Estimating respiratory syncytial virus-associated hospitalization in the first year of life among infants born at 32-35 weeks of gestation

Background: Prophylaxis against respiratory syncytial virus-associated hospitalization (RSV-H) with anti-RSV monoclonal antibody is not considered cost-effective for routine use in most jurisdictions. The aim of this study was to develop a scoring tool to estimate local risk of RSV-H in the first year of life among moderately premature infants to assist in prophylaxis decision making. Methods: A retrospective cohort was constructed from population-based databases in Nova Scotia, Canada, to follow 32- to 35-week gestation infants from the prenatal period to <12 months of age, from 1998 to 2008. Potential risk factors were entered into the logistic regression model, where the dependent variable was RSV-H. Receiver operator characteristic analysis demonstrated cutoff scores producing the highest predictive accuracy, and the likelihood ratio test was used to select the final set of variables for the predictive tool. Results: In 2811 eligible infants, the overall RSV-H rate was 3.1% (88/2811). Of 17 variables considered, 3 were used to create the scoring tool: birth during December to February, household smoke exposure and household crowding. The positive likelihood ratios of predictive tool scores for high, moderate and low of RSV-H were 3.57, 3.38 and 1.95, whereas posttest probabilities (risk of RSV-H) were 11.4%, 10.8% and 1.6%, respectively. Conclusions: While able to predict infants at low risk of RSV-H, the tool did not discriminate high from moderate risk infants. The tool could be used in anticipatory care to help educate families about reducing risk of serious RSV illness in their newborn.