Nosocomial Acinetobacter baumannii infection in children in adult versus pediatric intensive care units

Abstract

Background We hypothesized that children treated in adult intensive care units (ICUs) might be at increased risk of hospital-acquired infection. We therefore compared the incidence of Acinetobacter baumannii infection in children treated in pediatric ICUs (PICUs) and in those treated in adult ICUs. Methods Patients aged A. baumannii isolates and calculated the A. baumannii incidence during ICU admission. Cox regression was used to identify possible risk factors for A. baumannii infection. Results Eight hundred and thirty ICU cases were included; 508 were treated in PICUs ("PICU group") and 322 in adult ICUs ("adult-ICU group"). Acinetobacter baumannii was isolated 19 (3.7%) and 23 (7.1%) times in the PICU and adult-ICU groups, respectively. Incidence densities for A. baumannii acquisition in PICU and adult-ICU groups were 1.63 and 5.62 per 1,000 ICU days, respectively. In the Cox regression model, the hazard ratio for A. baumannii acquisition associated with treatment in PICUs was 0.33 (95% confidence interval: 0.17-0.61; P = 0.001). In multivariate analysis after adjusting for age and sex, treatment in PICU remained significantly associated with a reduced risk of A. baumannii infection (hazard ratio: 0.40; 95%; confidence interval: 0.20-0.76). Conclusions Pediatric patients treated in a PICU had a lower rate and incidence density of A. baumannii nosocomial infection than those treated in adult ICUs, suggesting that PICUs provide better infection control for children than adult ICUs.