Mortality attributable to carbapenem-resistant Pseudomonas aeruginosa bacteremia: a meta-analysis of cohort studies

Whether carbapenem resistance is associated with mortality in patients with Pseudomonas aeruginosa bacteremia is controversial. To address this issue, we conducted a systematic review and meta-analysis based on cohort studies. We searched PubMed and Embase databases to identify articles (up to April 2015). The DerSimonian and Laird random-effect model was used to generate a summary estimate of effect. Associations were evaluated in subgroups based on different patient characteristics and study quality criteria. Seven studies with a total of 1613 patients were finally included, of which 1 study had a prospective design, and the other 6 were retrospective. Our meta-analysis showed patients with carbapenem-resistant P. aeruginosa bacteremia were at a higher risk of death compared with those with carbapenem-susceptible P. aeruginosa bloodstream infections (pooled odds ratio (OR) from three studies reporting adjusted ORs: 3.07, 95% confidence interval (CI), 1.60–5.89; pooled OR from 4 studies only reporting crude ORs: 1.46, 95% CI, 1.10–1.94). The results were robust across a number of stratified analyses and a sensitivity analysis. We also calculated that 8%–18.4% of deaths were attributable to carbapenem resistance in four studies assessing the outcome with 30-day mortality, and these were 3% and 14.6%, respectively, in two studies using 7-day mortality or mortality during bacteremia as an outcome of interest. Carbapenem resistance had a deleterious impact on the mortality of P. aeruginosa bacteremia; however, the results should be interpreted cautiously because only three studies reporting adjusted ORs were included. More large-scale, well-designed prospective cohorts, as well as mechanistic studies, are urgently needed in the future.