Differential effects of inappropriate empirical antibiotic therapy in adults with community-onset gram-positive and gram-negative aerobe bacteremia

Abstract

Bacteremia is associated with high morbidity and mortality, which contribute substantially to health care costs. A beneficial influence of appropriate empirical antimicrobial therapy (EAT) on patient outcome is evidenced; However, the evidence highlighting a comparison of clinical manifestations and of the effects of inappropriate EAT between Gram-positive and Gram-negative bacteremia is insufficient. In a retrospective 6-year cohort study, the total 2053 adults (Gram-positive, 566; Gram-negative 1487) presenting with community-onset monomicrobial aerobes bacteremia were recruited. Inappropriate EAT was defined as the first dose of an appropriate antimicrobial agent not being administered within the first 24 h after blood cultures were drawn. Although the bacteremia severity (a Pitt bacteremia score) at onset, comorbidity severity (the McCabe-Johnson classification), and 28-day mortality rate were similar in the two groups. Furthermore, after adjustment of independent predictors of 28-day mortality respectively recognized by the multivariate regression model in Gram-negative and Gram-positive groups, the Kaplan–Meier curve and Cox regression analysis revealed a significant difference (adjust odds ratio [AOR], 2.68; P < 0.001) between appropriate and inappropriate EAT in the Gram-negative group, but not in the Gram-positive group (AOR, 1.54; P = 0.06). Conclusively, patients with Gram-positive and Gram-negative bacteremia exhibited the similar presentation in bacteremia severity, but a greater impact of inappropriate EAT on survival of patients with Gram-negative aerobe bacteremia was evidenced.