Understanding the local-level variations in seasonality of human respiratory syncytial virus infection: a systematic analysis.

Abstract

Background: While previous reports characterised global and regional variations in RSV seasonality, less is known about local variations in RSV seasonal characteristics. This study aimed to understand the local-level variations in RSV seasonality and to explore the role of geographical, meteorological, and socio-demographic factors in explaining these variations.

Methods: We conducted a systematic literature review to identify published studies reporting data on local-level RSV season onset, offset, or duration for at least two local sites. In addition, we included three datasets of RSV activity from Japan, Spain, and Scotland with available site-specific data. RSV season onset, offset, and duration were defined using the annual cumulative proportion method. We estimated between-site variations within a region using the earliest onset, the earliest offset, and the shortest duration of RSV season of that region as the references and synthesised the variations across regions by a multi-level mixed-effects meta-analysis. Using the three datasets from Japan, Spain and Scotland, we applied linear regression models with clustered standard errors to explore the association of geographical, meteorological, and socio-demographic factors with the season onset and offset, respectively.

Results: We included 7 published studies identified from the systematic literature search. With the additional 3 datasets, these data sources covered 888,447 RSV-positive cases from 101 local study sites during 1995 to 2020. Local-level variations in RSV season within a region were estimated to be 6 weeks (41 days, 95% CI: 25-57) for season onset, 5 weeks (32 days, 13-50) for season offset, and 6 weeks (40 days, 20-59) for season duration, with substantial differences across years. Multiple factors, such as temperature, relative humidity, wind speed, annual household income, population size, latitude, and longitude, could jointly explain 66% to 84% and 35% to 49% of the variations in season onset and offset, respectively, although their individual effects varied by individual regions.

Conclusions: Local-level variations in RSV season onset could be as much as 6 weeks, which could be influenced by meteorological, geographical, and socio-demographic factors. The reported variations in this study could have important implications for local-level healthcare resources planning and immunisation strategy.

Trial registration: PROSPERO CRD42023482432.