Rapid Antigen Testing to Detect Respiratory Syncytial Virus Performs Well in Neonates

Abstract

Background: Infection with respiratory syncytial virus (RSV) in neonates is associated with significant morbidity and mortality. Early and accurate detection of RSV infection is important for appropriate isolation and cohorting, appropriate antibiotic usage, antiviral treatment, and public health awareness. However, RSV rapid antigen test performance has not been evaluated for neonates. Methods: A retrospective analysis of rapid RSV antigen testing in neonates (≤30 days) using a laboratory database of the Diagnostic Virology Laboratory at Texas Children's Hospital acquired between 1997 and 2008 was performed. Of 3691 respiratory samples collected consecutively, 784 (21%) were positive for RSV by rapid test and 402 (10.9%) showed growth of RSV in viral culture. Sensitivity, specificity, positive predictive values, and negative predictive values were calculated using viral culture as the reference standard. P values were calculated utilizing Pearson chi-square with SPSS Statistics 17.0. Result: Overall sensitivity was 90.3% and specificity was 88.2%. Sensitivity was lower in nonseason samples (76.5% vs. 90.9% [P = 0.048]), but specificity was higher in nonseason samples (95.6% vs. 87.1% [P ≤ 0.001]). Lower specificity was seen in samples submitted from neonates evaluated in the emergency department compared with inpatient floors, neonatal intensive care unit, and unspecified venues (85.1% vs. 95.6%, 94.3%, 95.5%, respectively P ≤ 0.001). Overall, there was no statistical difference in test performance based on specimen type. However, when RSV was in season, nasal washes were more sensitive than tracheal aspirates (P = 0.017). Conclusions: Rapid RSV antigen testing was sensitive and specific in detecting RSV infections in neonates. Moderate differences in test performance occurred in and out of RSV season, by sample type, and by the location within the hospital from where the sample was obtained.