Getting Doctors to Listen: Ethics and Outcomes Data in Context

PAUL J. EDELSON

Practice guidelines originate in the phenomenon of practice pattern variations described by Wennberg and his associates in a continuing series of studies beginning in the mid-1970s. ¹ These papers demonstrated substantial geographical variation in the aggregate use of physician ² and hospital ³ services and in specific utilization rates of common surgical ⁴ and radiologic ⁵ procedures. Such variations, occurring over relatively small geographic areas, have now been documented for hospital services in several countries, ⁶ for children's hospitalizations, ⁷ and for rates of rehospitalization of elderly patients, ⁸ both by Wennberg's group and by others. ⁹ Although these authors ¹⁰ have argued that these variations neither reflect an intrinsic variation in the prevalence of specific conditions ¹¹ nor are explained solely by differing patient preferences, ¹² what they do represent is still a matter of considerable debate. ¹³ Nonetheless, there appears to be general agreement that such variations are a symptom of something seriously wrong with our medical system, ¹⁴ which should be addressed either for its own sake ¹⁵ or as a way of controlling medical-care costs. ¹⁶

Such a thorough rejection of practice variations, however, ¹⁷ is far from the attitude that most American physicians have held about the nature of medical practice. As long as physicians have focused on the treatment of disease as "the great end of all our studies"—as the famous nineteenth-century American physician Austin Flint, Sr., ¹⁸ expressed the preeminent significance of therapeutics for medical practice—it has been recognized that an essential characteristic of medical therapy, i.e., drugs, was that to be conscientious, it could not be routine. There were at least two reasons for this. First, because the expression of disease varied not only with the nature of the disease but also with the nature of the patient, the drugs used to treat the condition would reflect both the specific disease and the specific individual who was sick. And second, because all drugs had many different actions, the risks of what we might call the side effects of a drug in a given patient might vary

-31-