Health facilities are vulnerable to impacts from disasters, such as earthquakes or floods. These disasters can disrupt health facility functionality either through direct impacts, such as building damage, or indirect impacts, such as disruptions to supply chains. In addition, disasters impact local communities, increasing health needs and disrupting health demand through reduced accessibility due to disrupted infrastructure and services.

Open-Source Solutions

Understanding how facilities and communities are potentially impacted by disasters is critical to developing relevant resilience and preparedness strategies. This understanding can be informed by geospatial analyses, which assess disaster risk to facilities and communities. Through layering locations of facilities and communities with hazard risk maps, researchers and analysts can gauge levels of risk and develop operational recommendations and strategies for communities and the health system. In the modern age, the data required for these types of analysis is becoming more openly available and freely accessible (open source). Leveraging relevant methodologies on this open-source data is both time- and cost-effective.

The Afghanistan Case Study

Afghanistan is prone to disasters, particularly flooding and earthquakes, which lead to devastating consequences for communities and hamper an under-resourced health system. Health facility preparedness has remained a challenge in Afghanistan. The health system has been strained by decades of conflict, along with ongoing challenges in human resources and health governance.

Searching for earthquake casualties in Herat Province. Source: Voice of America / Wikimedia

Since 2017, KIT epidemiologists have been working on monitoring and evaluation programs with local and international partners in Afghanistan. In 2024, KIT researchers from the Centre for Applied Spatial Epidemiology (CASE) launched a pilot study mapping facility disaster risk using free and open-access data. This pilot study was conducted as part of the Disaster Impact Mapping for Health Facilities in Afghanistan programme, in collaboration with MSF. The pilot focused on two facilities that play a critical role in service delivery for their respective communities. For the analysis, open-source road and hazard data were procured through OpenStreetMap¹ and the World Bank data catalogue², respectively.

Our Methodology: Assessing Direct and Indirect Risks

To identify the risk of direct disaster impacts to a facility, we assessed the risk of various hazard types (flooding, avalanche, drought, earthquake, and landslide) based on the facility’s location. Thresholds were established for each hazard type in order to gauge the level of risk. For example, earthquake risk was based on the amount of peak ground acceleration and the extent of the earthquake, while flood risk was based on the proximity of the flood to the facility and the water level of the flood. These risks were then graded into four possible classes of risk (no risk, low risk, medium risk, and high risk). In addition, a facility’s overall disaster risk was cross-tabulated, providing a final risk score for each facility.

Beyond direct impacts, we assessed the indirect impacts on a facility’s accessibility through anticipated impacts to infrastructure. Potential damage to major roads could disrupt supply chains, while damage to minor roads leading to communities could impact accessibility for the population. The analysis assessed the exposure of infrastructure (ex: roadways and access points leading to each facility) to understand possible disruptions in services and demand of the population in catchment areas (10km from each facility). The extent of potential damage to local roads allowed us to grade and report indirect damage risk for each facility by disaster type.

The figure above shows the location of a critical health facility in Khost Province in comparison to flooding risk and roadways. This information can help understand the potential challenges of facilities in accessing supply chains and human resources. At the end, the findings were presented to relevant stakeholders using maps to present the geographic distribution of risk and using simple and interpretable infographics to present risk gradients, as seen in the figure below.

Recommendations

Overall, this disaster mapping exercise conducted by the epidemiology team demonstrated the use of combining various sources of open-source data in highlighting vulnerabilities of health facilities towards disasters, which can also be applied in other settings. Researchers are encouraged to utilise open-source data, such as the ones used in this study, to their full potential, and, in collaboration with operational stakeholders, provide operational insights to drive preparedness. Analysis should be expanded beyond facilities and into community risks as well, since communities are key drivers of health needs and post-disaster demand, while at the same time accessibility of health services might be hampered. Insights should be delivered through appropriate media, such as simple infographics for non-researchers.

Looking Forward

With the recent increases in conflict-related health impacts worldwide and the looming threat of climate change, understanding and preparing for disaster impacts is paramount to delivering consistent health services worldwide. The Afghanistan pilot demonstrates that open source geospatial and disaster data can be used for simple and cost-effective analysis to provide the necessary operational insights before, during and after disaster events. Initiatives to record and collate disaster events and impacts into open and accessible platforms should be further encouraged.

References
¹OpenStreetMap contributors. (2025). Map of Afghanistan. OpenStreetMap.
²Global Facility for Disaster Reduction and Recovery (GFDRR). (2024). GeoNode: Afghanistan [Data set]. GitHub. https://github.com/GFDRR