Small-area analysis of nationwide heat-related mortality risk using meta-regression and satellite reanalysis data

Due to climate change, extreme heat events are becoming more frequent and intense, and their impacts on health are an increasing major concern in public health. For more effective interventions, it is crucial to identify precisely high-risk areas considering regional characteristics. Therefore, this study aimed to assess the impact of high temperatures on mortality in South Korea at a small geographic scale using high-resolution data. We collected mortality and temperature data during the summer season (June-September) from 2015 to 2019 across 247 districts in inland South Korea. Also, we used satellite datasets to gather regional variables at a 1 km grid level. Through a three-stage analysis, we 1) calculated the heat-related mortality risk for each district, 2) estimated a pooled risk at the national level and derived more stable predictions for each district, and 3) estimated the risk at a 1 km grid resolution using regional variables and fitted a meta-regression model. The small-area estimates showed variations in risk across grids, even within the same district. Additionally, the heat-related mortality risk was higher for cardiovascular mortality than all-cause mortality. Among the regional variables, the proportion of the elderly population, urban cover fraction, and distance to emergency medical facilities were positively associated with the heat-related mortality risk. In conclusion, this study suggests the need to identify areas vulnerable to heat-related mortality risk at high resolution. Since the risk can vary depending on regional characteristics, it is crucial to develop response strategies considering these factors.