The role of systemic antibiotics in acquiring respiratory tract colonization with gram-negative bacteria in intensive care patients: a nested cohort study.

Abstract

Colonization of the respiratory tract with Gram-negative bacteria in intensive care patients increases the risk of subsequent infections. Application of systemic antibiotics may prevent colonization with Gram-negative bacteria, but this effect has never been quantified. The objective of this study was to determine associations between systemic antibiotic use and acquisition of respiratory tract colonization with Gram-negative bacteria in ICUs. A nested cohort study. A university hospital and a teaching hospital. Patients with ICU stay of more than 48 hours and absence of respiratory tract colonization with Gram-negative bacteria on ICU admission. None. Acquisition was determined through protocolized surveillance. Associations were investigated with Cox regression models with antibiotics as a time-dependent covariate. In all, 250 of 481 patients (52%) acquired respiratory tract colonization with Gram-negative bacteria after a median of 5 days (interquartile range, 3-8 d) (acquisition rate, 77.1/1,000 patient-days at risk). Antibiotic exposure during ICU admission was present in 78% and 72% of the patients with and without acquired Gram-negative bacteria colonization, respectively. In Kaplan-Meier curve analysis, the median times to acquisition of Gram-negative bacteria were 9 days (95% CI, 7.9-10.1) and 6 days (95% CI, 4.8-7.2) in patients receiving and not receiving antibiotics, respectively. In time varying Cox regression analysis, however, the association between acquired colonization and systemic antibiotics was not statistically significant (hazard ratio, 0.90; 95% CI, 0.70-1.16). Among patients not colonized with Gram-negative bacteria in the respiratory tract at admission to ICU, systemic antibiotics during ICU stay were not associated with a reduction in acquisition of Gram-negative bacteria carriage in the respiratory tract during the ICU stay.