Spatiotemporal analysis of leptospirosis in Colombia from 2007 to 2021. An environmental health metrics approach

Leptospirosis is a global zoonosis and environmental health problem because of its strong association with environmental factors. Although spatiotemporal statistics can estimate area-specific risk indicators, very few spatiotemporal analyses are done at the municipality level for extensive periods in tropical and subtropical regions, including Colombia. This study aims to assess the spatiotemporal distribution of leptospirosis in Colombia to identify significant environmental factors and produce municipality-level risk indicators. A Bayesian hierarchical model was used to estimate the association of rainfall, temperature, and overflooding with monthly leptospirosis incidence rates in all Colombian municipalities between 2007 and 2021, adjusting for sociodemographic confounders. The posterior mean was used to map the municipality-specific risk, probability of excess risk, and the municipality-specific emerging risk trend. Rainfall was associated with a higher risk of leptospirosis (rr = 1.023 [1.006–1.04]). Municipalities in the south, east, central, and west regions had the highest risk of leptospirosis, with low-risk municipalities in the central, east, north, and west regions. The overall risk decreased across the study period with only two periods of minimal variations. Multiple municipalities in the central region had a significant increasing risk. Spatiotemporal analyses provide a rich set of indicators to identify specific at-risk communities and support the development of environmental health indicators.