Clinical predictors of respiratory syncytial virus infection in children

Aim: The aim of this study was to develop a clinical prediction model that identifies respiratory syncytial virus (RSV) infection in infants and young children. Methods: Children <= 36 months of age with respiratory illness, who were suspected of having RSV infection, were enrolled in this prospective cohort study during the study period between January and February 2002. RSV testing was performed on all patients. Results: Of the 197 patients enrolled in the study, 126 (64%) were positive for RSV and 71 (36%) patients were either negative for RSV or had a positive culture for viruses other than RSV. The mean age of patients was 5 months and 57% were male. Backwards stepwise logistic regression analysis identified cough (p = 0.000), wheezing (p = 0.002), and retractions (p = 0.008) as independent variables predictive of RSV infection. The prediction model had a sensitivity of 80% (95% CI, 71-87%), specificity of 68% (95% CI., 54-79%), positive predictive value 82% (95% CI, 74-89%), negative predictive value 66% (95% CI, 52-77), positive likelihood ratio 2.5 (95% CI, 1.8-3.7) and post-test probability of 82%. Conclusion: The combination of cough, wheezing and retractions predicts RSV infection in infants and young children.