Talk:Evidence-based medicine

From Wikipedia, the free encyclopedia

	This article is within the scope of WikiProject Medicine. Please visit the project page for details or ask questions at the doctor's mess.
B	This page has been rated as B-Class on the quality assessment scale
Top	This article has been rated as Top-importance on the importance assessment scale

[edit] Moved unencyclopedic comment:

The expert is dead! Long live the expert!

-Montréalais

Epidemiology is evidence-based - read epidemiology. Also greatly tightened up the case study bit. - David Gerard 16:30, Jan 16, 2004 (UTC)

Excuse me, but Epidemiological research is not randomized clinical trial double-blind research. The medical scientism people are definitely laughting at all Epidemiological research. And, case-studies positively were the hot-item before the arrival of evidence-based medicine. If you want it tighter, than make it tighter. It looks good to me, as is. -- Mr-Natural-Health 17:45, 16 Jan 2004 (UTC)

Strunk and White are my heroes ;-) - David Gerard 19:20, Jan 16, 2004 (UTC)

[edit] I was invited to write here!

I was invited to write here! I added two objective agruments and they were deleted, of course, without comment. What gives? Was your offer more double talk. Wholesale deletion of text is strictly prohibited. -- Mr-Natural-Health 17:38, 16 Jan 2004 (UTC)

Did you write "Belief in only one way of doing medical research totally trashes the branch of medicine called epidemiology.

Is this the deletion you are talking about? Because David has commented. See above. Also the fact that you were invited to write an article does not mean that other people cannot edit it. theresa knott 17:49, 16 Jan 2004 (UTC)

Well excuse me, but David's comments actually support my position that an article on Medical Scientism is required. David says epidemiology is valid science. I say that epidemiology is valid science. But, the medical scientism people say that it is quackery.

Nutrition used to treat medical conditions is classified as CAM research by PubMed. Nutrition is not quackery becuase it is based on epidemiological research. Therefore, there is more than one way of medical research, and the Medical Scientism people are positively full of POV. Frankly, I am tired of having to carry on 5 conversations at the same time, with the same biased person. -- Mr-Natural-Health 18:22, 16 Jan 2004 (UTC)

"But, the medical scientism people say that it is quackery." - Refs? (And not ones requiring a subscription.) - David Gerard 19:12, Jan 16, 2004 (UTC)

???? We cannot quote comments made by RK and his other cronies. Just follow the talk pages. -- Mr-Natural-Health 19:30, 16 Jan 2004 (UTC)

Just a comment: Please do not use PubMed classifications as absolute truth. PubMed CAM classifications are nothing more than a search filter. You can view the contents here: [1]. And just to disprove the assertion above, here is an article about the effects of nutrition on heart problems, which is not classified as CAM: [2]. Rasmus Faber 23:27, 16 Jan 2004 (UTC)

No problem! It is easy for YOU to be confused! Searching on PubMed with this text string:

cam [sb] nutrition

yields 14292 hits.

Searching on PubMed with this text string:

cam [sb] epidemiology

yields even more hits at 17073.

Ergo, my point is perfectly valid and is not POV. -- Mr-Natural-Health 02:01, 17 Jan 2004 (UTC)

I was not claiming that your point was POV. Only that it used a bad argument. PubMed does not classify nutrition or epidemiology as anything. Rather when entering "cam [sb]" into the search field, they expand your search string into a larger search string, which among other things include "supplement AND vitamin c". To use your argument: here is an article, which popped up when searching for cam [sb] with the limit : Randomized Controlled Trial: [3]. So chemotherapy and stem-cell transplantion is CAM?! Probably not, but the article contained the right combination of keywords. Likewise with nutrition and epedemiology, which often deals with vitamins and supplements, and thus often contains words also associated with CAM. Rasmus Faber 12:19, 17 Jan 2004 (UTC)

I still consider YOUR POINT, too minor to even bring up. I never make absolute claims, because they are absolutely always wrong. I am not responsible for how PubMed has managed to bungle up their database.

What I find important/interesting is that the very first controlled research study ever done was done in the 1700's on the subject of scurvy in the British Navy. The concept of using diet rather than a medication (like seawater) to treat the medical condition called scurvy classifies it as Alternative Medicine as well as a Natural therapy. The fact that it is also classified as nutrition, etc., etc., etc., and that medicine will also claim it as one of their very own studies is totally besides the point. -- Mr-Natural-Health 16:03, 17 Jan 2004 (UTC)

We now begin discussing the definition of Alternative Medicine. I have moved the discussion to Talk:Alternative medicine. Rasmus Faber 08:13, 19 Jan 2004 (UTC)

What in the world are you talking about? Boy are you strange! -- Mr-Natural-Health 01:11, 20 Jan 2004 (UTC)

The case study para - should that be in the case study entry? - David Gerard 19:20, Jan 16, 2004 (UTC)

No! Being that evidence-based medicine is all about population evidence, more than half a sentence on case study should be present in the criticism section. -- Mr-Natural-Health 19:37, 16 Jan 2004 (UTC)

Medical Scientism cannot be referenced!

The mantra of medical scientism is randomized clinical trials, double-blind, peer reviewed, studies published in respectable journals. Epidemiological research is not the randomized clinical trials and double-blind stuff that drug research is made of.

If any one element of this mantra is missing, then that published research study is classified as being invalid research. This kind of decision is made quite arbitrarily by the medical scientism people. And, their mantra clearly renders the vast majority of all published health research invalid and virtual quackery by implication.

The primary treatment method of medicine is medication. And, their mantra is clearly designed for drug testing. So, forcing a drug testing paradigm upon other kinds of health research is patently absurd. Unfortunately, people like RK don't see it that way.

This phenomenon of medical scientism is quite real because it dominates all science newsgroups, mailing lists and of course the health related articles in Wikipedia.-- Mr-Natural-Health 21:00, 16 Jan 2004 (UTC)

This is the Evidence-based medicine article, not the Medical Scientism article. There's a book, there's a website. There is reference material for you to back up your assertions with. - David Gerard 10:43, Jan 17, 2004 (UTC)

I know what Evidence-based medicine is, probably more so than you do. According to BMJ physicians are supposed to spend on average 2 hours researching (ie, as a transitive verb) the best medical treatment for each patient that walks through their door; even though they spend an average of only 10 minutes in each visit with a patient. EBM in many respects has a lot in common with patient empowerment as any patient expecting to survive medical treatment better be in control to the point of making their own decisions.

I have already defended my position on the talk pages of both Medical Scientism and Scientism. Whether or not I can or cannot come up with a reference is purely an academic question. In Alternative medicine I have come up already with about 20 references. If you were to care to look, the reference section of Alternative medicine is actually longer than the text. Evidence-based medicine frankly is not even set up for a serious discussion of the topic. Your method of referencing is too cumbersome to fool with.

Set up a proper Reference Section with my references to research etc. fully disclosed and I might consider spending more time on improving this article. -- Mr-Natural-Health 15:50, 17 Jan 2004 (UTC)

[edit] reasons for removing material

I removed the following text from the criticism section because the individual assertions are either (1) true but applicable to medical practice in general rather than being specific to EBM, or (2) erroneous, reflecting a misunderstanding by the contributor of what EBM is and isn't.

Removed material

Some treatments take a more holistic approach, which may be difficult to fit into a testing model that assumes the patient is a passive object acted upon by the treatment.

EBM includes no "assumption that the patient is a passive object." The primary texts of EBM explicitly include the patient's individual values and preferences and provide methods for judging treatment outcomes that are not restricted to drug treatment.

Critics also raise conflict of interest. Journals such as the New England Journal of Medicine, The Lancet, JAMA, and the British Medical Journal have been unable to prevent papers ghostwritten by pharmaceutical companies from being published. These same pharmaceutical companies are a primary source of funding for medical and drug research. In some cases, doctors listed as authors on ghostwritten research papers did not review the raw data, only tables compiled by a medical writing company. (See also Flanagin A, Carey LA, Fontanarosa PB: Prevalence of articles with honorary authors and ghost authors in peer-reviewed medical journals; Larkin M: Whose article is it anyway?)

Conflict of interest is no more specific to EBM than any other types of journal article. In fact, EBM explicitly encourages awareness of sources of information and potential conflict of interest.

Some critics also claim that evidence-based medicine seems more concerned with the job security of researchers than with solving health problems.

You could say this about anybody you don't like. Doesn't belong in an encyclopedia.

Long before evidence-based medicine came along, case studies were an acceptable form of medical research. Case study methodology can be applied effectively to the study of men with chronic coronary heart disease, for example. "Case study methodology can be used as a creative alternative to traditional approaches to description, emphasizing the patient's perspective as being central to the process. Contemporary practitioners and researchers have come to appreciate the subjective richness of patients recounting their experience and the meanings implicit in them to help guide practice." (Zucker, DM: Using Case Study Methodology in Nursing Research)

The importance of the individual patient in the case study method runs counter to evidence-based medicine's emphasis on population evidence.

EBM does not reject case studies or any other type of evidence; it provides criteria for comparing the usefulness and strength of different types of evidence for a specific purpose. The quoted paragraph is perfectly true but perfectly irrelevant to EBM.

However, whereas in the past population trials have had to group patients according to very broad criteria such as age or according to basic parameters such as blood pressure, with increasing access to comprehensive genetic and physiological testing future trials may well be able to combine the best aspects of large scale epidemiology with the detailed investigation of individual patients. That is, the trials will be of thousands of 'individuals' and the data provided by increasingly be more relevant and more easily applied to the patient before the doctor.

Again, probabably true but certainly not a criticism of EBM. This is simply an assertion that new evidence based on new technology will be better because more variables can be taken into account.

I'd be happy to defend specific points if anyone disagrees. alteripse 14:52, 15 Jan 2005 (UTC)

[edit] Possible revisions to Criticism section

Regarding the first part of the first sentence of the Criticism section: "Critics of evidence-based medicine state that doctors were doing these things already..." Without specific reference to what the author meant by "these things", the point is unclear. Is the author attempting to say that doctors were already practicing evidence-based medicine before the term "evidence-based medicine" came into popular usage? If so, why is that in the criticism section? That seems like good support for evidence-based medicine!

Regarding the second part of the first sentence of the Criticism section: "Critics of evidence-based medicine state...that good evidence is often deficient in many areas..." In essence, this particular criticism is directed towards therapies that are performed despite the lack of evidence (i.e. alternative therapies), not those that are performed with evidence (i.e. evidenced-based therapies).

Regarding the third part of the first sentence of the Criticism section: "Critics of evidence-based medicine state ...that the more data are pooled and aggregated the more difficult it is to compare the patients in the studies with the patient in front of the doctor. i.e. EBM applies to populations, not necessarily to individuals." The point here is not clear. Having more information makes it more difficult to treat an individual?! Why is that?

These points should be clarified or the sentence removed. Edwardian 06:37, 25 Apr 2005 (UTC)

EBM deserves to be tagged as "quackery" as much as any other modality. Any attempt to intimidate me into silence is just that: intimidation. 68.226.125.108 21:40, 30 March 2006 (UTC)

One of the key aspects of an EBM approach is to consider how similar this patient is to the patients in the trials. In practice it isn't that hard for a jobbing GP to do, but putting it on an exact numeric basis would be non-trivial. Aggregating like with like, rather than lumping together all studies is expected as well. Midgley May 2005

[edit] in Criticism of...

"In The limits of evidence-based medicine (http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=11728302&dopt=Abstract), Tonelli argues that "the knowledge gained from clinical research does not directly answer the primary clinical question of what is best for the patient at hand." Evidence-based medicine seems to them to discount the value of the case study."

But ... while a case study is interesting, and is written to stick in our memory, a study of one patient or case also considerably less than directly answers the question of what i should do with another patient who is beside me now.

EBM is pretty good for distinct explicit medical-model conditions, particularly diseases, but I think one area of confusion is around the straw man that this is all one is allowed to do - much of family practice deals with non-disease, and with MUPS (medically unexplained physical symptoms). Midgley

[edit] does this really belng in "criticism of"?

"In managed healthcare systems evidence-based guidelines have been used as a basis for denying insurance coverage for some treatments some of which are held by the physicians involved to be effective, but of which randomized controlled trials have not yet been published." User:Midgley

It seems basically true to me, and worth mentioning, although I did not add it. The insurance angle was not anticipated by early EBM proponents, but it is fairly easy for an insurance company to deny coverage for something on the basis that published, high level evidence is weak. Of course some might agree and some might disagree as to how strong the evidence should be before insurance covers it. Messy topic, with no pure motives or unbiased perspective from any party. alteripse 23:23, 28 May 2005 (UTC)

I comment _only_ on the positioning of it, not on the entirely reasonable points above. Perhaps it should be under a heading of "use of EBH" or "adverse use ..." although that latter is not a NPOV.

I wonder to what extent it was not contemplated that health care commissioning and funding organisations (a more general term that fits the NHS and some other European systems as well as the commercial and public ones in the USA) would use an evidence-based approach? I think the idea came in fairly early but I can't cite a source.--Midgley 09:02, 30 May 2005 (UTC)

[edit] Category

Not medicine? Midgley (talk · contribs)

Adrian, Category:Medical informatics is a subcategory of Category:Medicine. But I agree it should have its own category. JFW | T@lk 08:56, 1 December 2005 (UTC)

EBM should be considered quackery, considering it's minimal contributions to health, its opportunistic exploitation by pharmaceutical companies, and the fact that those who claim to support it conveniently ignore case studies that do not support their modalities. Also, this article is very biased, and any attempt to remove the disputed tag at the top should be considered vandalism. The poor little MafiaDoctors can dish it, but they can't take it. Can anyone say cartel behavior?

Cartel behaviour.

No seriously now, it is your tagslapping that is vandalism. Your personal opinion is really irrelevant here; what is relevant is whether large groups of people disagree with the premises of EBM. You will need to quote serious sources. If the Americal Association of Homeopathy publishes a pamphlet against EBM that may be notable. Accusations of complot theories or cartel behaviour are actually buzzwords for Wikipedians to become really sarcastic. Please collaborate like a serious adult. JFW | T@lk 21:38, 3 April 2006 (UTC)

[edit] inaccurate reference

This comment refers to the (current) first paragraph in the section 'criticism of evidence-based medicine'. Starting "Critics of evidence-based medicine maintain...

The Tonelli reference seems to be misrepresented, at least by implication. Tonelli's point is that evidence-based medicine isn't enough by itself. He is strongly opposed to basing patient care solely on EBM, but strongly in favor of EBM as one of several bases for clinical care. His own summary of the article begins "The importance of clinical research for the practice of clinical medicine is immense and undeniable. Yet the type of knowledge gained from clinical research, referred to here as "empirical evidence," is itself insufficient to provide for optimal clinical care."

This is not a criticism of EBM, but rather an inherent limitation of EBM as currently practiced. Would it fit better in a different section?

129.255.102.104 14:51, 9 February 2006 (UTC)bill-morris@uiowa.edu

[edit] criticism section??

The majority of this section seems to be about limitations of EBM, all but the last paragraphs seem supportive of using scientific evidence, but critical of basing all decision on only published articles, and particularly critical of basing all medical decisions solely on published randomized blinded controlled trials. A section title of limitations might be more NPOV.

The last paragraph, and some of the sense in the first two paragraphs do seem to be criticism, but criticism of the misuse of EBM rather then criticisms of EBM. Should that be another category?? As noted in the the opening paragraph EBM is about the the judicious use of the current best evidence, not about refusal to accept any evidence that isn't absolutely positively perfect in every way.

I would also like some references in the last paragraph. my feeling is that such denials have usually been overturned on appeal.

129.255.102.104 15:17, 9 February 2006 (UTC)bill-morris@uiowa.edu

Perhaps the section might be more accurately labeled "limitations" of EBM because no proponents ever advocated the EBM was anything more than a system for assisting the determination of what is best treatment for an individual patient. alteripse 22:38, 30 March 2006 (UTC)

Agreed -Jim Butler 04:56, 31 March 2006 (UTC)

Concur, we should change that title. Midgley 15:20, 31 March 2006 (UTC)

It looks like most of the criticism is about the overapplication of ebm. For great justice. 20:24, 7 April 2006 (UTC)

[edit] Reorganized and over added comments

To be honest I read this article and thought it was badly organized, but included some good info. I took out some stuff from the 'overview' and placed it in the Qualification section since it was about that, and added a pargraph of three types of EBM practice. --Mikerussell 20:27, 10 May 2006 (UTC)

[edit] make EBM point to disambig page?

Is there support for changing EBM from a redirect to Electronic Body Music to a disambig page that points to Electronic Body Music and Evidence-based medicine? I'm not sure what the technical/political issues are about making this kind of change.

Done. Gnusmas 07:29, 4 October 2006 (UTC)

[edit] In medicine, evidence can be confusing

http://www.usatoday.com/news/health/2006-10-15-medical-evidence-cover_x.htm Rita Rubin, USA TODAY 10.15.2006

[edit] Daniel Davies

I have once again removed the (alleged - no source given) quote from the Guardian columnist Daniel Davies because (a) I don't see why one among the many thousands of journalists who have made well-informed, fairly-well-informed and downright-misinformed comments on EBM should be given a special place in this article; and (b) what he is quoted as saying about EBM is largely simply wrong. The notion that the "cult of the 5% significance level" plays a key part in EBM is untrue. Gnusmas 08:22, 18 October 2006 (UTC)

[edit] Off label use

Where does off label use of prescription medicines fit into Evidence-Based Medicine ? Should it be mentioned in this article somewhere as an example either or EBM or of non EBM? Or could off label use be a form of alternative medicine ? Robert2957 15:36, 27 October 2006 (UTC)

OLU can be considered experimental, or can be a useful exploitation of other effects than those normal used. Many medicines have multiple effects. Those that are the primary intended effects are the ones we are hoping to use, while the undesirable effects are called side effects, and are usually of lesser strength than the main effect, and often not even noticed by users. OLU is a common practice in modern medicine. Whether one could call it EBM is another matter. Not all of modern medical practice is adequately evidence-based as yet, but they're working on it.

In some cases it can be used in ways that go beyond legitimate experimental use, and then it might be classified as "alternative medicine." This use of the word alternative (as is the case with all of "alternative" medicine) is misleading, since a true alternative is a legitimate and logical choice among other effective choices. An alternate choice of an ineffective or even dangerous method or product is not a true alternative. This is one of the basic criticisms leveled against alternative medicine. For more on the subject of definitions, try the bottom of this page, and this one. -- Fyslee 18:00, 27 October 2006 (UTC)

Dear Fyslee,

Thank you for your reply. When you say PLU above, do you mean OLU ? Robert2957 20:28, 27 October 2006 (UTC)

Yes, that was a typo. Now corrected. Thanks! -- Fyslee 20:34, 27 October 2006 (UTC)

In my understanding, off-label is an administrative definition, not an EBM definition. On-label uses are EBM, but off-label uses may or may not be EBM.

In the US, at least, the "label" refers to the US FDA-approved package insert that comes with the drug and is an extension of the label.

The claims on the label for a particular use have been approved by the FDA, on the basis of clinical trials, and the drug company can only market the drug for those claims.

One or more clnicians, with or without the cooperation of the drug company, can do a clinical trial for a different use, and publish those results, and get high-quality evidence for a *different* use. That would be off-label, but it would still be EBM. The drug company could not advertise that use, however, unless they submitted that additional use to the FDA for approval. Since FDA approval is an expensive and time-consuming process, they often do not (especially with a drug whose patent will have expired by the time the FDA approves a new use).

Alternately, a clinician could appropriately use the drug off-label without high-quality evidence (which is common in e.g. oncology, especially in treating a disease that is too rare for RCTs to be practical).

Alternately, a clinican could inappropriately use the drug off-label for an unproven or inappropriate indication, such as testosterone or DHEA as an anti-aging drug.

(This is an excellent entry, BTW.) Nbauman 15:33, 10 November 2006 (UTC)

Thanks Nbauman for that excellent explanation. Feel free to contribute. -- Fyslee 19:30, 10 November 2006 (UTC)

[edit] NNT of antibiotics for Helicobacter pyroli

For example, an NNT of 4 means if 4 patients are treated, only one would respond. An NNT of 1 is the most effective and means each patient treated responds, e.g., in comparing antibiotics with placebo in the eradication of Helicobacter pylori.

I have no knowledge in this area, but to an amateur, this sounds suspicious. Every treated patient with Helicobacter pyroli responds to antibiotics? Is that correctly interpreted? I'd love a reference there. Also, I don't really understand how a NNT of 1, as explained, is used to compare antibiotics with placebo. I'd assume there would be one number for antibiotics and another for placebo. Can someone clarify this? /skagedal^... 15:51, 19 December 2006 (UTC)

It's sadly nothing so interesting. NNT is the number of patients you need to treat to get an effect in 1 person on average. In this case, I suspect the author was making up random numbers to demonstrate what NNT was. In this case, with a NNT of 4, that would imply that antibiotic treatment had a 25% chance of effectiveness. It's basically a different way of displaying data to patients (eg, a patient who is bad at statistics might not understand a 25% chance of effectiveness, but they may understand "imagine there are 4 people who take the medication, this will have an effect in 1 person”). It could also be used to see whether the cost of the chance of getting ulcers in the 1 patient the medicine would have been effective in is worth the cost of treating 4 people.

The author next writes that a NNT of 1 implies everyone is helped by the intervention. That said, I don't think this is the "best" result, since I could conceive a medication potentially treating one person and effecting more than one person (perhaps by preventing that one person transmitting it), although this I'm not too sure about.

As for one number for antibiotics and one number for placebo, I'm not sure about this. I think that NNT doesn't take into account the placebo effect, and refers to treating people in a clinical setting.

Anyway, that's just my 2 cents 60.242.64.202 14:29, 12 April 2007 (UTC)

[edit] EBM nuances, criticism and more shades

Hi people,

I just ran accross a page which could be used to add to the criticism section: http://www.holistichealthtopics.com/HMG/quack.html

+ History EBM: term from 1990, investigators from McMaster's University began using the term during the 1990s defined in 1996. http://cat.inist.fr/?aModele=afficheN&cpsidt=16960172 Ben Meijer 22:48, 10 January 2007 (UTC)

[edit] Over-claim in definition?

Evidence-based medicine (EBM) applies the scientific method to medical practice.

. Surely this is an overclaim which nobody will be happy with. At best, EBM attempts to more uniformly apply the standards of the scientific method (provided we could all agree on just what the scientific method was...) to help in some aspects of medical decision-making. But of course we all know that even if the "scientific method" was available to supply all the evidence we liked of efficacy and "risk" (risk of death? Risk of disfigurement? Risk of losing 5% of your retirement money??), there would still be a HUGE amount of "medical practice" which wouldn't be impacted at all, because it's based on totally appropriate but scientifically intangible values like: social cost, resource allocation, patient resoures, patient and doctor individual values (including patient and doctor tolerance for pain, boredom, frustration, criticism, cosmetic problems, etc etc), social mores (ethical problems in general), changes in fashion and politics and legal regulation, and endless other factors. Remember, science at best tells what you can do, not what you should do. Medical practice is very often about what people WANT. Science AT BEST only tells them what they can perhaps GET, at what COST (in money and risk). So there's always that gap in any praxis (technology, medicine, engineering, art) that needs to be filled from philosophy and practicality. EBM won't fix that, and I don't think it pretends to. But the first paragraph of this article is written as if it did. S B H arris 23:27, 10 January 2007 (UTC)

You know, you are absolutely right. That is a terrible introductory sentence. It is incaccurate, as scientific method has almost nothing to do with EBM. EBM is the application of scientifically researched evidence to the important questions about the objectively measurable benefits and risks of various medical practices. EBM does not pretend to provide value judgements of subjective risks and benefits nor does it deny their importance. So why not change it? alteripse 01:08, 11 January 2007 (UTC)

Agreed!Felix-felix 08:28, 11 January 2007 (UTC)

I'll give it a shot. Just wanted to lay out rationale here before making such a radical move. S B H arris 08:35, 11 January 2007 (UTC)

[edit] External links needs clean-up

The section is a mess, with too many links, and of varying quality. It needs to be boiled down to just a few links, or create groups of links with subheadings. -- Fyslee/talk 14:21, 8 April 2007 (UTC)

[edit] Criticisms section

This is an unreferenced rant, that makes claims that are neither sourced, nor demonstrated to have any relationship to reality, e.g. the suggestion that EBM guidelines are used to deny better drugs. I have no objections to criticism, but they have to be sourced, and any appropriate rebuttals need included. This section fails on both counts. Adam Cuerden ^talk 15:26, 13 April 2007 (UTC)

Some references have been added. And not all the rants are here, because some of them are too difficult to explain. Example: Although EBM has been criticised as being econometrically based, it isn't. However, one of the main problems with use of EBM to set national health policies is that too often EBM is in bed with econometric people, and that fact screws up drug development in subtle and no-so-subtle ways that are ultimately not good for patients.

For example, suppose drug companies have new and very expensive drugs which are active against tumors in animals. What groups of humans do you test them in, first? Well, if you test in late-stage cancer patients, your placebo or non-active or best-previous-treatment group's going to be dead in two years, so you can get EBM-quality data, statistically valid, very fast. So if you test in late stage patients, you can make money fastest getting the data to bring said expensive drug to market. So you focus your experimentation there, and meanwhile the patients who may really be cured by your drug (those with early stage cancers who need it as adjuvant, where it may help as a real cure) are left out in the cold, unable to get it at any price.

Two years later you have your answer: your drugs don't save any lives in late stage patients, but lengthen them by 4 or 5 months at a cost of $20,000 or more. Now, EBM helps you get rich in rich countries. And (at the same time) it keeps the drug off the market entirely in poorer ones, where the government wonks decide econometrically, using the best EBM evidence, that that little life isn't worth that much money. The drug cures people in neither types of countries.

Meanwhile, the early-stage patients who might really have been helped, get screwed in BOTH kinds of countries, because EBM has helped the system not pay any attention to them, because they have the bad luck to be slow experimental animals. All this happens very scientifically, because EBM has inhibited the most rational and potentially helpful experimentation, due to the fact that the drug makers need EBM results, quick, for their cash flow. Which is a real thing, and not some evil enterprise. So here, the existance and use of EBM hasn't helped, and in fact has hindered, clinically significant progress in drug development. If you need a concrete example, check out NICE's recommendation that the Brit NHS not pay for the cancer drugs Avastin and Erbitux for stage IV cancer, in 2006. And because of all the stage IV tests, the stage II and III tests have gone lacking for funds, and are still collecting subject/patients. Go figure. [4][5] [6] S B H arris 01:41, 29 April 2007 (UTC)

[edit] Inductive quality?

What does "inductive quality" mean in the introduction? I couldn't find it in Steadman's.

The introductory paragraph in Wikipedia is supposed to be written for a layman, but I can't figure out what this means. (And maybe I'm not a layman; I read Science, NEJM and BMJ every week.) Nbauman 23:58, 28 April 2007 (UTC)

Inductive quality is the quality of inductive evidence. If a drug works on 50 year olds, will it work on 40 year olds? 20 year olds? 10 year olds? Toddlers? Neonates? Much evidence of many kinds goes into guessing at the answer to that, and it's all induction, especially if the exact study hasn't been done. For a a new experimental antibiotic, the odds may be better than new experimental organ transplant procedure. In fact, it's induction even if a very close study has been done, since one 50 year old isn't exactly in all ways like another. I'll see what I can do to fix up the LEAD. S B H arris 01:09, 29 April 2007 (UTC)

I just searched BMJ and PubMed for the phrase "inductive quality" and "inductive evidence", and I didn't get any hits. The MeSH headings for "inductive" refer to something completely different. As I said, I couldn't find it in Steadman's Medical Dictionary either. It doesn't seem to be a standard medical term. It may be a term from a different discipline, such as philosophy. Can you give me a source for the definition of this term? Can you give me an example of its use in the medical literature? Nbauman 03:01, 29 April 2007 (UTC)

You get more hits on pubMed from "inductive logic". Medicine usually comes up against inductive logic from the viewpoint of "Bayes' theorem" and "Baysian reasoning," [PMID 11206199] both of which turn up many hits. For example [PMID 15338074]. But this stuff is often quantitative. Most of the real-life problems a physician faces are not quantitative, because the input parameters (the sensitivity and specificity of a fact in a given situation) are not known. The world is too complex. If Lipitor decreases heart attacks and overall mortality in 50 year old men with known CAD, will it do so in 50 year old men with no medical history at all? 50 year old male smokers? What about 50 year old women smokers? 50 year old women with high cholesterol only? We can take guesses and they are better guesses than if we had no data at all about Lipitor, but it's all an inductive exercise. It depends a lot on premises. Are women's coronaries like men's? If so, how much are they? Do women tolerate high LDL as well as men do? Or does LDL hurt them only if above a threshhold (and what is that number?), or they have other risk factors like smoking and hypertension? We don't really know. Should be we correct hormones in women after menopause? Induction is how we guess yea or nay, as the evidence comes in, before the ultimate large controlled blinded randomized clinical trial. And even then, we only know about the group we treated, and only about the exact treatment we used. Depart from that, and you're back to induction. EBM uses induction. But induction is not a cut and dried process. There's no formula for it, due to the assumptions that go into Baysian logic, many of them mechanical. Will a pill that decreases PVD's decrease the incidence of V-tach and V-fib? Or is that just in people with CAD and other risk factors? Place your bets. If you get it wrong, you kill people. S B H arris 03:34, 29 April 2007 (UTC)

If "inductive evidence" is a term that is not generally used in medicine -- never in Stedman's, PubMed or BMJ -- should we use it in the lead? Nbauman 04:35, 29 April 2007 (UTC)

The phrase "inductive evidence" seems to violate the rule against neologisms. Unless you can provide a reliable source, it has to go. Nbauman 14:52, 29 April 2007 (UTC)

For Heavensake, just because YOU have never heard of it, doesn't mean it's a neologism. Entering inductive evidence gets a million hits on Wikipedia, many of them with the words used together (in fact, about 20,000 of them are, as you can ascertain by entering "inductive evidence" complete with quotes in GOOGLE, so it is searched for as a phrase).

*I* didn't particularly mean it as a special phrase, or I would have hyperlinked it. I'm using "inductive" as a modifier of "evidence" since in fact nearly all the evidence, and argument, provided by EBM *is* of the inductive type. Evidence only COMES in two varieties: inductive and deductive. Evidence about what will happen in the future based on what has happened in the past, is inductive by definition. That's what EBM is about. You mentioned that you are a layman. Fine. I'm sorry that you've missed a major part of the underlying theory of science, and all scientific disciplines, including this one. But those of us who work with it everyday, don't have this problem. I personally think that nobody who is in charge of thinking about the nature of evidence and what it means, should not have a basic grounding in logical terms, and should certain know what "induction" and "deduction" mean (enough to know that what Sherlock Holmes called "deduction" was actually "induction"!

Reasoning from specific observations toward general conclusions is induction. Evidence that something will happen based upon that is inductive evidence. The amount and type of this kind of thing is inductive evidence. Educated people should know this. So take this as an opportunity to learn something new, rather than to decide you want to do something reactionary. I promise you that every single historical constructor of the ideas of evidence-based medicine, is fully aware of what "inductive evidence" is and means. "Neologism," my rear end. With 20,000 GOOGLE hits, are you maybe joking here? Don't waste our time, if so! S B H arris 21:27, 29 April 2007 (UTC)

There are 2 issues here: (1) Whether the term "inductive evidence" is a neologism as defined by WP:neo and (2) Whether a general reader, as defined by WP:Audience, could understand the term. I could accept it if it were further down in the article, but not in the lead.

(1) Neologism. When you say,

I* didn't particularly mean it as a special phrase, or I would have hyperlinked it. I'm using "inductive" as a modifier of "evidence" since in fact nearly all the evidence, and argument, provided by EBM *is* of the inductive type.

it sounds like a violation of WP:NEO#Reliable_sources_for_neologisms:

"An editor's personal observations and research (e.g. finding blogs and books that use the term) are insufficient to support use of (or articles on) neologisms because this is analysis and synthesis of primary source material (which is explicitly prohibited by the original research policy)."

because you yourself are linking ideas, rather than citing somebody else who linked those ideas.

I know that if you search Google for "inductive evidence" you get 2,300 hits. That's the first place I looked. But in the context of medicine, I couldn't find anything that explained it. It's not a term that doctors use in their literature. It's not in Steadman's, the BMJ, or PubMed. It's not a term that doctors normally use. It's not a term that laymen would understand.

One of the Google hits I got was in a NEJM book review using it as an example of unclear language:

[7]339:353-354, July 30, 1998

To See with a Better Eye: A life of R .T.H. Laennec

By Jacalyn Duffin. 453 pp. Princeton, N.J., Princeton University Press, 1998. $49.50.

...Finally, her style of writing is not characterized by simplicity and clarity — for example, "As in his formulation of pulmonary signs, the decision-making criteria he imposed on this ensemble of inductive evidence expressed an inkling of statistical probability, typifying one whose existence straddled the sensualist philosophy of the late eighteenth century and the dawn of positivist thought." [Italics added]

(2) Audience. I didn't say that I was a layman. I said that I'm not a layman -- I read NEJM, BMJ and Science every week. I've written reports and news stories for doctors about PSA tests from presentations at urology and cancer conferences, and I regularly write about randomized controlled studies, so I know about Baysean statistics. (I've also written about Lipitor.) But Wikipedia is written for the general reader:

Wikipedia:Manual_of_Style_(medicine-related_articles)#Audience

Wikipedia is written for the general reader. It is an encyclopaedia, not a comprehensive medical or pharmaceutical resource or first-aid manual. While healthcare professionals and patients may find much of interest, they are not the target audience.

If I can't understand it, then a fortiori a general reader can't understand it. And I can't understand it. (Or rather, from my current understanding of "inductive evidence", it seems trivial. You're just saying, "EBM uses inductive evidence to make rational decisions." What am I missing?)

It's especially important for the introduction to be clear. I think you're loading too many ideas into the introduction. If you feel this idea is important enough to include in the article, then you should (preferably) describe it clearly in layman's language, and place it further down in the entry.

Can you explain to me what that sentence, "EBM seeks to apply judgments about the quality of inductive evidence, to those aspects of medicine that depend on rational assessments of risks and benefits of treatments," means, in language that a general reader would understand? Can you rephrase it in a sentence?

Can anybody else give me a second opinion on this? Nbauman 23:44, 29 April 2007 (UTC)

I found 132 papers on PubMed which use the phrase "inductive reasoning". So doctors don't talk about this? I'll be glad to change the phrase inductive evidence to inductive reasoning, if that will make you happy. But according to you, it won't help your understanding. My Google must be different from your Google, because the phrase "inductive evidence" really does hit 19,800 times on mine, and if you add medicine without quotes to that phrase (i.e. search: "inductive evidence" medicine you get 701 hits (some duplicates, but still very many), and if you read the first four or five of them, perhaps we wouldn't be having this conversation, since they're mostly about inductive inference in Evidence Based Medicine. Which (EBM) you can't discuss without coming to grips with the different kinds, or levels, of evidence [8][9][10] [11] [12]
If it works in a rat will it work in a human? If it worked in 5 other humans will it work for you? How do you know? How about if it works in 2000 other humans? Where does your extra confidence come from? Why more in one case than in another?

You say you're not a "layman". That would make you a professional, then, by definition. This is a conversation about logic and medicine. So which of these are you a professional in? Present credentials, please. Reading a lot does not make you a professional. Reading a lot does not save you from being a layman. Only doing it for a living does that. Usually that requires a degree, but not always.

Not all articles in wikipedia are for the general reader. Wikipedia has large sections on mathematics and mathematical physics. There's no help for it. Read general relativity

Can you explain to me what that sentence, "EBM seeks to apply judgments about the quality of inductive evidence, to those aspects of medicine that depend on rational assessments of risks and benefits of treatments," means, in language that a general reader would understand? Can you rephrase it in a sentence?

You can simply remove the word "inductive" if you don't know what it means. It will become clear in the article later that EBM grades evidence on quality, and that the difficulties in doing this stem mainly from the problem of induction. Probabilities of future events are always based on inductive processes, reasoning inferentially in tricky ways, from past events. The very idea of inference from past experience refers to an inductive process. There's no getting around understanding the idea of performing inductive reasoning (no matter what you call it) before you go anywhere near EBM. I put it in the LEAD to give the reader heads-up on what's coming. Not every word in a LEAD has to be understandable immediately. The LEAD can partly be a warning of what concepts are going to be needed later. Feel free to put hyperlink brackets around inductive, to show that. S B H arris 00:55, 30 April 2007 (UTC)

"Not every word in a LEAD has to be understandable immediately."

I disagree. When I worked at McGraw-Hill, my editors told me that if your reader can't understand the lead, they'll never read the rest of your story. (They used to hire consultants to observe magazine readers as they read and do scientific studies of reading behavior.) When I read an article, I want to understand what I'm reading as I go along. If I can't understand the first paragraph, I'll stop and try to figure it out. If I can't figure it out, I may or may not read the rest of the article. Most readers don't. Sometimes readers take a concept they don't understand and "put it in a box" until the rest of the article comes together, but that's a hard way to read and makes it difficult to get your point across.

But it doesn't matter what I think. We have rules here. Wikipedia:Manual_of_Style_(medicine-related_articles)#Audience: "Wikipedia is written for the general reader. " If the general reader can't understand it, you can't put it in. I'm at least as dumb as the general reader, and I can't understand it. If you want to make a point about inductive logic, you have to first prepare your reader by explaining the terms you're going to use, and then make the point. You can't make a point with terms that your reader doesn't understand.

I think evidence-based medicine is very important, for a lot of people. Who are we writing this for? Well, one of the people we're writing for is a cancer patient who is being offered quack cures, or research still in vitro or in animals like Dichloroacetic acid) who should understand how to evaluate those treatments, and understand why his or her doctors advise against it. That person should be able to understand as much of this article as possible, and certainly the lead, without struggling or misinterpreting it.

I think you should take it out of the lead, and discuss it further down. And start out with a sentence that everybody can understand. —The preceding unsigned comment was added by Nbauman (talk • contribs) 19:47, 30 April 2007 (UTC).

Okay, I've taken the word "inductive" out of the LEAD. Remove all hard words, lest they cause fear. Theeeere you go. S B H arris 23:55, 30 April 2007 (UTC)

I'll examine this more carefully when I get done reading the rhesus monkey genome. Nbauman 19:02, 1 May 2007 (UTC)

[edit] Wikipedia sucks on science?

OK, here's the problem with this entry:

Epidemix

Why Does Wikipedia Suck on Science?

"on science, there’s a oneupmanship going on, and a topic will be honed to an ever-greater level of expertise. That’s great for precision and depth, but horrible for the general user, who is often brought to Wikipedia through a top hit on Google."

Nbauman 05:06, 13 May 2007 (UTC)

Let me disagree here: IMO this article, concise but fairly complete, balanced but not in overly confrontation-averse way, with interesting, non-trivial and civilized talk page, is an example of value of Wikipedia. --bonzi 16:59, 25 July 2007 (UTC)

Answer: LOL, that's one of a very many reasons, and not one of the big ones. Wikipedia does remarkably well on science, given the many constaints it labors under. For an example, here's a quote from somebody who is aguing for a more historical (but also more convoluted and confusing) account of mass in special relativity, on TALK:photon:

Oh, and by the way, Wikipedia is not for learning, though it's a valuable learning tool. It is a reference work, a place to look things up. This is a subtle but important distinction. If you want to lead readers through your preferred way of learning something, rather than simply presenting the facts, the place to do it is Wikibooks. --Trovatore 21:50, 9 May 2007 (UTC)

This is from a guy arguing that since Wikipedia isn't a textbook, there's no particular reason why it should be as accessible as one, for the poor tyro reading it. And if its editors take that attitude, it's amazing that Wikis on science topics suck as little as they do.

For my part, I merely observe that when Wikipedia decides what it wants to be when it grows up, it will suck a lot less. S B H arris 05:43, 13 May 2007 (UTC)

I don't know if WP will ever grow up. Whatever it becomes will be the result of anarchist changes, like people making introductory paragraphs more readable and understandable to the general reader.

BTW, this is from the nutshell section of the newly adopted guideline, Wikipedia:Manual of Style (medicine-related articles):

Write for the average reader and a general audience — not professionals or patients.

Explain medical jargon or use plain English instead if possible.

Nbauman 19:52, 13 May 2007 (UTC)

Excuse me? Weren't you the man arguing that the word "inductive" (as in "inductive quality of evidence," which you seemed to find completely understandable without the use of the first word) was a pharse that was bad precisely because it was NOT medical jargon, and wasn't used in medicine? You've sort of gotten yourself completely turned around in this debate, and are now taking the other side. For a reason nearly opposite to the one you first gave. Methinks you need to look into your own motives a bit.
As for "inductive", as applied to logic it IS "plain English". "Inductive logic" and "inductive evidence" both return more than a million hits on a google search. Any phrase used that commonly is common enough to be used in a LEAD paragraph, especially if linked, without having to explain it right away. On a google search it's about as common as "pumpkin pie," and more common than "key lime pie." And way more common than "mincemeat pie." Er, which of these do you think need to be explained to the average English speaker? "Inductive logic" returns about the same number of hits on Google as the phrase "value of life," (don't forget to use the quotes) which you allowed to remain in the LEAD without objection. I suppose you feel that doesn't need explaining right away? How is "value of life" less medical jargon than plain English? Are we all agreed on what it means? Or did you like it in this case because it WAS medical jargon? S B H arris 20:41, 13 May 2007 (UTC)

No, that's not what I said. I said that "inductive evidence" was a term that wasn't used in medicine, or in ordinary, layman's non-technical English. It seems to have a technical meaning in philosophy which I don't understand.

I still don't understand the meaning of the following sentence:

Specifically, EBM seeks to apply judgments about the quality of evidence [PMID 15338074], to those aspects of medicine that depend on rational assessments of risks and benefits of treatments (including lack of treatment)

Could you explain to me in simple English what that means? Nbauman 00:37, 14 May 2007 (UTC)

[edit] This article is far from done

I have provided a structure of some topics that I think are lacking. We should define what criteria constitute a good study. How do critical thinkers determine whether to implement an intervention? What are some examples of helpful and unhelpful clinical articles? Anyone care to join in?--66.74.75.39 06:52, 20 August 2007 (UTC)

[edit] Criticism

In this article (in norwegian) a doctor complains that evidence-based medicine can often be used to dodge making a proper theory and hypothesis, and as such gain support for an implied hypothesis that is inconsistent with theory. I'm a bit suprised the article doesn't mention this, as it's the criticism I've come across the most. —Preceding unsigned comment added by 83.243.152.134 (talk) 11:42, August 29, 2007 (UTC)

[edit] Evidence-Based Medicine vs. Evidence-Based Practice

Could someone explain how "evidence-based medicine" and "evidence-based practice" differ? They both have Wikipedia entries, unlinked to one another. Seems like they are essentially the same, however.

Hogvapor (talk) 22:16, 29 November 2007 (UTC)Hogvapor

EBM is the specialization of EBP to applications in medical treatment of patients with somatic or psychiatric disorders. EBP is more the more general application of the same model for transforming descriptive information (the scientific study or theory derived therefrom) into a prescriptive policy. Prescriptive policies are common professional tools in politics, management, sociology, finance, civil engineering, corrections, natural resource management, etc. Aminorex (talk) 04:40, 2 March 2008 (UTC)

[edit] Possible quote for inclusion

Would this quote be useful in the article?

A research finding is less likely to be true when the studies conducted in a field are smaller; when effect sizes are smaller; when there is a greater number and lesser preselection of tested relationships; where there is greater flexibility in designs, definitions, outcomes, and analytical modes; when there is greater financial and other interest and prejudice; and when more teams are involved in a scientific field in chase of statistical significance. Simulations show that for most study designs and settings, it is more likely for a research claim to be false than true.^[1]

(The author then goes on to explain that meta-analyses are necessary to get to the heart of the matter.)

regards, Jim Butler^(talk) 04:25, 28 January 2008 (UTC)

[edit] References

Please start new sections above this one.

^ Ioannidis, John P. A. (2005). "Why Most Published Research Findings Are False". PLoS Med 2 (8). doi:10.1371/journal.pmed.0020124.

[edit] External links

Hu12 deleted our list of external links and replaced it with a link to Evidence-based medicine at the Open Directory Project instead. I reverted it, and added the link to The Open Directory at the bottom.

Our list is a short list of well-selected articles, which I've used and found very helpful. The Open Directory, while also useful, is a huge list of 117 links (most of them sorted alphabetically by title), which for me was so long that it was less useful. Most significantly, our list gave bibliographic citations, while the Open Directory did not. I like to look down the list of references for journals I'm familiar with (and subscribe to), like BMJ. I can't do that with The Open Directory.

WP:NOT doesn't apply here. The editor seemed to be thinking of WP:NOT#LINK, which doesn't apply here; this is not a "mere" collection of external links. WP:SPAMHOLE doesn't apply here; there are no commercial links. Nbauman (talk) 17:35, 5 February 2008 (UTC)

WP:NOT applies everywhere on the project. Wikipedia is an encyclopedia (Not a repository of links). "There is nothing wrong with adding one or more useful content-relevant links to an article; however, excessive lists can dwarf articles and detract from the purpose of Wikipedia." Links should be restricted to the most relevant and helpful. Long lists of links are not appropriate, commercial or otherwise. We don't need to link to every site in existence that meets a certain criterion. This is an encyclopedia, articles need cited contentent, not linkfarms--Hu12 (talk) 14:35, 7 February 2008 (UTC)

Removed services, search links, directories (except dmoz), regional specific links, links to content that duplicates whats already in the article and links which go to the root domain of a sites. Biggest issue is why we needed so many "what it is ..." type links in an encyclopedic article explaining evidence-based medicine .. --Hu12 (talk) 14:58, 7 February 2008 (UTC)

The links you deleted are not "what it is..." links. They're useful, major sources of evidence-based reviews, that librarians and doctors routinely recommend as one of the first places to go. I use them regularly. They give evidence-based reviews, and let people understand what EBM is in a way that can't be explained in the article.

Why, for example, did you delete the Cochrane Collaboration library? Nbauman (talk) 19:43, 7 February 2008 (UTC)

Hu12, I'm still waiting for your answer. Nbauman (talk) 17:04, 8 February 2008 (UTC)

Hu12, I'm sorry you haven't replied. I've reviewed the links you removed. While some of them were commercial content, others were clearly important, useful links by the WP critera you cited. For example, the Cochrane Collaboration and USPSTF are major, important resources that everybody in EBM knows about and uses regularly. Others are important resources and articles about EBM. Medical research has to be constantly updated, and DMOZ doesn't have the resources (or responsibility) to do that.

I'm replacing the important links. If you want to discuss them on a case-by-case basis, please do.

Important links:

Cochrane.org - 'The Cochrane Collaboration: The reliable source for evidence in healthcare' (systematic reviews of the effects of health care interventions), Cochrane Library Major source of rigorous EBM evaluations.
AHRQ.gov - 'U.S. Preventive Services Task Force (USPSTF)', Agency for Health Care Research and Quality. Major source of EBM evaluations
GPNoteBook.co.uk - 'Evidence-based medicine (EBM)', General Practice Notebook Free content
JR2.ox.ac.uk - 'Bandolier: Evidence-based thinking about health care', Bandolier (journal) Free reviews online
SHEF.ac.uk - 'Netting the Evidence: A ScHARR Introduction to Evidence Based Practice on the Internet' (resource directory), University of Sheffield Extensive bibliographies and links to online articles
TRIP Database - 'TRIP Database - EBM search engine' (resource directory), TRIP Knowledge Service. Free.
BMJ.com - 'Evidence based medicine: what it is and what it isn't: It's about integrating individual clinical expertise and the best external evidence', (editorial) British Medical Journal, vol 312, p 71-72 (January 13, 1996)
BMJ.com - 'Evidence based medicine: Socratic dissent', (Education and debate) British Medical Journal, vol 310, p 1126-1127 (April 29, 1995)
CEBM.net - Oxford Centre for Evidence-Based Medicine (UK) Some free content
EBOnCall.org - 'Evidence compendia' (evidence-based summaries of 38 on-call medical conditions), Evidence-Based On-Call (EBOC) Free
BMJ.BMJjournals.com - 'Parachute use to prevent death and major trauma related to gravitational challenge: systematic review of randomised controlled trials', Gordon C S Smith, Jill P Pell, British Medical Journal, Vol 327, pp 1459-1461 (20 December 2003) (Classic argument that situations still exist where RCTs are unnecessary.)

Keep deleted:

[13] - American College of Physicians Journal Club offers evidence-based reviews of recent literature Subscription only
BEEMCourse.com - 'The Best Evidence In Emergency Medicine (BEEM) Team' Group based out of McMaster University (Hamilton, Ontario, Canada) that actively reviews the best emerging evidence based research in emergency medicine. Advertisement for course
EvidenceMatters.com - Personalized treatment outcomes from peer-reviewed research, in a database and search-engine format. Subscription required for full information
Evidence-based medicine, What is ? - The What is...? series explains the key concepts in a clear, concise and accessible format. Apparently a commercial company
Evidence-based Decision Making in Critical Care - On-line journal club and reference site. Critical care focus. not much free content
HERTS.ac.uk - Evidence-Based Medicine', University of Hertfordshire Just a home page
[14] - The Medical Letter of Drugs and Therapeutics, a leading resource of unbiased drug information Subscription required

Nbauman (talk) 14:30, 9 February 2008 (UTC)