Because passive immunization against respiratory syncytial virus (RSV) is costly, its use should be restricted to well-defined groups of high risk children. We aimed to develop a clinical prediction rule that estimates the individual monthly risk of hospitalization for RSV infection in young children.
A retrospective cohort study was conducted in the southwestern part of the Netherlands. We included children born between January 1, 1996 and December 31, 1998. Children hospitalized for proven RSV infection were compared with children not hospitalized for RSV infection. The monthly risk was estimated with a logistic regression model including 5 clinical predictors (gender, gestational age, birth weight, presence of bronchopulmonary dysplasia and age) and the mean seasonal monthly pattern of RSV infections. We compared the predictive performance of the prediction rule with the guidelines of the American Academy of Pediatrics (AAP).
Information was collected on 2469 hospitalized children and 140,661 children who were 1,181,790 months at risk. All predictors were statistically significant, with age and the seasonal monthly RSV pattern having the strongest effects. The clinical prediction rule that included these predictors could better discriminate between high and low risk children than the AAP guidelines and would potentially reduce the number of immunizations by 20%.
The prediction rule reliably estimates individual monthly risks of hospitalization for RSV infection in the population studied. It provides an improved index for passive immunization but further validation in other populations is required.
From the Divisions of *General Pediatrics, †Pediatric Infectious Diseases and Immunology and ‡Neonatology, Department of Pediatrics, Erasmus Medical Centre-Sophia, and the §Department of Public Health, Erasmus Medical Centre, Rotterdam, the Netherlands
Accepted for publication October 19, 2005.
Supported by The Health Care Insurance Council of the Netherlands (Project OG99-021). The council had no involvement in study design, data collection, analysis and interpretation, writing of the report or decision to submit the paper for publication.
E-mail email@example.com. Reprints not available.