Prediction of mortality after myocardial infarction by simple clinical variables recorded during hospitalization.

Simple clinical variables obtainable in any coronary care unit and in any patient were recorded in 769 consecutive patients who were admitted with acute myocardial infarction (AMI) and who were discharged from the hospital and followed for up to 3 years. To identify the patients at highest and lowest risk of posthospital mortality, a prognostic index was established from a stepwise logistic discriminant analysis of variables obtained in a consecutive series of 418 patients discharged alive from one of two coronary care units admitting new patients on alternate days. This prognostic index was validated by applying it to a comparison group of 351 consecutive control patients discharged from the other coronary care unit. In the training group, 59 of the 418 patients (14%) died during the first year after hospital discharge and 34 (8%) died during the second or third year. The stepwise logistic discriminant analysis made it possible to distinguish between 1-year survivors and nonsurvivors, but not between the patients who died during the second and third years and the 3-year survivors. Four variables were selected for obtaining a 1-year prognostic index: the maximum grade of left ventricular function during hospitalization (0 to 4), history of previous AMI (1 or 0), predischarge cardiothoracic ratio (0 to 0.99), and complete bundle branch block (1 or 0). Prognostic index = 7.0196-0.6515 function - 1.6623 previous AMI - 0.0729 cardiothoracic ratio - 1.0813 bundle branch block. This index was validated in the comparison group.(ABSTRACT TRUNCATED AT 250 WORDS)