Abstract
BACKGROUND: Respiratory syncytial virus (RSV) is a leading cause of acute lower respiratory tract infection (ALRTI) in young children, but data on surveillance case definition performance in estimating burdens have been limited. METHODS: We enrolled children aged <5 years hospitalized for ALRTI (or neonatal sepsis in young infants) through active prospective surveillance at 5 sentinel hospitals in South Africa and collected nasopharyngeal aspirates from them for RSV molecular diagnostic testing between 2009 and 2014. Clinical data were used to characterize RSV disease and retrospectively evaluate the performance of respiratory illness case definitions (including the World Health Organization definition for severe acute respiratory infection [SARI]) in identifying hospitalized children with laboratory-confirmed RSV according to age group (<3, 3-5, 6-11, 12-23, and 24-59 months). RESULTS: Of 9969 hospitalized children, 2723 (27%) tested positive for RSV. Signs and symptoms in RSV-positive children varied according to age; fever was less likely to occur in children aged <3 months (57%; odds ratio [OR], 0.8 [95% CI, 0.7-0.9]) but more likely in those aged ≥12 months (82%; OR, 1.7-1.9) than RSV-negative children. The sensitivity (range, 55%-81%) and specificity (range, 27%-54%) of the SARI case definition to identify hospitalized RSV-positive children varied according to age; the lowest sensitivity was for infants aged <6 months. Using SARI as the case definition would have missed 36% of RSV-positive children aged <5 years and 49% of those aged <3 months; removing the fever requirement from the definition recovered most missed cases. CONCLUSION: Including fever in the SARI case definition lowers the sensitivity for RSV case detection among young children hospitalized with an ALRTI and likely underestimates its burden.