Wikipedia talk:Identifying reliable sources (medicine)/FAQ

Biomedical information
This note is to explain the correction that I made here:

MEDRS applies to "biomedical information". It does not apply to all medical content; It does not apply to all health content; it definitely does not apply to all content that could be (mis)construed as being health content.

MEDRS only applies when you are making a statement that has both "bio" and "medical" components. Whether patients choose to comply with a prescription more frequently if the prescriber comes from their class is "medical" content. However, it is not subject to MEDRS, because there's no "bio" in that. Whether there are typically 46 chromosomes in a human cell is "bio", but it's not "medical". It's plain human biology. Therefore MEDRS is not required for that, either.

I realize that this "limitation" is not popular with certain woo-fighters, who occasionally wish (for example) to demand MEDRS-compliant sources on the strictly non-biological (and therefore non-biomedical) fact that a huge number of licensed physicians choose to recommend chiropractic or acupuncture treatment to their patients – but that kind of statement is not covered by MEDRS. What we think of as a "MEDRS-compliant" source is required only for the biomedical subset of all medical- and health-related information in Wikipedia. WhatamIdoing (talk) 06:15, 6 August 2015 (UTC)

Why
Under the question "Why do you have special rules for medical information?", it says that we want to get it right "Since Wikipedia's readers may make medical decisions based on information found in our articles". This is probably true for some people. However, I'm not sure that this is "the (sole) reason"; we want to get it right because we want to get it right. Also, the consequence doesn't follow logically: If the sole rationale is that people make health decisions with Wikipedia, then we need to get the information right. (We would also need to write in lay-friendly English and omit information that might cause someone to make the "wrong" choice.) Using high-quality sources is a means to an end, not an end in itself. WhatamIdoing (talk) 00:06, 13 August 2015 (UTC)

When
The section "When do I need to follow MEDRS?" says, "Content about human biochemistry or about medical research in animals is also subject to MEDRS if it is relevant to human health."

I think this is too sweeping. Is there any aspect of human biochemistry that isn't – somehow, perhaps indirectly – "relevant to human health"? Also, why human biochemistry and not human biomechanics? And how far do you want to take this? Do you need a MEDRS-compliant source to say that children sometimes cry when they fall down? There's a whole lot of biochemistry involved in that process.

Ditto for medical research in animals. This kind of oversimplification makes it very hard to write solid science about toxicology and environmental health research. I think this is going to lead to writing bad articles and blanking good content over needless sourcing disputes. It's a gift to POV pushers. We need to find a better way to say this. WhatamIdoing (talk) 00:15, 13 August 2015 (UTC)

Review: ideal?
I query the recommendation "For most (not all) purposes, the ideal source is a peer-reviewed review article", which goes along with "and the [reverting] editor said I needed to use a review". As MEDRS pyramid diagram shows, clinical practice guidelines are probably the closest we have to "ideal". When such guidelines come from an international body of experts, reviewing and grading evidence, and combining that with "best practice" for all the things that RCTs don't necessary cover, then they really can be a gold mine for sourcing. We are, after all, supposed to reflect the expert consensus.

There are various kinds of reviews. A narrative review may only reflect the opinion of the authors [who may in turn be the same folk who did the trials]. A systematic review can be so algorithmically constrained, that the authors end up saying little that is useful. A meta-analysis can only be done if there is homogeneity in the studies. Editors seem often over-focused on "efficacy" and perhaps "adverse effects" which are important when covering treatments, but in a disease-focused article, or a treatment that is more complex than just taking a pill, that may only cover a small aspect of the article topic. A professional and condition/treatment-specific textbook may outclass all of these for some aspects of an article, as it can be more comprehensive and holistic than a journal review paper. Of course, such a book requires one to make more of an effort and commitment than following a PubMed link. So, in short, I don't think we should push "review article" as "ideal". It is merely "good" and "better" than a research study paper.

I think we should also encourage editors to immerse themselves in the literature on a topic, before doing significant work on it. That is the only way we can determine WP:WEIGHT. I often see the addition of factoids linked to this or that review. The result is many of our articles are a fairly disorganised collection of facts, that neither introduce and explain nor are comprehensive. So really, what one needs is not "a review" but to collect several high quality sources, read them all, and cite one or two. That would be my advice in a FAQ. -- Colin°Talk 16:10, 8 January 2019 (UTC)


 * Agree. One problem I have noticed with review papers is that in aggregating the results of many separate trials a significant outlier may be omitted or downgraded, e.g. a result that only applies to a particular genetic population or diagnosis. Of course that is exactly what a review is supposed to do! But useful, reliable information (even from primary sources) may be lost, and the article may be misleading, especially with negative results such as "Xxxx has not been found to have any effect greater than placebo on Ccccc". D Anthony Patriarche (talk) 18:51, 2 January 2023 (UTC)

Fake heading
The navigation (esp. mobile) and editing of the page is a bit difficult now. This seems to have a reason, which I’m not sure I understand. Can we change it back? -- Dustfreeworld (talk) 18:15, 27 January 2024 (UTC)


 * This FAQ is transcluded at the top of WT:MEDRS (in a collapsed section). I suspect that @Seppi333 changed the formatting so that the section headings in the FAQ didn't appear in the Table of Contents for WT:MEDRS.  WhatamIdoing (talk) 18:36, 27 January 2024 (UTC)
 * Thanks WAID; your explanation clears things up! -- Dustfreeworld (talk) 21:30, 3 February 2024 (UTC)
 * It makes this page very hard to use though. Bon courage (talk) 06:05, 9 February 2024 (UTC)
 * We could put vanchor on each section heading and something like
 * • •  •  •  •
 * (only prettier) at the top of the page, if that would make it easier. WhatamIdoing (talk) 23:15, 9 February 2024 (UTC)