Risk, susceptibility and the epidemiology of proliferative neoplastic disease: descriptive vs. mechanistic approaches

Recently there has been a confusing and somewhat frustrating difference of opinion in epidemiology concerning the relative merit and utility of descriptive versus mechanistic knowledge in approaching the comprehension and control of human cancer. This distinction has both historical and evolutionary rationale; it generally being necessary to know something of WHAT is happening before one can be legitimately concerned with HOW it is happening. As outlined below, however, there is little justification for making value judgements based on this distinction, and the situation is more one in which descriptive approaches and mechanistic approaches are equally essential in the development of a systematic viewpoint, a viewpoint providing guidance to meaningful intervention.