BACKGROUND: Respiratory syncytial virus (RSV) immunoprophylaxis is given to high-risk infants during RSV season. Traditionally, the Centers for Disease Control and Prevention has defined RSV season in the United States by a threshold of 10% positive weekly RSV antigen tests (10%-positivity) reported to the National Respiratory and Enteric Virus Surveillance System (NREVSS). In recent years, RSV molecular testing has become more widely utilized; the extent of testing and its effect on determining seasonality are unknown.
METHODS: To describe national and regional RSV reporting trends, we assessed RSV testing and detections reported to NREVSS during July 2005 – June 2015, by viral isolation, antigen detection, and molecular testing (based on various diagnostic platforms). A convenience sample of NREVSS reporters was also interviewed to evaluate laboratory testing practices during 2009 – 2014. To characterize RSV season with molecular reports, we first assessed the traditional 10%-positivity threshold, and subsequently developed three methods based on either RSV detections or percent-positivity for use in different settings. We compared season characteristics of each approach to those derived from the traditional method.
RESULTS: Annual RSV molecular testing reports from consistent NREVSS reporters increased 200-fold during July 2005 – June 2015, and are now more frequent than reports based on antigen detection or viral isolation. Correspondingly, 62% (59/94) of interview respondents reported an increase in molecular testing during the previous 5 years, and 14% (13/94) reported discontinuing or decreasing antigen testing and viral isolation. Based on NREVSS data from consistent reporters, weekly RSV percent-positivity was consistently lower with molecular testing than with antigen detection; because of this, the 10%-positivity threshold was imprecise for characterizing RSV season with molecular tests, and the two diagnostic categories could not be meaningfully interpreted when combined. The new molecular-specific approaches typically yielded smoother curves, earlier season onsets and longer durations than the traditional method. The most consistent estimates were derived from a retrospective method (RS10) that assessed the weekly increase in normalized RSV detections; the two other approaches provided good real-time approximations to RS10. Based on RS10, national RSV seasons during July 2009 – June 2015 began between epidemiologic weeks 38 and 45, a median of 4 weeks earlier than traditionally defined.
CONCLUSIONS: RSV molecular testing is increasingly important for RSV surveillance but the traditional 10%-positivity threshold is not adequate to characterize RSV season with NREVSS data. RSV molecular diagnostics should provide a more comprehensive understanding of RSV circulation trends, potentially extending the benefits of immunoprophylaxis.