Spatial-temporal patterns, seasonality, and age-specific trends of varicella in Hangzhou, China, 2019-2024

Varicella has been subject to mandatory reporting to the China Information System for Disease Control and Prevention (CISDCP) by health agencies within 24 h of diagnosis since 2019. However, even if two-dose varicella vaccination has been recommended to be administered to children at 1 and 4 years of age in Hangzhou since 2014, emerging evidence of increasing breakthrough varicella cases in outbreaks challenges the present varicella vaccination schedule and its protective effect. We seek to identify hotspot areas and temporal trends of varicella at the township level in Hangzhou in the recent 6 years by using spatiotemporal analysis.

Varicella cases diagnosed by medical practitioners from 2019 to 2024, demographic data, and clinical data were extracted from CISDCP. Township-level population figures were estimated using a constant-share proportional allocation method based on the seventh census data in China. Global I statistics and the local index spatial autocorrelation (LISA) method were used to identify global autocorrelation and local autocorrelation, respectively. Retrospective spatial scan statistics were undertaken to explore potential spatiotemporal clusters of varicella. A harmonic regression model was used to quantify seasonality, and an age-specific trend was evaluated through the Cochrane-Armitage test.

A continuous decline in reported incidence of varicella in Hangzhou from 2019 to 2024 was observed, with 97.95 per 100,000 and 52.23 per 100,000 in 2019 and 2024, respectively. Seasonality of the bimodal peak was observed, with the first peak of varicella cases observed from May to July, whereas the second peak typically occurs from November to February of the following year. A pronounced reduction in varicella incidence among younger children (5-9) and a relatively slower decline in older pediatric and adolescent groups (10-19) were found. The spatial distribution pattern of varicella in Hangzhou at township levels was non-random, and hotspots tend to be more frequent in the suburbs than in downtown areas. A total of 34 significant varicella spatiotemporal clusters were identified by retrospective space-time scan statistics, the vast majority of which were located in suburban areas.

Varicella incidence has dramatically declined over the past 6 years. The 10-19-year-old age band exhibited a slower reduction than the 5-9-year-old age-band. Moreover, the tendency for varicella clusters to appear more frequently in suburban areas reflects disparities in varicella incidence geographically. Specific surveillance and control measures should be undertaken in high-incidence regions in Hangzhou.