Abstract Extreme heat exposure leads to excess cardiovascular morbidity and mortality among older adults, but misalignment in the geographic scales of health data and intraurban heat exposure make it challenging to inform local interventions. We introduce an adaptable framework for developing health burden estimates at finer geographic scales by connecting small area analysis (SAA) techniques with an epidemiologic model of heat related health risks in three steps: (a) estimating an exposure response function using individual level cardiovascular disease (CVD) hospitalizations for adults aged 65 and over; (b) downscaling daily CVD hospitalization incidence rates at the ZCTA‐level to census block groups (CBG) using SAA, adjusting for individual‐ and community‐level demographic factors, and (c) linking exposure‐response with daily incidence rates to estimate heat‐attributable burden at the CBG scale. Using the data for a metropolitan area in the southeastern United States for the summer of 2018, we estimated 1.8 to 22.0 excess hospitalizations per 10,000 people across CBG. We demonstrate the utility of this 3‐step approach to help inform localized intervention strategies by classifying neighborhoods as one of four risk groups: Low, Health‐driven, Heat‐driven, and Dual‐channel (driven by both heat and health). In comparison, using coarse‐resolution temperature resulted in a significantly smaller heat‐attributable health burden with different geographic distribution across the CBGs, highlighting the importance of using appropriately scaled exposure data. The findings from this study can support more effective heat interventions to address heat‐health related outcomes, rather than exposure alone.
Farr et al. (Mon,) studied this question.