Numerators Without Denominators

Abstract

EACH day, physicians face complex decisions in choosing accurate diagnoses and optimum treatments. We have become quite adept at marshaling objective data for diagnostic choices. Therapeutic decisions must often depend on less substantial evidence. In consequence, elaborate methods are required for clinical trials to test the effectiveness and safety of new drugs, but, somehow, the same type of rigorous criteria are not demanded for trials of new operations. Much thought and huge sums of money have been directed toward ensuring the outcome when the physician writes a prescription, but there is not this type of concern generated when he flourishes a scalpel. If this division of requirements were really proportionate, it might be justified as a backhanded acknowledgment that the pen is indeed mightier than the sword. Unfortunately, this is nottrue; there is no evidence that we need less demanding criteria for making surgical than medical choices. Comparing Treatments: The