Ecological Niche Modeling and Spatial Analysis of Snakebite Risk in Baringo County, Kenya: An Evidence-based Approach

International audience zIntroduction: Snakebite envenoming is a significant community health issue in Kenya, and Baringo County has the highest rates of the problem. The current study employs ecological niche modeling and spatial analysis to identify areas of high risk and environmental determinants of snakebite.Methods: Retrospective snakebite case data were obtained from health facility records in Baringo County, while environmental variables were derived from remote sensing and geospatial datasets (including elevation, land surface temperature, vegetation density (NDVI), and distance to rivers). Risk prediction was conducted using Maximum Entropy (MaxEnt) modeling, complemented by spatial autocorrelation analysis using Getis-Ord Gi* statistics. Model performance was evaluated using the area under the curve (AUC), Kappa statistic, and True Skill Statistic (TSS).Results: The MaxEnt model had a high predictive accuracy (AUC = 0.874, 95% CI: 0.832-0.916; Kappa = 0.618; TSS = 0.619). Distance to rivers (38.2%), elevation (25.7%), land surface temperature (19.3%), and NDVI (16.8%) were considered critical environmental factors. Moderate elevation (800-1200m), proximity to rivers (0-5km), optimal temperatures (25-32 °C), and moderate vegetation cover (NDVI: 0.4-0.8) were used to delineate the high-risk areas. Spatial analysis revealed significant case clustering (Moran's I = 0.805) and identified 82 hotspots.Conclusion: This study provides evidence-based spatial risk maps that can guide targeted snakebite prevention and intervention strategies in Baringo County. The integration of ecological niche modeling with spatial statistics provides a robust framework for assessing snakebite risk in similar settings.