Regular Medicaid Home Visits and Emergency Department Use Among Older Adults During Extreme Heat

Overview

This cohort study evaluated whether receipt of weekly Medicaid home visits during the month preceding extreme heat events was associated with reduced emergency department (ED) utilization among community-dwelling dual-eligible enrollees aged 65 years or older. Extreme heat days were identified at the zip code tabulation area level using a threshold of maximum temperature >90 °F and at or above the 97th percentile for that calendar day (reference period 2006–2017). Consecutive extreme heat days of 1 to 5 days were treated as discrete extreme heat events. The primary outcome was daily ED visits per 1000 enrollees during extreme heat and up to 2 days after event termination.

Methods and approach

Claims data for Medicaid and Medicare dual-eligible enrollees were linked with local weather observations for ZCTAs that experienced at least one extreme heat event between May and October in 2018 and 2019. The analytic cohort comprised individuals aged 65+ who had at least one Medicaid home visit in 2018 or 2019. The exposure definition compared enrollees who received weekly home visits in the month preceding an extreme heat event (treatment group) with enrollees who did not receive weekly visits (comparison group). The primary analysis estimated changes in daily ED visit rates during events and up to 2 days post-event, with stratified analyses by climatic region and presence of heat-sensitive chronic conditions. Analyses were conducted on data processed between June 2024 and October 2025.

Results

The sample included 597,388 dual-eligible enrollees (mean age 77.2 years; 70.9% female). During extreme heat, daily ED visits per 1000 enrollees increased by 1.29 (95% CI, 1.22 to 1.37) in the group that had weekly home visits in the preceding month and by 1.25 (95% CI, 1.19 to 1.31) in the group without weekly visits. The difference in the heat-associated increase between groups was 0.05 (95% CI, −0.04 to 0.14), not statistically significant. Stratified analyses by climate region and among subgroups with heat-sensitive chronic conditions yielded similar patterns without significant heterogeneity in the primary association.

Implications

Regular Medicaid home visits in the month preceding extreme heat events were not associated with a measurable reduction in ED utilization during such events in this cohort of older dual-eligible enrollees. The absence of an observed reduction does not exclude potential benefits of home visits for other outcomes, including earlier recognition of heat-related symptoms or facilitating appropriate care-seeking that may not reduce ED use but could affect severity or downstream care trajectories. Future research should evaluate alternative visit timing, visit content focused on heat mitigation, other health service endpoints (e.g., urgent outpatient visits, hospitalizations), and heterogeneity of effects across subpopulations and geographic settings to inform targeted adaptation strategies.