Abstract
Preview Physicians must continually make treatment decisions with incomplete or uncertain information. The probability of disease and the sensitivity and specificity of the available tests are thus of utmost importance. By use of a mathematical diagnostic model, Dr Sox explores these variables in regard to coronary artery disease. Despite the article's apparent complexity, according to one reviewer “its excellent technical quality should make the subject easily understandable for anyone.” Dr Sox offers some practical suggestions in regard to test ordering, and the basic principles outlined for coronary artery disease diagnosis are readily adaptable to defining other disease processes.