The assessment of patient prognosis using an interactive computer program.

Abstract

The use of chronic disease probabilistic models to calculate patient prognosis is presented. The method relies on the calculation of transition probabilities between discrete disease states from a patient data bank. Maximum likelihood estimates are used for each age and unwanted fluctuations are removed by a moving average. An interactive computer program was written and the method applied to the calculation of the probability of stroke and myocardial infarction for male patients on antihypertensive therapy. A clinician could obtain the prognostic information for a patient in less than one minute. Applications in medical student education are also discussed.