pieces. The ones I looked at tended not to have a neutral encyclopedic voice, but were advocating specific policies/actions, not doubt very sensibly, but the phrasing would need either quoting as the UNESCO view, or adjusting. Otherwise the source obviously needs to be clear, and where specific medical statements are made (rather than those on the social and economic aspects of public health issues), WP:MEDRS would probably prefer other sources, perhaps going directly to the work UNESCO cites. So I think there is a role, certainly as refs, or in quotations, but also some incorporating text directly. Hope that helps. Johnbod (talk) 11:25, 22 July 2016 (UTC)[reply]
Thanks very much @Johnbod:, this is really helpful, I've tried to articulate this for a little while but haven't been able to, that open license publications can be used as
A source of openly license text that can be directly copied into Wikipedia articles
A reference for Wikipedia articles
A source of references of other works that the UNESCO publications cite
Pretty much. There is a dislike, overdone imo, for quotations in medical articles, but sometimes (without overdoing things) this is the best way. Johnbod (talk) 13:23, 22 July 2016 (UTC)[reply]
I think it is very much down to encyclopedic voice, and from my experience this can vary much between publications. It's also something that is rather simple to rework, sometimes you just have to fix some of the most egregious examples, and then someone else will come along and help with the rest. When it comes to giving advice that is a problem, but as long as any editor is aware that this needs to be qualified as coming from UNESCO or just needs changing from "you should" -> "it is recommended" I think that it is a lesser problem. For an article that was nearly fully based upon CC-BY sources that were rewritten see Heart. If we're looking at outside Wikipedia uses, the materials can of course be reused by other UN bodies, or national bodies. Are the PDFs available on Commons under any specific category? Carl Fredrik 💌📧14:58, 22 July 2016 (UTC)[reply]
Also, a note. If anyone chooses to copy long paragraphs of UNESCO material, remember to tell them to add a reference at the end of every sentence. Otherwise someone else may stick something in the middle of the paragraph that isn't supported by the UNESCO reference and it may look like it is. (This I've found takes some explaining, and not all Wikipedians get it right either). Carl Fredrik 💌📧15:00, 22 July 2016 (UTC)[reply]
Thanks very much @CFCF: (sorry don't know how make a ping show your name correctly). I think that referencing after every sentence is super important and something that I hadn't thought about before. The way I'm thinking about it is using external text as a start to build upon, its an additional opportunity for external organisations to contribute to Wikimedia without having to learn how to edit directly (which has around 1% retention rates) using content they have either already written or could write and use for Wikipedia and other platforms. I'm wondering if there are any open license medical journals or other medical sources that may have suitable text to trial what I've done and improve upon it. John Cummings (talk) 15:35, 22 July 2016 (UTC)[reply]
There are some, but we'd have to choose properly so as not to introduce too much bias if we only select open source articles. We need review articles, which when it comes to open access are somewhat rarer than primary sources (for example PLoS doesn't publish reviews). I'll get back to you about that, but for now if your volunteers are willing to write about anatomy and physiology there is [8] which is an excellent source (all the images are on commons at commons:Category:CNX_anatomy), it just needs some rewriting for encyclopedic tone. Carl Fredrik 💌📧15:46, 22 July 2016 (UTC)[reply]
Thank you, @Johnbod:, this opening-up of licensing information on UNESCO sources is very encouraging. If there is any way that similar clarity can be brought to other UN bodies' publications it would be very welcome. For WPMed interests, that primarily means the World Health Organization, but there are often media published through the UN News Centre, the FAO, or UNICEF (just to name a few) that could be usefully employed in Wikipedia if compatible licensing permitted. So far, some other UN organizations have applied an incompatible "noncommercial" or "no derivatives" license similar to CC-By-NC-ND without even realizing that clause effectively prevents us from using the work. If we were able to get this idea through to the necessary people, it could be very productive. Even short of liberalizing the license, it is helpful to know clearly whether or not it is compatible.
I have been trying to convince WHO to go to an open license for many years now without much success. Plan to keep trying but would not recommend people hold their breath. Doc James (talk · contribs · email) 23:50, 22 July 2016 (UTC)[reply]
Oh man, that first sentence, "It has been suggested that Pokémon Go could potentially improve the mental and physical health of players suffering from depression and social anxiety."[11][12][13]→Aol.com and 2 other non-MEDRS sources. Can I just delete it? —PermStrump(talk)16:27, 24 July 2016 (UTC)[reply]
I would've deleted it but thought, even though it's quite new (the game), there might be medrs sources that, at least, talk about the subject outside the pop. press.--TMCk (talk) 17:40, 24 July 2016 (UTC)[reply]
It is of course too early for published research on this, but places like Baylor College of Medicine are touting health benefits[14]. Science Daily, too.[15] RS, not MEDRS, but the claims aren't too strong either: it encourages walking and walking can have health benefits. It is just a virtual version of geocaching.[16]. --Mark viking (talk) 18:37, 24 July 2016 (UTC)[reply]
Claims like "...praised by some medical professionals for potentially improving the mental and physical health..." in the lead are surely over the top, or at least it seems to me.--TMCk (talk) 18:59, 24 July 2016 (UTC)[reply]
I've tried to tighten the language used which seemed rather uncritical of the speculation seen in the sources so far. In the absence of MEDRS sourcing, we need to make clear the status of the claims against evidence-based medical standards. Is that any better? --RexxS (talk) 19:41, 24 July 2016 (UTC)[reply]
After reading the OP, I was expecting the health claims to be about increased physical activity and cardiovascular benefits or something, so I was surprised to see how many were mental health claims, which just seems ridiculous, especially the ones about autism. —PermStrump(talk)20:07, 24 July 2016 (UTC)[reply]
I'm not surprised. Every big fad sees a spate of sources that talk about how Popular Thing X relates to My Favorite Subject. So people who are interested in health will write about how it affects health. I don't know much about the game (Mark Viking's description as a virtual form of geocaching is the most sensible explanation I've seen), but I've seen headlines about its relationship to criminal activity (trespassing, drug deals), church attendance, agile software coaching, and other weird subjects. By comparison, the mental health benefits of walking and engaging in casual social interactions seem quite tame. WhatamIdoing (talk) 07:21, 25 July 2016 (UTC)[reply]
I think this is a pretty interesting discussion, because it related to what MEDRS should cover. I find that the pretty benign speculation on increased health due to walking ought to be mentioned in the article, as long as it is stated as unproven speculation — but then again I see the issues of where we ought to draw the line of benign vs. harmful speculation. Seppi333, you found some review article already? Carl Fredrik 💌📧07:43, 25 July 2016 (UTC)[reply]
There's no issue with content policies if we cite appropriate sources and say "Individuals who play Pokemon Go generally walk [possibly qualify this with 'X more steps' or 'Y more minutes' here] more than non-players.[WP:RS-quality citation] Regular light exercise like walking is associated with such-and-such health effects.[WP:MEDRS-quality citation here]" It's not ok to say "Playing Pokemon Go improves the health of players via [list of health effects of regular walking] health effects.[citation of any quality here]" because it attributes these to Pokemon Go, not light exercise; someone going around playing Pokemon Go on a motor scooter is obviously not getting those health benefits. It's definitely not ok to say that Pokemon Go improves mental health like the article did previously because there isn't a single study published on Pokemon Go at the moment. It basically read like: "Random guy named Bob with X credentials said Pokemon Go can improve player health by [list of health-promoting effects, according to Bob]". Seppi333 (Insert 2¢) 08:05, 25 July 2016 (UTC)[reply]
Funny, QuackGuru, I must have been replacing that mass removal of sourced criticism at the same time you were posting it here. CFCF, I could see an argument for a few accurately worded statements about the speculation of health benefits as far as increased physical activity and hypothetical increased socialization, but the claims in the article this morning were basically that this game cures depression, agoraphobia, and autism (I'm not even kidding). IMO it's better to leave that section out, because that material will make its way back into the article (and is/was probably already present in other areas), but have a section header about health benefits seems like an invitation for dubious and potentially harmful claims. —PermStrump(talk)08:07, 25 July 2016 (UTC)[reply]
To add to Seppi333's last comment, a lot of the claims about the potential mental health benefits originated from twitter posts by regular people without credentials who play the game. —PermStrump(talk)08:11, 25 July 2016 (UTC)[reply]
I think it was an oversight (just a guess). I don't think the paragraph you're referring to needed MEDRS. It's not like it was saying the game causes disabilities, just that people complained about the lack of accessibility for people with mobility issues. —PermStrump(talk)07:37, 26 July 2016 (UTC)[reply]
it has also been criticized for its inaccessibility to those with physical disabilities. The game is more difficult for those who are not able to move around, requiring players to walk around and have manual dexterity to experience the game.[1][2][3]
Thanks for your thoughts, Seppi - I've lodged an edit request that emphasises the social model a bit more, but does not make any medical claims. --211.30.17.74 (talk) 08:31, 26 July 2016 (UTC)[reply]
For what it's worth, I'm the one who initially added the section in question. Although science is my primary focus on Wiki, health is far outside my usual stomping ground of meteorology. I initially tried crafting in the fashion of "the game promotes walking, which is known to cause [x health benefits]", but that message got lost along the way (and probably within my own initial edit). Stumbled on this conversation by way of a link on the article's talk page. Was not aware of the additional layer of sourcing credentials for medical claims. Will be more cautious of this in the future should something similar to this game crop up. Cheers, ~ Cyclonebiskit (chat) 08:50, 26 July 2016 (UTC)[reply]
Naming suggestions for Wikiversity Journal of Medicine
As many of you might already be aware, we have a journal called Wikiversity Journal of Medicine. We invite your participation in the journal. Feel free to make submissions.
We also invite your views on whether the name of the journal should be modified. Please feel free to participate in the discussion and express your views. While some feel that it is better to leave the name unchanged, others feel that better options could be Wikipedia Journal of Medicine, Wikimedia Journal of Medicine, Wiki Journal of Medicine, Collaborative Journal and so on. Keen to know what you people feel. DiptanshuTalk05:42, 25 July 2016 (UTC)[reply]
In a 2011 survey, 292 clinical experts in Scotland ranked 19 common recreational drugs by how they perceived their personal and social harm.[1]Addiction experts in psychiatry, chemistry, pharmacology, forensic science, epidemiology, and the police and legal services engaged in delphic analysis regarding 20 popular recreational drugs.[2]
CFCF and I are having a debate over whether or not this chart (which appeared in several articles) makes a medical claim. I think it is more of a sociological claim about what some experts *believe* about the relative harms of drugs, not a statement on the *actual* harms of drugs. For comparison, (not that I'm making an OTHERSTUFF argument) we have information on electronic cigarettes about how people *perceive* e-cigs as safer regardless of their actual safety ("Many e-cigarette users use them because they believe they are safer than conventional cigarettes."). CFCF thinks the chart makes a medical claim. What do ya'll think? Sizeofint (talk) 17:43, 25 July 2016 (UTC)[reply]
^Nutt, D; King, LA; Saulsbury, W; Blakemore, C (24 March 2007). "Development of a rational scale to assess the harm of drugs of potential misuse". Lancet (London, England). 369 (9566): 1047–53. PMID17382831.
I think it probably is subject to WP:MEDRS, and ideally we would incorporate the somewhat different results produced by the similar Nutt (UK) and Dutch studies. This is not "many people believe" info. Johnbod (talk) 17:57, 25 July 2016 (UTC)[reply]
It incorporates sociological claims, that doesn't make it immune from MEDRS because it also includes medical information. A chart concerning only crime among users wouldn't fall under MEDRS, but this does because it sums up different factors, including medical ones. Carl Fredrik 💌📧18:08, 25 July 2016 (UTC)[reply]
It sums up *perceptions* of medical factors. I guess the central question is whether beliefs about medical claims are also medical claims. Roughly the same study was done in 2013 with drug users rather than experts [17] which would, I think, constitute "many people believe" info. What about the Global Drug Survey's Net Pleasure Index [18]? Are data about the subjective preferences of drug users a medical claim since the drug users are implicitly taking into account harm in their scoring? Sizeofint (talk) 18:35, 25 July 2016 (UTC)[reply]
I'm not sure what you mean about the Nutt and Dutch studies Johnbod. If this Scottish study can't be used because it makes a medical claim and isn't MEDRS then neither can those other studies. Sizeofint (talk) 18:42, 25 July 2016 (UTC)[reply]
Those who assessed the harm were "clinical experts" — the article explains how they took medical factors into consideration. Did you read it? Carl Fredrik 💌📧19:52, 25 July 2016 (UTC)[reply]
I read it a while ago but just took another look at it. The article states "there was explicit guidance provided emphasising that the harm rankings should be based on the experts' global clinical experience of the population seen in addictions services (ie, not based on an understanding of ‘milder’ wider society use patterns)." So these rankings were based on their opinions from personal clinical experiences? Is your objection more that the subjective opinions of experts might be read as a medical fact? There are studies of the perception of drugs among the general populace that could be used instead if that is more palatable to you. Sizeofint (talk) 21:07, 25 July 2016 (UTC)[reply]
In response to Sizeofint's question, "I guess the central question is whether beliefs about medical claims are also medical claims", I'd say yes, especially this one where there isn't an opportunity to appropriately contextualize the claim since it's an image, but I also think that's pretty much the crux of MEDRS (that unsubstantiated beliefs about medical claims don't belong in an encyclopedia). I do think MEDRS applies to this, but MEDRS aside, I also think think including this image in almost all of the drug articles is drastically undue given it's primary research that was a survey of experts in one relatively small country. —PermStrump(talk)22:03, 25 July 2016 (UTC)[reply]
Is it behind a paywall? I can't tell when I'm at work because it automatically logs me in. I'm going to read it later tonight, but at first glance it appears to be a review. Do you want me to email you the PDF? It looks interesting. —PermStrump(talk)00:08, 26 July 2016 (UTC)[reply]
No, I don't believe it is behind a paywall, although the Lancet will try to make you register. I found a copy on Google Scholar. Also found the Encyclopedia of Health Economics reproduces some of this information (Table 4) [20]. Sizeofint (talk) 00:21, 26 July 2016 (UTC)[reply]
This discussion is basically the perennial "shall we have an evidence-based-only encyclopedia, or shall we have a comprehensive encyclopedia" all over again. I'm in favor of comprehensive coverage of medical subjects myself. If the image needs context, then write some context! There's no shortage of pixels in that section. WhatamIdoing (talk) 11:57, 26 July 2016 (UTC)[reply]
No, that I disagree with entirely. This is about whether we should employ appropriate guidelines when it comes to medical professional's assessment of harm concerning a specific disease/condition. Opening this up would allow for any manner of "expert opinion" on any manner of subjects — there are loads of "experts" that think colloidal silver is fine, but pretty much no MEDRS-sources that do. I don't especially disapprove of the image as such, but we need to be consistent in how we employ MEDRS — anything undermines the whole purpose of having it at all.
This discussion has nothing to do with if we should be a comprehensive encyclopedia: there are sources that are better than this, so use them instead. Carl Fredrik 💌📧14:06, 26 July 2016 (UTC)[reply]
I have a big issue with the amount of times it was used. Even the better sources shouldn't be on every single drug article. It's undue weight that makes it seem like there's some objective ranking of drugs that is stable over time and location, which doesn't even apply Nutt (2004 2007*) that synthesizes ~9 or so the different ways harm can be measured. "Societal cost" is going to vary region to region and year to year, depending on what's popular and where it's coming from, but it's easy to forget that when you see the same graph on every article. Plus, when it's used in that many articles, space for contextualization is limited to the caption in many places. I think Nutt (2004 2007*) or a similar graph should only be on like Recreational drug use and Substance use disorder, but not on every single individual drug article. —PermStrump(talk)15:06, 26 July 2016 (UTC)[reply]
Alright, let's so let's use Nutt (thought it was 2007 not 2004) then? There have been several variants of Nutt's ranking in the last decade and all of them show similar orders. It is a subjective ranking but a subjective ranking that is mostly agreed upon by different teams of researchers. It was on so many drug articles in various forms for about nine years, I think, because readers find it informative. Sizeofint (talk) 15:22, 26 July 2016 (UTC)[reply]
What about the views of the more general populace then CFCF? For example, information that in such and such survey X% of smokers don't think smoking has adverse health effects? Sizeofint (talk) 15:38, 26 July 2016 (UTC)[reply]
MEDRS would apply — it isn't a very high standard to hold for a source like that. All it requires is a decent secondary source, which WP:RS does for controversial statements anyway. Such a statement would also need to be put into context explaining that this belief is entirely false. Carl Fredrik 💌📧19:29, 26 July 2016 (UTC)[reply]
NIDA comes out with the Monitoring the Future survey every couple of years which captures kids' perception of the safety of different drugs and it always gets discussed later in a lot of reliable secondary sources. I can only name resources about teens off the top of my head, but there's way more research on adults so I'm sure there's some counterpart to MTF for the adult/general population. —PermStrump(talk)19:48, 26 July 2016 (UTC)[reply]
If you are excluding information because it is "expert opinion" rather than "evidence based", then, yes, this actually is exactly the "shall we have a comprehensive encyclopedia?" question.
If there were "loads of experts" saying "colloidal silver is fine" – i.e., a majority of a reasonably representative sample of actual experts, not merely a few guys who managed to buy a PhD somewhere and work for companies that sell the stuff – then we really ought to say that in the relevant articles, too, complete with an explanation of exactly what "fine" means. (Personally, I wouldn't be the least bit surprised if such a group said that colloidal silver is "fine" in the sense of "a fine way to part fools from their money" or "fine" in the sense of "very finely ground", or even "fine" in the sense of "relatively harmless in low doses", but I doubt that you'll find any saying that it's "fine" in the sense of "efficacious for treating any actual disease".)
However, unlike that red herring, all of the sources basically agree: some drugs are more (or less) harmful to individuals, and some drugs are more (or less) harmful to society, and these can be ranked on a relative scale, and multiple attempts to rank them have produced basically the same results. I feel like the objections here are amounting to a game of Mother, May I?. The point behind preferring recent, independent, secondary sources is to get the material right. The point is not to show that we like fancy-sounding sources or to set up some bureaucracy that effectively excludes most editors. The point is not even to protect other articles from a hypothetical slippery slope, in which one "merely reliable" source causes the downfall of all standards. When you know that the material is right, i.e., all of the reliable primary sources say (more or less) the same thing, and no reliable sources contradict this information, then it's not helpful to nitpick over the source. An accurate summary of expert views on the harmfulness of various drugs is DUE, regardless of the exact type of the source that's currently supporting that information.
It may be worth pointing out again that the policy requires only "any source that he or she believes, in good faith, to be sufficient", and that once such a source is provided to support information that is verifiable, DUE and appropriate for an encyclopedia, the role of editors who want a different or better source is to find a better one themselves, not to blank everything for being less than perfect. I know that this isn't popular with some editors, but this is the policy. There are no "except for medical content" or "except for FRINGE material" or any other exceptions to this policy. If that kind of material is verifiable and belongs in the article (and I say expert views on the relative harmfulness of drugs does belong, and I suspect that almost every editor outside this discussion would agree with me), then you need to collaboratively improve it instead of just blanking it. WhatamIdoing (talk) 21:09, 27 July 2016 (UTC)[reply]
If the data are included in the Nutt review in some format (e.g., graphically, numerically, etc) and the image caption accurately reflects that the data in the graphic is derived from expert opinion, then I don't think there's an issue with MEDRS or any other content guidelines/policies. The important thing is to accurately and clearly state where the data is derived from in the image caption, otherwise there could potentially be issues with WP:V and/or WP:UNDUE, depending upon the context and how the caption is worded. Personally, I don't really agree with the ordinal relationships between a few compounds in the bottom diagram based upon their acute and long-term effects from repeated overdose, but I don't particularly care that much since I recognize that it's not based upon an objective measure. Seppi333 (Insert 2¢) 23:32, 27 July 2016 (UTC)[reply]
Looking back through earlierdiscussions it seems clear that we need a guideline (if we don`t already have one, somewhere) on the topic, and that it should reflect a consensus with wp:WikiProject Biography.
I`d further suggest that "Person X, who was notable, later developed condition Y" isn't good enough, even with good sources. We need evidence that "Person X's condition Y was notable". For inclusion on a manageable, finite list we need further evidence that "Person X had one of the most notable cases of condition Y". To take a nonmedical example, David Beckham includes sourced assertions about his 32 tattoos, but it is neither an extraordinary number that would set him high among all people with tattoos, nor is it a defining characteristic of his that would ``generally`` be seen in the lede of sources covering him. Hence it should not justify his appearance on a List of people with tattoos. In contrast, J. Edgar Hoover or Alan Turing might justifiably belong in a List of closeted homosexual men, while Stephen Hawking or Lou Gehrig would sensibly be included in any List of people with ALS. Their condition was not defining of them, but it is nearly always discussed when they are described. A borderline example might be Ronald Reagan, whose Alzheimer's disease came long after his principal notable activities, but whose case was used widely in fundraising and awareness efforts, with Nancy's support.
Some of these categories and lists exist primarily for maintenance tracking purposes, to ensure that problem assertions of a controversial nature are given visibility. Would not a hidden maintenance category be the better mechanism to use in such cases?
Just a few thoughts, I'm sure there are many others. Should we bang it around for a while, or just start work on a draft directly and let the ideas evolve around the draft? LeadSongDogcome howl!18:17, 28 July 2016 (UTC)[reply]
The criteria will depend upon the subject. "People with extremely common thing" needs a different set of rules than "People with rare thing". If it's common, then you should exclude as many entries as possible, to make the list usable. But if it's rare, you should tend to include most entries that can be reliably sourced (e.g., to a self-identifying news article), so that the list can be reasonably representative. WhatamIdoing (talk) 16:43, 29 July 2016 (UTC)[reply]
On a related note every one of this six diseases currently have no related page on our sister project Wikiquote. Anyone here willing to start the pages there? Solomon796807:11, 28 July 2016 (UTC)[reply]
These mostly look like lists. I think that articles could be written and sourced for AIDS in popular culture, and I believe that malaria appeared in British literature during the late 19th and early 20th century, but from a Western/American POV, I'm not sure that the others appear much in (Western) popular culture. Tuberculosis is easy precisely because it was originally not considered a disease of poverty. (See Disease#Language of disease and elsewhere on that page for some sourced information.)
A French lit professor I had in college told us that a lot of times in literature TB was a euphemism for syphilis. (I have sources for that.) Can I put that nugget anywhere? :) —PermStrump(talk)10:29, 28 July 2016 (UTC)[reply]
Please consider adding information about the stigma to both Syphilis and Tuberculosis.
If your source supports a more general statement (the general case of people lying about medical conditions because of stigma, not just these two particular conditions), then it might be good to get that in Disease and/or Terminal illness. There are probably some other conditions that could be sourced. For example, I understand that SIDS deaths have significantly decreased (on paper) since coroners quit using SIDS as a euphemism for suffocation. WhatamIdoing (talk) 15:35, 29 July 2016 (UTC)[reply]
Any improvements to the article. It's something we do here at WPMED: if someone we know dies, then we try to improve an article related to that person. Since this editor was an organ donor, articles on organ donation are a good match.
I've worked on those articles in the past, although it's been a long time since I looked at them. Most of them have a lot of information about the sociolegal issues of convincing people to be donors. It might be interesting to see whether we could turn up some sources about how the surgeries happen, and expand the articles with that technical information. It sounds like his heart was transplanted, so perhaps that's an area to focus on. WhatamIdoing (talk) 12:47, 30 July 2016 (UTC)[reply]
.
Facepalm. >.> Seppi333 (Insert 2¢) 05:16, 25 July 2016 (UTC)
.svg)
