Relations between heart failure, ejection fraction, arrhythmi suppression and mortality: Analysis of the Cardiac Arrhythmi Suppression Trial

Objectives. We studied the relations between heart failure, ejection fraction, arrhythmi suppression and mortality. Background. Both left ventricular ejection fraction and functional class of heart failure are strongly associated with mortality after acute myocardial infarction. Both are also related to the presence of ventricular arrhythmias and have been identified as factors related to the ability to suppress ventricular arrhythmias. Little has been reported about the relations between these two factors and arrhythmi suppression or mortality. Methods. Baseline dat from the Cardiac Arrhythmi Suppression Trial were used to define several categories of heart failure and to relate both the resulting categories and ejection fraction to arrhythmi suppression and mortality using logistic and survival regression analytic methodologies. Results. Regardless of the prospective baseline definition of heart failure used, the dat consistently showed that heart failure was more powerful predictor of subsequent congestive heart failure events and arrhythmi suppression and was equally powerful in predicting death. However, each variable provided incremental information when included in the prediction model. Heart failure and ejection fraction appeared to be independent predictors of death. Interactions were observed: low ejection fraction was more predictive of failure of arrhythmi suppression in patients with than without evidence of heart failure before or at baseline; low ejection fraction was more predictive of subsequent congestive heart failure events in patients without than with evidence of heart failure before or at baseline. Conclusions. Although heart failure as prognostic feature appears to be somewhat superior to ejection fraction, both are powerful predictors of arrhythmi suppression and cardiac events in patients with ventricular arrhythmi after myocardial infarction. Each provides incremental prediction