Survivorship patterns in the posthospital phase of myocardial infarction.

Abstract

A prospective postinfarction follow-up study was used to identify subsets of patients with different survival patterns.Nine hundred forty patients who survived the hospital phase of an acute myocardial infarction were followed for 12-60 months.During the 5-year follow-up, 115 patients died of cardiac causes.Univariate analysis showed that prior myocardial infarction (PMI), left ventricular dysfunction (LVD) in the CCU, one or more ventricular premature depolarizations (VPDs) on a 6-hour Holter recording, and anterior myocardial infarction were significantly (p < 0.01) more frequent in patients who died of cardiac causes than in survivors.Survivorship analyses revealed a variety of survival patterns, depending on the presence or absence of the risk factors PMI, LVD, VPD and anterior infarction, as well as their interactive combinations.A com- bination of anterior infarction with LVD and VPDs identified a high-risk subset that made up 15% of the myocardial infarction population, and this group had 6-month and 3-year survival rates of 85% and 70%, respectively.After we excluded the high-risk subset, PMI, LVD and VPD each had significant yet independent influence on survival, with PMI having a greater effect on mortality than either LVD or VPDs.A low-risk sub- set that made up 24% of the population was identified by the absence of PMI, LVD and VPDs, and this group had a 3-year survival of 94%.PATIENTS WHO SURVIVE an acute myocardial infarction (MI) are at an increased risk of cardiac death and recurrent MI during the subsequent post- hospital years.Poor posthospital prognosis has been associated with older age, ST-segment depression and ventricular conduction defects on the resting ECG, cardiomegaly, advanced New York Heart Association functional class, left ventricular dysfunction (LVD) and frequent and complex ventricular premature depolarizations (VPDs).?-' Although these factors have been used to classify pa- tients in terms of their risk of cardiac death, very little is known about variable interactions as they pertain to the mechanisms influencing mortality.Techniques using multiple variables for modeling survivorship of postinfarction populations are useful in this regard because they utilize the entire experience of all patients, i.e., take into account the length of follow- up.'0 11 However, only limited use has been made of the full power of these techniques.In 1973, a prospective, longitudinal, posthospital follow-up program was initiated in Rochester, New York for patients discharged alive after recovering from acute MI.' The purposes of the present study are twofold: 1) to determine in this population whether clinical univariate risk factors act independently or in combination with each other to influence posthospital