Mortality of Enterococcal Bacteremia: Are Inappropriately Treated Cases Associated with Higher Mortality?

2test was used to determine risk factors for death and comparisons of appropriate therapy versus inappropriate therapy. A multivariate logistic regression model (EPI INFO and STAT ADV computerized package of CDC and Postgraduate Medical School) was also used. Of 101 patients with enterococcal bacteremias, 40 died (39.9%). Predictors for inferior outcome identie ed in univariate analysis were 2 or more positive blood cultures ( pB0.0061), burns or decubital ulcer as underlying diseases ( pB0.004). The latter was the only predictor of death also in multivariate analysis ( pB 0.0006, OR 5.07 Cl (1.4‐ 18 5)). Surprisingly neutropenia and vascular catheters seemed to be protective in univariate analysis (pB0.0475 and pB0.0417, respectively). In multivariate analysis neutropenia was found to be signie cantly ( pB0.007) corrected to a lower mortality probably because of the use of early empirical therapy in neutropenia patients. In this subgroup of patients, death was 3.48 times less likely than in enterococcal bacteremia in non-neutropenic patients. We performed another univariate analysis comparing those who were appropriately treated ( nae81) versus inappropriately treated ( nae20, 19.9%) for enterococcal bacteremia. Both groups were comparable except in 6 of 36 risk factors. Only endocarditis ( pB0.0026) and 2 positive blood cultures (pB0.0267) were more frequently observed in inappropriately treated patients. Appropriate treatment was signie cantly more common in patients with cancer ( pB0.0165) and cytotoxic chemotherapy ( pB0.0147). In summary both analyses showed, that inappropriate therapy (either late therapy of drugs ineffective against enterococci) were associated with higher mortality, in patients with febrile neutropenia, fungaemia and endocarditis (1‐ 3). Vancomycin or other antibiotic resistance in enterococci was however, not associated with higher mortality.