Assessment and Validation of Syndromic Case Definitions for Respiratory Syncytial Virus Infections in Young Infants A Latent Class Analysis

Background: Respiratory syncytial virus (RSV) is a major cause of pediatric morbidity and mortality worldwide. Standardized case definitions that are applicable to variety of populations are critical for robust surveillance systems to guide decision-making regarding RSV control strategies including vaccine evaluation. Limited data exist on performance of RSV syndromic case definitions among young infants or in high-resource settings. Objective: The purpose of this study was to evaluate existing and potential syndromic case definitions for RSV among young infants in an urban, high-income setting using latent class analyses (LCA). Methods: We used data collected on infants <6 months of age tested for RSV as part of routine clinical care at Children's Healthcare of Atlanta between January 2010 and December 2015. We computed the sensitivity, specificity, positive and negative predictive values of clinical features, existing syndromic case definitions used by the World Health Organization (WHO) and alternative definitions we constructed using LCA to detect RSV infection. Results: Among 565 infants tested for RSV, 161 (28.5%) had laboratory-confirmed RSV infection. Among all case definitions evaluated, the definition developed through LCA (cough plus shortness of breath plus coryza plus wheeze plus poor feeding plus chest in-drawing) was the most specific (95.8%, 95% CI 93.8-97.8) and had the highest positive predictive value (51.4%, 95% CI, 34.9-68.0). WHO-acute respiratory infection (cough or sore throat or shortness of breath or coryza, plus a clinician's judgment that illness is due to infection) was the most sensitive (98.1%, 95% CI, 96.1-100.0; negative predictive value 96.3%, 95% CI 92.2-100.0). Conclusions: The WHO acute respiratory infection definition could be useful for initial screening for RSV among infants <6 months, whereas our alternative syndromic case definition may serve as the strongest confirmatory case definition in the same population. Appropriate case definitions will vary depending on the content and setting in which they are utilized.