[Evidence-based medicine: modern scientific methods for determining usefulness]

For quite some time, clinical epidemiology has introduced the art of critical appraisal of evidence as well as the methods of how to design sound clinical studies and trials. Almost unnoticed by most medical institutions a new hierarchy of evidence has emerged which puts well thought out trials, able to document unbiased treatment benefit in terms of patient suffering, above pathophysiological theory. Many controlled trials have shown, in the meantime, that the control of laboratory or other kind of pathologies and the correction of anatomical abnormalities do not necessarily mean a benefit for the patient. Concepts relating to this dissection of evidence include: Surrogate fallacy ("cosmetics" of laboratory results or ligament or cartilage "cosmetics" in surgery), confounding (spurious causal relationships), selection bias (comparison with selected groups) as well as lead-time bias (mistaking earlier diagnosis as increase of survival), length bias (overlooking differences in the aggressiveness of diseases as determinants of disease stage distributions) and overdiagnosis bias (mistaking the increasing detection of clinically silent pathologies as improvement of prognosis). Moreover, absolute instead of relative risk reduction needs to be used to measure patient benefit. The incorporation of decision-analysis and of the concepts or clinical epidemiology will improve the efficiency and quality of medicine much more effectively than the sole focus on technical medical performance. Evidence based medicine is the systematic and critical appraisal of medical interventions, based on the understanding how to avoid the fallacies and biases mentioned.