Ventilator-associated pneumonia in a multi-hospital system: differences in microbiology by location.

OBJECTIVE To determine whether there were differences in the microbiologic etiologies of ventilator-associated pneumonia in different clinical settings. DESIGN Observational retrospective cohort study of microbiologic etiologies of ventilator-associated pneumonia from 1998 to 2001 in a multi-hospital system. Microbiologic results were compared between hospitals and between different intensive care units (ICUs) within hospitals. SETTING Three hospitals--one pediatric teaching hospital, one adult teaching hospital, and one community hospital--in one healthcare system in the midwestern United States. PATIENTS Patients at the target hospitals who developed ventilator-associated pneumonia and for whom microbiologic data were available. RESULTS Seven hundred fifty-three episodes of ventilator-associated pneumonia had culture data available for review. The most common organisms at all hospitals were Staphylococcus aureus (28.4%) and Pseudomonas aeruginosa (25.2%). The pediatric hospital had higher proportions of Escherichia coli (9.5% vs 2.3%; P < .001) and Klebsiella pneumoniae (13% vs 3.1%; P < .001) than did the adult hospitals. In the pediatric hospital, the pediatric ICU had higher P aeruginosa rates than did the neonatal ICU (33.3% vs 17%; P = .01). In the adult hospitals, the surgical ICU had higher Acinetobacter baumannii rates (10.2% vs. 1.7%; P < .001) than did the other ICUs. CONCLUSIONS Microbiologic etiologies of ventilator-associated pneumonia vary between and within hospitals. Knowledge of these differences can improve selection of initial antimicrobial regimens, which may decrease mortality.