Effects of the COVID-19 pandemic, vaccination, and return to in-person classes on dengue incidence in major Brazilian cities: an interrupted time-series analysis

Abstract

Dengue transmission is closely linked to human mobility and vector-control activities, both of which were substantially disrupted during the COVID-19 pandemic. We conducted an interrupted time-series analysis of 477 epidemiological weeks (2015–2023) to estimate the effects of the pandemic, the national COVID-19 vaccination campaign, and the full resumption of in-person classes on weekly dengue incidence in ten major Brazilian cities. City-specific negative binomial models were fitted with adjustment for meteorological variables, and effect estimates were pooled using random-effects meta-analysis. At the pooled level, no significant immediate changes in weekly dengue incidence were observed after pandemic onset or vaccination initiation. However, both interventions were associated with significant changes in incidence trends, including a modest decline following pandemic onset and a gradual increase after vaccination initiation. No significant immediate or trend effects were detected for the full resumption of in-person classes at the pooled level. City-specific analyses revealed substantial heterogeneity in the direction, magnitude, and statistical significance of immediate and trend effects for each intervention. Between-city heterogeneity was high (I² = 75–91%), highlighting marked context dependence of intervention effects. Sensitivity analyses yielded consistent results. These findings indicate that the COVID-19 pandemic and related population-level interventions had heterogeneous impacts on dengue incidence across major Brazilian urban centers, underscoring the need for strengthened, locally adapted dengue surveillance and vector-control strategies during periods of major systemic disruption.