Healthcare Disruptions and Health System Resilience under Climate Change in Malawi

Introduction
Climate change is increasingly associated with extreme weather events that disrupt healthcare delivery, yet the system-wide health consequences of these disruptions remain poorly quantified. While damage to health facilities following extreme events is well documented, far less is known about how climate-related disruptions to service accessibility propagate through health systems and affect population health. In Malawi, for example, Cyclone Freddy in 2023 led to the closure of at least 79 healthcare facilities, in some cases for several months, substantially reducing access to care in already resource-constrained settings.

Methods
We use Malawi, one of the world’s most climate-vulnerable countries, as a case study to investigate the interactions between extreme precipitation, health system functioning, and population health. We integrate empirically-estimated damage functions for the impact of precipitation on healthcare service delivery into Thanzi La Onse, an all-disease, health-system model calibrated to Malawi. Using two climate and socioeconomic futures (SSP2-4.5 and SSP5-8.5), we project the impacts of climate-related disruptions on healthcare access between 2025 and 2040, accounting for heterogeneous healthcare-seeking behaviour and changes in service accessibility.

Results and Discussion
We estimated, for the first time, the population health impact of precipitation-mediated disruptions to healthcare services. We estimate that up to 4% of healthcare appointments may be disrupted by precipitation-related events over the study period. Disruptions disproportionately affect conditions requiring continuous or long-term engagement with care, such as chronic pain, mental health conditions, and contraceptive services (Figure), where interruptions increase the likelihood of individuals falling out of care entirely. Additionally, acute care such ante- and postnatal care were disrupted. Despite these effects, we project only modest changes in aggregate DALYs, reflecting both pre-existing barriers to healthcare access and conservative assumptions regarding the scope of service disruption. Notably, our analysis does not yet capture complete precipitation-driven changes in disease prevalence, suggesting that our estimates likely represent a lower bound of true impacts. Nonetheless, the projected scale of disruption highlights a substantial and growing strain on healthcare systems under climate change, particularly in rural and infrastructure-poor areas. Future work will extend this framework to explicitly model facility closures, transport disruptions, and climate-sensitive diseases, providing a more comprehensive assessment of health system vulnerability and resilience.

 

Figure: Disruption to appointments due to precipitation-mediated disruptions under the SSP2.45 scenario (compared with the "no dispruption" Baseline) between 2025 and 2040. Those services that either required a long-term engagement with care, or were acute, were most affetced.