The evidence based approach is not a panacea for all medical decision making.
Some clinical practices have become established without "evidence" e.g. the use of intravenous diuretics in acute pulmonary oedema has evolved without any formal randomised controlled trial. In such circumstances an application for a trial comparing such an intervention with a placebo would be unethical. Ignoring obviously successful interventions because of lack of evidence would be taking the EBM paradigm to ridiculous extremes.
Also collating "evidence" for uncommon diseases is difficult and, despite the use of tools such as meta-analysis, it may not be possible to adequately analyse "evidence" relating to aspects of uncommon diseases.
Evidence based medicine is based on data on defined populations of patients. Often criteria for entry into particular trials are, necessarily, rigid. Clinical trials can only analyse a particular scenario of those that may exist in clinical practice. It would be impossible to have trials relating to interventions with every possible permutation of clinical populations in which an intervention may be applicable. Therefore conclusions relating to trial populations are induced to also be true for other populations that require similar intervention. A patient may suffer multiple pathologies and be unlike any patients in particular studies on which "evidence" is based. Is the "evidence" valid for that particular patient? There is no means to know the answer, only of inducing an answer.
Therefore though "evidence" is very important one must recognise the limitations of adopting a puritan scientific approach. Medicine is art and science. Decisions must consider the "evidence" (or lack of "evidence") in the context of individual patients.
Annotations allow you to add information to this page that would be handy to have on hand during a consultation. E.g. a website or number. This information will always show when you visit this page.