Nosocomial pneumonia in the pediatric trauma patient: A single center's experience

Objectives: To evaluate a single center's experience with the frequency rat e, patterns of occurrence, and impact an outcome of nosocomial pneumonia in the critically injured child. Design: Retrospective review of prospectively collected data. Setting: Level I university trauma center with a pediatric trauma intensive care unit. Patients: A total of 523 consecutive critically injured children admitted t o the pediatric intensive care unit during an 80-month interval. Measurements and Results: Thirty-five episodes of nosocomial pneumonia were identified in 29 children (frequency rate of 5.5%). The mean age of the ch ildren was 9.2 yrs, and the mean Injury Severity Score was 27 +/- 9. In 91% of patients (26 children), nosocomial pneumonia was associated with mechan ical ventilation, This represented a 13% frequency rate in injured children who were ventilated during the study period. The most common organisms rec overed were Staphylococcus aureus (21%), Haemophilus influenzae (19%), Pseu domonas (11%), and Enterobacter (11%). Early pneumonia (diagnosed 17 days a fter injury) was predominantly caused by Haemophilus species. In contrast, Enterobacter and/or Pseudomonas were isolated primarily in late pneumonia ( diagnosed >7 days after injury), Staphylococcus was prominent throughout th e hospitalization. overall, children with nosocomial pneumonia were more se verely injured (Injury Severity Score 27 vs. 17, p < .001) and had a longer hospital stay (26 vs. 7 days, p < .001). Despite this, mortality (6.9% vs. 7.9%, p = NS) was not significantly different from injured children withou t pneumonia. Conclusions: In this study of a single pediatric trauma center, nosocomial pneumonia occurred in a small but significant percentage of injured childre n. The frequency rate increased two- to three-fold with mechanical ventilat ion. Microbiology varied with day of onset. In contrast to the adult, morta lity did not seem to be significantly altered by this complication. Analysi s of additional pediatric trauma centers is encouraged to confirm these cha racteristics of nosocomial pneumonia in the injured child.