LAUSR

BACKGROUND Carbapenem resistance is a growing concern. Applying a novel algorithm, we examined epidemiology and outcomes of carbapenem resistance among gram-negative pathogens in hospital-acquired pneumonia (HAP) and ventilator-associated pneumonia (VAP). METHODS In a retrospective cohort design within the Premier Research database (2009-2016), all hospitalized adult patients with a gram-negative organism in a respiratory or blood culture specimen who fit criteria for HAP/VAP based on International Classification of Diseases, Ninth Revision, Clinical Modification, codes were included in the study. RESULTS Among 8,969 patients with HAP/VAP, 1,059 isolates (11.8%) were carbapenem-resistant (CR) organisms. Patients with CR organisms were more likely female (41.4% vs 33.2%; P < .001) and medical admissions (33.8% vs 27.4%, P < .001) than those with carbapenem-susceptible (CS) organisms. Patients with carbapenem resistance had higher comorbidity burden than those with carbapenem susceptibility (median [interquartile range] Charlson Comorbidity Index score, 3 [1-4] vs 2 [1-4]; P < .001). Pseudomonas aeruginosa was the most common gram-negative pathogen overall (24.9%) and among CS organisms (23.5%), and was second to Stenotrophomonas maltophilia (44.0%) among CR organisms (35.3%). Acinetobacter baumannii accounted for 11.8% of CR organisms and 2.5% of CS organisms (P < .001). Patients with carbapenem resistance were more likely than those with carbapenem susceptibility to receive inappropriate empiric therapy (25.8% vs 10.0%; P < .001). Carbapenem resistance did not affect adjusted mortality (22.9% CR vs 21.6% CS) or postinfection length of stay (except among survivors of VAP), but it was associated with excess costs ($8,921; 95% CI, 3,864-13,977). CONCLUSIONS Using administrative data, our novel algorithm identified patients with pneumonia at high risk for death, consistent with HAP/VAP. Among them, carbapenem resistance occurred in 12% of all cases and was associated with substantial excess in hospital costs.