964

Abstract

Introduction: Carbapenem resistance is increasing and correlates with worsened outcomes. Unfortunately, patients with carbapenem-resistant gram-negative infections rarely receive appropriate empiric therapy despite an association of early appropriate antimicrobials with improved mortality and morbidity in critically ill patients with septic shock. However, there is limited data on the impact of early appropriate antibiotics and mortality with carbapenem resistance. Methods: Adult patients admitted to an ICU with sepsis due to gram-negative bacteremia resistant to a carbapenem were included. Exclusion criteria included polymicrobial, recurrent, or breakthrough infections, or patients who expired before appropriate antibiotics were initiated. Patients were divided into those alive or deceased at 30 days after index culture. Classification and regression tree analysis (CART) was used to determine the breakpoint between early and delayed treatment. Independent factors associated with mortality were identified through multivariate logistic regression. Results: Of 115 patients included, there were 69 (60%) survivors and 46 (40%) non-survivors at day 30. Baseline criteria were similar except non-survivors were older (63.9 vs. 55.7 years, P=0.003) and had higher Sequential Organ Failure Assessment (SOFA) (11.7 vs. 9.7, P=0.009) and Charlson Comorbidity Index scores (5.4 vs. 3.9, P=0.02). Non-survivors were more likely to have catheter-related bacteremia (52.2 vs. 29%, P=0.012) and less likely to have bacteremia from a urinary tract source (4.2 vs. 30.4%, P=0.001). Type of organism, overall source of infection, first appropriate antibiotic, and use of combination therapy did not affect mortality. The CART-established breakpoint for early antibiotics was 0.8 hour (mortality 0%, <0.8 hour vs. 42%, >0.8 hour, P=0.082). On logistic regression, only age (OR 1.037, 95% CI: 1.004-1.071, P=0.027) and SOFA score (OR 1.165, 95% CI: 1.033-1.314, P=0.013) were independently associated with mortality. Early appropriate antibiotic use was not independently associated with mortality on multivariate analysis. Conclusions: Older age and higher SOFA score were independent risk factors for 30-day mortality. Early appropriate antibiotics was not associated with mortality.