Analysis of the Spatiotemporal Spread of COVID-19 in Bahia, Brazil: A Cluster-Based Study, 2020-2022

Background: The COVID-19 pandemic progressed unevenly across the 417 municipalities of Bahia, Brazil. Pinpointing where and when risk peaked is vital for guiding interventions and preparing for future emergencies. Methods: We performed an ecological spatiotemporal study using confirmed cases recorded in e-SUS Notifica and Sivep-Gripe from January 2020 to December 2022. A discrete Poisson space-time scan in SaTScan (spatial window ≤ 50 % of the population; temporal window ≤ 6 months) identified clusters. For each cluster we calculated relative risk (RR) and log-likelihood ratio, considering p < 0.05 significant. Results: Thirty-three clusters were detected, 25 statistically significant. The largest cluster (164 municipalities; May 2020–June 2021) comprised 702 720 observed versus 338 822 expected cases (RR = 2.8). Two overlapping large clusters (185 and 136 municipalities) during January–February 2022—coinciding with Omicron circulation—showed RR ≈ 2.2. Localised clusters reached RR = 3.37. Spatially, risk concentrated in the south, southwest and east of the state, with isolated contryside outbreaks. Conclusions: The heterogeneous spatiotemporal dynamics of COVID-19 in Bahia underscore the value of real-time cluster detection for targeted surveillance and resource allocation. Tailored strategies, supported by genomic surveillance and socioeconomic indicators, are essential to mitigate future respiratory events.