A medical decision making system for predicting the probability of cardiac events in patients with residual ischemia after uncomplicated acute myocardial infarction

摘要

The goal was to develop an integrated consultative system, able to store data relative to patients' clinical history and noninvasive diagnostic tests, retrieve them and perform a statistical analysis, and finally present to the physician the probability of a future cardiac event for the individual patient, with a definite sensitivity and specificity. In the clinical scenario of residual ischemia at the exercise stress test performed prior to hospital discharge for a first acute myocardial infarction, the physician must take relevant management decisions, sometimes in the presence of confusing results of noninvasive techniques. A consultative system, based on a logistic regression analysis, was developed to assist the cardiologist in the interpretation of clinical data and provocative tests in this critical setting. This system allows a 91.7% correct classification of high or low risk patients for future cardiac events.