Environmental Pharmacoepidemiology

5G-01 - The effect of Hurricanes Matthew and Florence on rates of hospitalization or death among North Carolina Medicare beneficiaries using electricity dependent durable medical equipment

Tuesday, August 26, 2025

8:30 AM - 8:45 AM ET

Presenting Author(s)

Jennifer L. Lund, PhD, MSPH

Professor
Department of Epidemiology, Gillings School of Global Public Health, University of North Carolina at Chapel Hill
Chapel Hill, North Carolina, United States

Co-Author(s)

BJ

Bradford Jackson

Cancer Information and Population Health Resource, UNC Lineberger Comprehensive Cancer Center
Chapel Hill, North Carolina, United States
CB

Christopher Baggett

Cancer Information and Population Health Resource, UNC Lineberger Comprehensive Cancer Center
Chapel Hill, North Carolina, United States
AP

Arrianna M. Planey

The University of North Carolina at Chapel Hill, United States
AL

Amy Lowman

Department of Epidemiology, UNC Gillings School of Global Public Health
Chapel Hill, North Carolina, United States
PD

Paul Delamater

Department of Geography
Chapel Hill, North Carolina, United States
EM

Eman Metwally, MD-PhD, MSCR, MPS

University of North Carolina at Chapel Hill, United States
EB

Eboneé Butler

Department of Epidemiology, UNC Gillings School of Global Public Health
Chapel Hill, North Carolina, United States
JW

John Wallace

North Carolina Institute for Public Health, University of North Carolina at Chapel Hill, North Carolina, United States
SR

Shabbar Ranapurwala

Department of Epidemiology, UNC Gillings School of Global Public Health
Chapel Hill, North Carolina, United States

Background: Hurricanes are increasing in frequency and intensity, potentially exacerbating pre-existing health inequities. Older adults who rely on electricity-dependent durable medical equipment (EDDME, e.g., ventilators, infusion pumps, dialysis machines) may be particularly vulnerable to hurricane hazards.

Objectives: We evaluated the impact of Hurricanes Matthew and Florence on rates of hospitalization or death among North Carolina (NC) Medicare beneficiaries who used EDDME.

Methods: Using Medicare enrollment and claims data for from NC residents age ≥65 years with 1+ claim for EDDME, we carried out a controlled interrupted time series (CITS) analysis to assess the immediate change and sustained impact of Hurricane Matthew (striking Eastern NC in October 2016) and Hurricane Florence (striking the same region in September 2018) on monthly rates (per 100,000 person months) of hospitalization or death, overall and stratified by race (Black or White). County-level hurricane exposure was defined using FEMA declarations for individual or public assistance. We constructed twenty-four month-level cohorts (November 2015-October 2019) comprised of older adults enrolled in fee-for-service or Medicare Advantage plans, with 1+ claim for EDDME in the previous 12 months.

Results: Monthly study populations ranged in size from 117,617 to 129,364. Pre-Hurricane Matthew, rates of hospitalization or death were higher in hurricane-exposed vs. unexposed counties (3,974 vs. 2,791 per 100,000 person-months). Within hurricane-exposed counties, rates of hospitalization or death were higher among Black vs. White beneficiaries before Hurricane Matthew (4,547 vs. 3,851 per 100,000 person-months). Per the CITS analysis in the overall population, there was a small immediate increase of 114 hospitalizations or deaths per 100,000 person-months (95% CI: -203, 432), followed by no sustained change in hospitalizations [6.1 more hospitalizations every month after hurricane (95% CI: -40, 52)] in hurricane-exposed vs. unexposed counties. Analyses focused on Hurricane Florence and stratified by race showed similar findings.

Conclusions: Before both hurricanes, hospitalization and death rates among older Medicare beneficiaries who used EDDME were higher in hurricane-exposed vs. unexposed counties. We observed an immediate increase in the rate of hospitalization or death in these counties immediately after the hurricanes hit, but no long-term impacts. Future work will explore the use of detailed hydrology and residential building damage-based models to assess hurricane exposure.