RESUMO
OBJECTIVES: Nosocomial rotavirus infections induce difficult problems for pediatric wards during winter epidemics. This prospective study was carried out to measure their incidence in a general pediatric ward by using epidemiological tools, clinical and molecular methods, and to examine the main factors increasing the risk of cross contamination. MATERIAL AND METHODS: The study was carried out on children, younger than three years old, hospitalized between 15 November 1996 and 1 March 1997. We examined the feces of all children for rotavirus on admission and during their hospital stay if they had developed diarrhea. The strains were typed by RNA electrophoresis. A cohort study was done to identify the factors of exposure to risk. RESULTS: Three hundred twenty-six of the 415 hospitalized infants were studied. One hundred and five were hospitalized for gastroenteritis, including 39 (37.1%) with rotavirus infection. Among 221 infants admitted without diarrhea, 11 (5.0%) had an asymptomatic community infection and nine (4.3%) developed nosocomial gastroenteritis. The only significant predisposing factor (P = 0.003) for nosocomial infection was the distance between the hospital rooms and the nurses station. Despite the low level of typing (51%), the molecular study suggested a wide diversity of nosocomial and community strains. CONCLUSION: The relatively low incidence of nosocomial infections found may be due to routine assignment to individual rooms, the isolation of infants admitted with acute diarrhea, and the policy of routine detection of asymptomatic carriers with the use of similar isolation measures for these children.
