PREDICTING CLINICAL STATES IN INDIVIDUAL PATIENTS

A probability model expresses the relation between the presence of cli nical findings (input or independent variables) and the probability th at a clinical state will occur (the dependent variable); for example, it expresses the probability that a disease is present or will develop or the probability that an outcome state will be reached. Probability models are developed by using selected study groups. Although these m odels are most often used to make predictions for groups of patients, they can also predict clinical states for individual patients. The fol lowing seven criteria provide a basis for the critical appraisal of pr obability models. In particular, physicians can use these criteria to decide when a specific probability model should be used to make a pred iction in an individual patient. Five of the criteria are concerned wi th the applicability of a model to a particular patient: 1) the compar ability of the patient and the study group used to develop the model; 2) the congruence between the clinical state of interest to patient an d physician and the model's outcome; 3) the availability of all input variables where and when the prediction is to be made; 4) the usefulne ss of a quantitative estimate of the predicted clinical state; and 5) the degree of uncertainty in the probability estimate. The other two c riteria are concerned with how well the probability model ''works'': 6 ) the fit of probabilities calculated from the model to the outcomes a ctually observed and 7) the model's ability to discriminate between ou tcome states relative to chance and to other, more traditional, predic tion methods. We illustrate the use of these criteria by applying them , in the form of questions, to a convenient, tabular version of a mode l that estimates a patient's chances of surviving for 10 years after h aving definitive surgical therapy for primary cutaneous melanoma.