Infection in prolonged pediatric critical illness: A prospective four-year study based on knowledge of the carrier state.

OBJECTIVE: This study was performed to determine the rate, timing, and incidence density of infections occurring in a subgroup of patients requiring a prolonged stay in a regional pediatric intensive care unit.

DESIGN: Prospective, observational cohort study over 4 yrs.

SETTING: This epidemiologic descriptive study was performed in a university hospital 20-bed pediatric intensive care unit.

PATIENTS: Critically ill children requiring > or = 4 days of intensive care.

INTERVENTIONS: The microbial carrier state of the children was monitored by surveillance cultures of throat and rectum, obtained on admission and twice weekly afterward.

MEASUREMENTS AND MAIN RESULTS: Data are presented on a total of 1,241 children, accounting for 1,443 admissions to the unit, corresponding to 18,203 patient days. The median pediatric index of mortality was 0.063 (interquartile range, 0.025-0.131), and the mortality rate in this subset of children was 9.6%. Five hundred twenty children had infections, an overall infection rate of 41.9% (520 of 1,241); 14.5% (180 of 1,241) of the children developed viral and 33.0% (410 of 1,241) developed bacterial/yeast infections. The incidence of bloodstream infection was 20.1 and lower airway infection 9.1 episodes per 1,000 patient days. We found that 13.3% of the children were infected with a bacterial/yeast microorganism acquired on the pediatric intensive care unit; 4.0% (50 of 1,241) of children developed infections due to resistant microorganisms. There were a total of 803 bacterial/yeast infectious episodes, of which 59.8% (480) were due to microorganisms imported in the patients' admission flora. These primary endogenous infections predominantly occurred within the first week of pediatric intensive care unit stay. The other 38.9% (312) were caused by microorganisms acquired on the pediatric intensive care unit. A total of 38 viral infections (24.5%) were acquired during pediatric intensive care unit stay.

CONCLUSIONS: Two thirds of all infections diagnosed in children with prolonged illness on pediatric intensive care unit were due to microorganisms present in the patients' admission flora.