Evidence vs Expert Opinion
A central thesis of EBM is the putative superiority of scientific evidence over opinion. The truth is that both evidence and opinion have their limitations. As already noted, scientific evidence is absent for much of medicine and, when available, often lacks internal and external validity.
The problems with opinion, whether it be clinical judgment or the beliefs of experts, are also well documented.  These include the selective use of evidence (inadvertently or consciously ignoring studies suggesting another view), biases about magnitudes of effect and appropriateness that stem from personal experience (eg, how one was trained, a notably bad outcome in a past patient), flawed assumptions about the frequency or natural history of diseases, and external influences (eg, professional norms, business pressures, patient expectations, medicolegal concerns). 
Medical history is replete with examples of interventions that were thought to be beneficial and that received enthusiastic support from clinicians, even in the face of emerging evidence to the contrary. The tendency of physicians to assume that a treatment is beneficial -- based on biological plausibility, prevailing theories about natural history, and intermediate (rather than health) outcomes -- has inspired many mistakes in medical dogma. For example, cardiologists have long believed that routine antiarrhythmic therapy is beneficial or even life-saving in patients with acute myocardial infarction because of the evidence that arrhythmias are associated with increased mortality. This inference went unchallenged until the early 1990s, when clinical trials showed that antiarrhythmic therapy increased mortality. 
However, science is not the last word, nor the only consideration, in determining what is best for patients. Although science is of paramount importance, good clinical decisions and sound public policies address other considerations, such as expert opinion and clinical experience, patient expectations and preferences, social circumstances, the time, equipment, and personnel available in the clinical setting, community support services, access to care, ethics, insurance coverage, and medicolegal risks.
Cancer Control. 2000;7(4) © 2000 H. Lee Moffitt Cancer Center and Research Institute, Inc.
© Copyright by H. Lee Moffitt Cancer Center & Research Institute. All rights reserved.
Cite this: Evidence-Based Medicine and Practice Guidelines - An Overview - Medscape - Jul 01, 2000.