| Issue/Objective |
Background: Quantifying the population-level burden of climate-sensitive diarrheal disease at subnational scales remains a critical gap in official health statistics, limiting evidence-based adaptation planning in climate-vulnerable settings. We demonstrate a scalable framework for generating standardized, district-level climate-health burden estimates using routine surveillance data. |
| Methodology/Approach |
Method: A spatiotemporal Bayesian hierarchical model, integrated with Distributed Lag Nonlinear Models, was applied to 12 years of monthly diarrhea surveillance data from 261 Ghanaian districts (2013 to 2024). We estimated relative risks, attributable numbers, and population-standardized attributable rates for maximum temperature and cumulative rainfall, adjusting for humidity, surface runoff, and sociodemographic factors. |
| Results |
Result: Temperatures of 27 to 32°C and rainfall of 85 to 394 mm were each associated with elevated diarrheal risk within two months (RR = 1·015, 95% CrI: 1·011 to 1·019; RR = 1·023, 95% CrI: 1·014 to 1·031). Despite modest relative risks, approximately 138,600 cases over the study period were attributable to these exposures, with 69,938 linked to temperature and 68,662 to rainfall, demonstrating how widespread population exposure amplifies small risks into substantial absolute burdens. A pronounced north-to-south gradient emerged, with northern districts bearing disproportionately higher attributable rates, and seasonal peaks concentrated between June and September. |
| Discussion/Conclusion |
Conclusion: These findings highlight the importance of integrating climate-informed burden metrics into routine health surveillance systems to enable more targeted, data-driven adaptation strategies. The proposed framework offers a practical pathway for national health systems to translate climate-risk signals into actionable subnational policy insights, particularly in high-burden, climate-vulnerable regions. |
| Presenters and Affiliations |
Etse Yawo DZAKPA African Institute for Mathematical Sciences Research and Innovation Centre (AIMS RIC) |
