Nosocomial rotavirus infections: a meta-analysis.

Abstract

Nosocomial rotavirus (nRV) infections represent an important part of rotavirus (RV)-associated morbidity. The incidence of nRV influences the estimated total RV disease burden, an important determinant of cost-effectiveness of RV vaccination programs. Our aim is to summarize the existing evidence and produce reliable estimates of nRV incidence, in pediatric settings in Europe and North America. We searched electronic databases for studies on nRV incidence among pediatric inpatients. To ascertain complete case reporting, only studies describing active nRV surveillance in their methodology were included. Random effects meta-analysis was performed. Meta-regression was used to obtain results adjusted for important study characteristics. Twenty surveillance studies met the quality criteria for inclusion. The pooled unadjusted nRV incidence was 2.9 per 100 hospitalizations (95% confidence interval [CI]: 1.6-4.4). Incidence was significantly influenced by studies' seasonality-months (RV epidemic season only or year-round) and the age range of included patients. Highest nRV incidence was found for children <2 years of age, hospitalized during the epidemic months (8.1/100 hospitalizations; 95% CI: 6.4-9.9). The adjusted year-round nRV incidence estimate without age restriction was 0.4/100 hospitalizations (95% CI: 0.1-2.1) and 0.7 (95% CI: 0.0-1.8) for children <5 years of age. This is the first meta-analysis to summarize results of surveillance studies on nRV incidence. nRV is an important problem among hospitalized infants during the winter months. The lower season and age-adjusted nRV incidence estimate seems more appropriate for application in population-based burden of disease analysis.