It's best decision making to follow the evidence when it is there, is of high quality, scientific, peer-reviewed, with time for confirmation.

The evidence-based practice movement consists of a variety of efforts to bring into widespread use, programs, policies, and other practices already shown by experimental evaluation to affect psychological or behavioral functioning. Beginning in the 1960s, psychologists and educational researchers began to report treatment procedures that were demonstrably more effective than the procedures with which they were compared.

What is best possible evidence is also not absolutely clear. Many meta-analyses suffer from data being biased (e.g., publication bias) that tends to overestimate effects. But so do individual studies. Because of this, I think EBM needs to be based on well-powered pre-registered studies with heterogenous samples. This is the best way, I think, to evaluate effects and variability in effects.

Another issue is the question of whether a study is applicable to the individual case you are working with. This is referred to as the general-to-individual problem. Sure, group-level estimates provide a good starting point (as long as the group-level data at least to some degree pertains to your case), but since there is always considerable variation, a clinician need to also estimate the effect of a procedure on the case being treated. Averages are averages and in most studies, the effect for most individuals is far from the average estimate). This means that wile many will respond well to an EBM treatment, you will still have individuals who do not or who get worse. Blindly relying on the average effect in this cases is simply detrimental.

Because of this, I think the main importance of the EBM paradigm is to decrease the amount of treatments that completely lack basis in the scientific literature.

I think it cut off the end of my question! I am really interested in how the EBM has been interpreted in practice rather than being interested in the theory. It seems that two of the three domains of evidence (client value and needs and clinician experience) are overpowered by the published scientific gold standard EVIDENCE. What do you think?!

I believe that ideally patients facing a clinical decision should enter into a shared decision-making process. The clinician should enable the patient to learn about and understand the relevant evidence which should come either from clinical trials or epidemiologic studies. The patient should understand the limitations of the evidence. Then, the patient's values and preferences should come into play. The physician may want to present his/her clinical experience, but a) it is not evidence, and b) it makes sense primarily when there is no good evidence, or conflicting evidence, in the literature.

Here is a link to just one piece of work by John (Jack) Wennberg and his colleagues about the unjustifiable variation in use of clinical services that occurs when physician preference dominates clinical decision making: https://academic.oup.com/jnci/article/95/8/570/2520702

Thanks all. So it seems that heavy reliance on gold standard evidence is considered to be the best option overall. What about in community services/development programs where a counterfactual is difficult and potentially unethical and there are significant contextual variations? Additionally, what about co-morbidities/co-occurring issues/challenges etc that aren't included in experimental studies?

Ms. Kelly, I have thought for many years that while it is wonderful when a "gold standard" clinical trial exists that can be used appropriately as evidence to present to a patient, it is also important to collect new forms of evidence to address some of the situations you mention. That is what led to the editorial I wrote about 25 years ago and that I'm adding to this entry. Even 25 years ago, it was possible in certain places to develop some databases along the lines I suggested. Today, with much more extensive use of computers in health care it is certainly possible; but unfortunately, I'm not aware that it is being done.

I undertand it in simple way as a practice where our decsion is based on evidence whether it is related to Clinical /laboratory /field/Community/developmental work relate decision making. The another issue may be how the evidence is generated?. In my opinion it should be based on high quality. peer reveiwed and high impact publication.

I think that both your questions focus on determining applicability of generalized evidence to one patient or situation, which is the crux of clinical application of EBM. While occasionally studies identify characteristics that change disease risk (eg. women have lower risk of abdominal aortic aneurysm) or therapy utility (eg. lower efficacy of ACE inhibitors in blacks), our knowledge is usually limited. That is why the concepts of likelihood ratios and numbers needed to treat/harm are so valuable to estimate effect. Then we can gamble with this one patient (using shared decision making), with an idea of our odds. Despite inherent errors, it’s usually better than our fallback of heuristics based on easily-recalled, somewhat-similar patients. I believe that EBM is best defined as a questioning of cognitive shortcuts using evidence. At some point, the evidence can be so poor that we would be better off with the heuristic. That’s why we look for gold standards and are so disheartened by bias and misconduct.

Where doctors treat sick people, evidence based medicine is used. Obstetrics is another story. 99% of birth now takes place in hospitals, and hospital setting has never been shown to improve outcomes. Despite this, nearly all published obstetric research in the past 70 years is done in hospitals, a non-evidence based setting. In addition to the setting being non-evidence based, the interventions done to women are not evidence based. There are no evidence based protocols performed on low risk women in hospital. Hospitalizing women at birth is dangerous, resulting in more maternal infections and deaths. Breaking the sac is dangerous, yet done routinely. Episiotomy is simply genital mutilation, because it has never been shown in a single study to improve outcomes. There is no evidence to induce for postdates or do a cesarean after the woman has been pushing for 2 hours. Hemorrhaging only occurs in human birth, no other animal bleeds more than a spoonful, even animals who have the same placenta, such as chimpanzees and great apes. So when it comes to obstetrics, there are no evidence based practices used at normal births. If this interests you, you might be interested in some of the following:

Article To what extent do English language RCT meta-analysis justify...

https://www.researchgate.net/publication/317350722_Prolonged_second_stage_Three_cases_14_17_and_24_hours_long_with_perfect_outcomes

Article Instead of attacking home birth, hospital practitioners shou...

Presentation Critique of Grobman etal. and the ARRIVE RCT to induce birth...

Article Minimalizing bleeding and tearing at vaginal birth: One prac...

Leanne - great question!

I'm not a clinician so perhaps can't offer you the insight you require. However, I highly recommend the recently published book by one of the top EBM researchers in the world - Professor Trisha Greenhalgh. Her book discusses in detail some of the challenges and complexities in implementing EBM in practice (as she is also a primary care physician).

Book details: How to implement evidence based healthcare by Trisha Greenhalgh (2018) published by Wiley Blackwell.

Thank you for the brilliant answers everyone! So interesting to get your perspectives and thank you for all the links to other references. I will check them all out. I hadn't seen that Trisha Greenhalgh recently released a book either Liesl so thanks for that. I have read a number of her articles and found her very enlightening.

If you are writing something about the evidence based medicine, i hope you are brave to compare it to the ANTI-VAX movement which is the opposite- and a prime example of how people get confused about that is evidence and what is based on Andrew Wakefield's greed and the greed of other interest groups and just crowd fearmongering.