Wikipedia:WikiProject Medicine/Referencing case studies for MEDRS

The main purpose of this page is to gather some representative case studies for the way the WP:MEDRS guideline is applied.

The preliminary resources immediately below build up accounts of the normal situation in medical referencing. Case studies should also exhibit edge cases, where they occur, against a realistic background.

Resources

"Why MEDRS?" essay

Wikipedia:Why MEDRS? sets out to justify the rigorous nature of the guideline.

What is biomedical information?

From Wikipedia:Biomedical information, an annexe to MEDRS:

"Biomedical information is information that relates to (or could reasonably be perceived as relating to) human health. Generally speaking, such information should be supported by a reputable biomedical source, such as review articles, higher-level medical textbooks, and professional reference works."

In a nutshell

The "nutshell summary" for MEDRS is this:

Ideal sources for biomedical material include literature reviews or systematic reviews in reliable, third-party, published secondary sources (such as reputable medical journals), recognised standard textbooks by experts in a field, or medical guidelines and position statements from national or international expert bodies.

Tutorial summary

There is a quick tutorial Editing Medical Topics. A summary from it is:

Good sources:

• are typically no more than 5 years old
• summarize secondary sources or many primary sources
• provide an overview of the current understanding of the topic
• combine the results of several studies
• are published independently of the research

From the MEDMOS manual page

The Wikipedia:Manual of Style/Medicine-related articles#Citing sources medical manual of style has these points to make under "citing sources":

Citing sources (copied across 2018-06-20)

Sources should be used to make verifiable statements, but they should not be discussed by themselves in an article.

• Do not provide a detailed analysis of an individual study unless the analysis itself is taken from a published reliable source. Wikipedia should concisely state facts about a subject. It should not discuss the underlying literature at any length. Generally speaking, the facts will be found in the conclusions or results section of a study, not in the detailed methodology. Articles that rely on secondary sources are less likely to fall into the trap of discussing the size of a single study, its methodology, its biases, and so forth. Thus, "washing hands after defecating reduces the incidence of diarrhea in the wilderness", not "An uncontrolled survey involving 132 experienced long-distance backpackers on the Appalachian trail in 1997 concluded that washing hands after defecating reduces the incidence of diarrhea in the wilderness."
• Do not hype a study by listing the names, credentials, institutions, or other "qualifications" of their authors. The text of the article should not needlessly duplicate the names, dates, titles, and other information about the source that you list in the citation. Always omit professional titles and academic degrees: use "Smith" or "Jones" rather than "Dr Smith" or "Prof Jones". It is necessary to specifically include such information only when a specific individual is being cited as an example of a person holding a minority view: You might write, "The AIDS Denialist Society says that HIV is entirely harmless", but just use a plain statement for the widely accepted fact, "HIV causes Acquired Immune Deficiency Syndrome."
• Do not publish your own views about studies.
• Adding sources to the lead is a reasonable practice but not required as long as the text in question is supported in the body of the article

Case 1: Antidepressants

See Talk:Antidepressant/Archive 2#comparative efficacy and tolerability table. The table was criticised as

1. "Original research" (say breaches of the WP:OR content policy requiring referencing, and possibly WP:SYNTH for aggregation);
2. Relying on outdated references, see WP:MEDDATE, "In many topics, a review that was conducted more than five or so years ago will have been superseded by more up-to-date ones [...]".

Case 2: Exercise

See Talk:Exercise. The page is tagged with {{WikiProject Medicine |class=c |importance=High |cardiology=yes |cardiology-imp=low}}, meaning that it comes within the scope of MEDRS. The potential inclusion of material on animal exercise (i.e. for non-humans) has led to dispute on whether and how MEDRS might apply to such material. There is a formal record at Wikipedia:Dispute resolution noticeboard/Archive 163#Talk:Exercise.

Case 3: Kava

See Talk:Kava/Archive 2#What will constitute sufficient evidence that kava helps with anxiety and that it is being investigated for its potential anti-cancer properties?.

This discussion walks through the "health food" or treatment implications of the literature on kava tea. It was argued that:

1. A "double-blind, randomized, placebo-controlled study" is to be treated as "a limited short-term study on a small number of subjects", not qualifying for WP:MEDSCI and WP:MEDASSESS.
2. A Cochrane review reference, at that time 14 years old, passed muster.
3. Some Therapeutic Goods Administration studies were regarded as of "supplement products with no proof of safety or efficacy".

Case 4: WHO guideline for chiropractic

See Wikipedia talk:WikiProject Medicine/Archive 45#Is WHO guideline a MEDRS.

1. A WHO guideline from 2005 was stated (in 2014) to be reliable source (provided it is not contradicted by a newer Cochrane review), there being then no newer guideline that applied from the WHO.
2. The application of this guideline was criticized for use in chiropractic, in alternative medicine, on the basis that it was "written for the WHO by a bunch of industry insiders, largely in response to political pressure".
3. On the other hand, WP:MEDRS #Medical and scientific organizations specifically mentions the WHO as a reliable source.

The main point at issue was the safety of chiropractic. This paper, a systematic review, was stated to be the only paper on the topic within the five-year limit. It was stated to be inconclusive.

NB There is a long history of contention here. See Wikipedia:Mediation Cabal/Cases/2010-08-23/Chiropractic. But one should note also that MEDRS is not static, and comments there about the "review cycle" are absent from the current version.

Case 5: Pancreatic cancer

See Talk:Pancreatic cancer/Archive 3#Risk factors: herbicide use. This paper, from 2009, was characterized as "preliminary research", being "a beginning finding from a limited cohort study [...]" on what it had to say to indicate a possible link of herbicides to pancreatic cancer.

Case 6: Curcumin

See Talk:Curcumin#Recent additions.

1. A source that was a comment in Nature was rejected under WP:MEDREV.
2. Other related sources were rejected as "appear to be out-of-date or off-topic", in contrast to a 2017 review article that was in use.
3. Another reversion of an edit, half "not in the cited source", and half under WP:UNDUE as based on a single study, proved contentious.

Case 7: Mirtazapine

See Mirtazapine and Talk:Mirtazapine. Mirtazapine (Remeron) is an atypical antidepressant. At the time of this writing, the article itself contains nine uses of the {{unreliable medical source}} template, all dated to October 2011. No talk page discussion of these.

The talk page contains a discussion of the appropriate conclusions to extract, relating to mirtazapine, from a National Institute for Health and Care Excellence (NICE) study. It concerns a 2010 follow-up study, one type of longitudinal study. NICE's clinical guidelines pass MEDRS as evidence-based and of high quality. What was cited, it was argued, was (a) not in the NICE conclusions, and (b) subject to WP:MEDDATE.

Case 8: Bacon

See Talk:Bacon#Health risks. The issue is the due handling of health risk from eating bacon. Points raised:

1. Specificity: The health risk for bacon should be discussed independently of the risk for processed meat in general.
2. Type of source: not journalism. At the time of this writing, on the other hand, the article contains a reference to the BBC website, tagged with {{unreliable medical source}}, and dealing with the health risk, that cites the WHO.
3. Prominence: Does the health risk warrant a mention in the lead section?

A list of English Wikipedia articles containing {{unreliable medical source}} can be generated with this query.

Fringe cases

Tagging of an article talk page by {{Ds/talk notice|acu}} and/or {{Ds/talk notice|ps}} gives standing to the classification of the topic as alternative and complementary medicine or pseudoscience. (The tagging will be on the actual talk page, and is not applied to archives of talk pages.)

Case 9: Emotional Freedom Techniques

See Talk:Emotional Freedom Techniques/Archives/2014/November#Horribly_biased. It was argued that a reference to Current Research in Psychology was unreliable, because of the closeness of the author to the treatment, and failed MEDRS because the journal had a predatory publisher.

Sources defining "predatory publishers" include Beall's List.

Case 10: Young blood transfusion

See Talk:Young blood transfusion.

1. Science-Based Medicine, which is a blog, is customarily considered a reliable source for matters covered by WP:FRINGE. While blogposts as a general rule are not accepted as reliable sources, the general guideline at WP:RSSELF admits some special cases.