Statistical Approaches to Clinical Predictions

Abstract

What is the chance that a patient will have a particular disease, a particular test result if the disease is present, or a favorable response to therapy? Such questions are becoming increasingly common in clinical medicine as the development of predictive indexes becomes more important.In a recent issue of the Journal, Dolgin and his colleagues¹ attempted to develop a method to predict which patients with radiographically detectable gallstones have cholesterol stones and which have pigment stones. Medical therapy with chenodiol is effective only for the former, and since treatment is expensive and time-consuming, it is important to be . . .