Abstract Rationale Broad-spectrum antibiotics are key to sepsis management. A prior randomized trial found similar short-term outcomes between cefepime and piperacillin-tazobactam for adults with acute infection. However, recent data suggests receipt of antibiotics with anaerobic coverage, like piperacillin-tazobactam, may be associated with increased long-term mortality. Whether the choice of initial antibiotics affects long-term outcomes is unknown. Methods We conducted an ancillary study of ACORN, a single-center randomized trial comparing cefepime versus piperacillin-tazobactam in adults with acute infection. All patients were included in this analysis. The primary outcome was all-cause mortality within 365 days of enrollment. Mortality data was obtained from the electronic health record and state vital statistics; individuals who were not known to have died using these sources were assumed to be alive. Secondary outcomes included 90-day mortality and 30-day readmission, defined as any unplanned inpatient admission at the enrolling center. Binary outcomes were compared using a chi-square test. Survival analysis was performed using Kaplan-Meier analysis with between group comparisons calculated using Cox proportional-hazards regression. Additional analyses included an adjusted analysis of mortality, performed using a logistic regression model adjusted for age, sex, acute kidney injury at enrollment, chronic kidney disease, baseline SOFA score, receipt of vasopressors and/or mechanical ventilation, treatment location, and source of infection. Results A total of 2,511 patients were enrolled with 1,214 assigned to cefepime and 1,297 assigned to piperacillin-tazobactam. The median age was 58 years. The most common sources of infection were intra-abdominal (24.4%) and lung (22.2%). A total of 303 (23.4%) patients in the piperacillin-tazobactam group and 293 (24.1%) in the cefepime group were confirmed dead by 365 days (p=0.66). A total of 205 (15.8%) patients in the piperacillin-tazobactam group and 220 (18.1%) patients in the cefepime group were confirmed dead by 90 days (p = 0.12). Readmissions within 30 days were similar between the groups (22.0% vs 23.6%, p=0.35). Kaplan-Meier analysis demonstrated no significant difference in survival (HR = 0.96; 95% CI 0.81-1.12; p=0.58; Figure 1). In adjusted analyses, piperacillin-tazobactam was not associated with increased mortality at 365 days (OR 0.95 0.77-1.16) or 90 days (OR 0.82 0.65-1.03). Conclusion In this ancillary study, patients assigned to piperacillin-tazobactam experienced similar long-term mortality and readmission rates as patients assigned to cefepime. Our results suggest that long-term clinical outcomes are comparable between these empiric regimens and do not support a detrimental effect of antibiotics with anaerobic coverage. This abstract is funded by: None
Reed et al. (Fri,) studied this question.