Antibiotics associated with acquisition of carbapenem-resistant Pseudomonas aeruginosa in ICUs: a multicentre nested case–case–control study

Abstract

Background\nCarbapenem-resistant Pseudomonas aeruginosa (CRPA) strains are involved in severe infections, mostly in ICUs. Exposure to antibiotics other than carbapenems may be associated with isolation of CRPA; therefore, we aimed to identify those antibiotics using the case-case-control study design.\n\n\nMethods\nA case-case-control study was conducted in 2015 in a prospective multicentre cohort that included 1808 adults hospitalized in 2009 in 10 French ICUs. Patients were screened for P. aeruginosa at admission to the ICU and then weekly. Cases were patients with CRPA and patients with carbapenem-susceptible P. aeruginosa (CSPA) isolation. Controls were patients without P. aeruginosa isolation, matched with each case according to centre, length of stay and hospitalization period. Effects of antibiotic exposure were explored, after adjusting for prior treatment with carbapenems and confounding factors comprising colonization pressure with two logistic regression models. The two models were compared to identify specific risk factors for CRPA isolation.\n\n\nResults\nFifty-nine CRPA, 83 CSPA and 142 controls were compared. In adjusted multivariable analyses, exposure to carbapenems and to antibiotics belonging to the group of β-lactams inactive against P. aeruginosa were independent risk factors for CRPA isolation (OR, 1.205; 95% CI, 1.079-1.346 and OR, 1.101; 95% CI, 1.010-1.201, respectively). Conversely, exposure to β-lactams active against P. aeruginosa was an independent protective factor for CSPA isolation (OR, 0.868; 95% CI, 0.772-0.976).\n\n\nConclusions\nBesides carbapenem exposure, exposure to β-lactams inactive against P. aeruginosa was a specific risk factor for CRPA isolation. Clinicians should counterweigh the potential benefits of administering these antibiotics against the increased risk of CRPA infection.