THE LIMITS OF GENERALIZED CARDIAC SCREENING-TESTS FOR PREDICTING CARDIAC COMPLICATIONS AFTER INFRAINGUINAL ARTERIAL RECONSTRUCTION

We examined the relative efficacies of different cardiac screening str ategies for infrainguinal arterial bypass. The outcomes of 205 electiv e leg bypass procedures over a 10-year period, including myocardial in farction (MI), total cardiac complications, and mortality were tallied . Clinical risk factors popularized by Goldman and Eagle, and the resu lts of dipyridamole thallium myocardial imaging (DThal) were recorded. The overall mortality rate was 3.4%, with a 3.4% incidence of MI and a 5.4% total cardiac complication rate. Both abnormal DThal (p = 0.011 ) and Goldman class II-IV (p = 0.030) were significant predictors of M i and cardiac death, but both suffered from poor specificity and posit ive predictive value. Because logistic regression analysis identified a correlation between angina, CHF, and an abnormal DThal, a customized screening strategy was developed to include the presence of angina, C HF and an abnormal DThal. Eighty-eight percent of patients suffering M I or death met these criteria, while only 11% of the complication-free group did. This screening strategy provided a superior sensitivity of 88%, specificity of 89%, positive predictive value of 25%, and 99% ne gative predictive value. A customized screening strategy (angina, CHF, abnormal DThal), developed from a 10-year experience with a single pa tient group, provided better predictive accuracy than any generalized screening formula.