Carbapenem Antibiotics for the Empiric Treatment of Nosocomial Pneumonia A Systematic Review and Meta-analysis

BACKGROUND: Previous meta-analyses suggested that treating hospital-acquired pneumonia (HAP), including ventilator-associated pneumonia (VAP), with empiric carbapenems was associated with lower mortality rates but higher rates of clinical failure for pseudomonal pneumonia. This study was an updated meta-analysis with sensitivity analyses and meta-regression to better understand the impact of carbapenem use in HAP/VAP. RESEARCH QUESTION: What is the efficacy of carbapenems for empiric treatment of nosocomial pneumonia? STUDY DESIGN METHOD: Databases were searched for randomized controlled studies evaluating empiric treatment for HAP and/or VAP, and studies were included comparing carbapenem- vs non-carbapenem-containing regimens. The primary outcome was all-cause mortality. Secondary outcomes included subgroup stratification and resistance development. RESULTS: Of 9,140 references, 20 trials enrolling 5,489 patients met inclusion criteria. For mortality, carbapenem use had a risk ratio (RR) of 0.84 (95% CI, 0.74-0.96; P = .01). Stratified according to VAP proportion (< 33%, 33%-66%, and > 66%), RRs were 0.95 (95% CI, 0.77-1.17; P = .66), 0.78 (95% CI, 0.57-1.07; P = .13), and 0.81 (95% CI, 0.65-0.99; P = .04), respectively. Stratified according to severity, only groups with Acute Physiology and Chronic Health Evaluation II scores < 14 and between 14 and 17 showed mortality benefit (RRs of 0.64 [95% CI, 0.45-0.92; P = .01] and 0.77 [95% CI, 0.61-0.97; P = .03]). Meta-regression did not show an association between Pseudomonas prevalence and mortality (P = .44). Carbapenem use showed a trend toward developing resistance (RR, 1.40; 95% CI, 0.95-2.06; P = .09) and a 96% probability of resistance emergence. INTERPRETATION: Carbapenem-based empiric regimens were associated with lower mortality rates compared with non-carbapenems, largely driven by trials of VAP. The mortality effect was not observed in trials with high disease severity and was not associated with Pseudomonas. The mortality difference was observed mainly in studies that used ceftazidime as control. There was a trend toward increasing resistance associated with carbapenems.