This community innovation has been accepted at the 2026 DHIS2 Annual Conference as a physical poster.
Climate-Sensitive Malaria Risk Modelling in Rwanda
Malaria remains a primary public health challenge in Rwanda. While national surveillance shows a dramatic decline in annual incidence, from 409 cases per 1,000 population in 2016/17 to 47 per 1,000 in 2022/23, this 85% reduction masks significant seasonal and spatial variability. Historically, Rwanda’s malaria modelling focused on demographic and programmatic factors. However, the high sensitivity of transmission to environmental shifts necessitates integrating climate factors to refine spatiotemporal risk assessments. This study utilized monthly malaria case data (2015–2024) within a Bayesian spatiotemporal Poisson regression framework. Spatial dependence was accounted for using the BYM2 model, which disentangles structured and unstructured spatial effects. Temporal dynamics were captured via a first-order random walk and space-time interactions. We integrated standardized ERA5 climate covariates, precipitation, temperature, and relative humidity, to quantify environmental drivers. Model inference was conducted using INLA, with results reported as posterior means and 95% CrI. Significant spatial heterogeneity was observed; high-risk clusters in Gisagara, Nyamasheke, Rutsiro, Kirehe, and Ngoma exhibited incidence rates two to six times higher than the national baseline. Posterior probability mapping identified hotspots with > 0.95 probability of elevated risk. Incidence increased with precipitation (IRR = 1.03; 95% CrI: 1.01–1.05) and humidity (IRR = 1.16; 95% CrI: 1.09–1.23), while higher maximum temperatures were associated with reduced incidence (IRR = 0.94). Risk peaked during 2015–2017, stabilised post 2020, and showed a marginal resurgence in 2024. Although national gains are substantial, transmission remains climatically sensitive and spatially concentrated. Integrating climate variables into Bayesian frameworks enhances the precision of hotspot identification. These findings provide a data-driven foundation for climate-informed surveillance and targeted resource allocation to sustain Rwanda’s progress toward malaria elimination.
Primary Author: Similien NDAGIJIMANA
Keywords:
Malaria Elimination , Rwanda, Spatio-temporal Modeling, INLA, Climate-Informed Surveillance, ERA5 Reanalysis, DHIS2, Public Health Intelligence
2 Likes