Document Type

Presentation

Publication Date

4-10-2026

Faculty Mentor

Corey Morrow

Abstract

Title: Variation in Rehabilitation-Related Durable Medical Equipment Use and Cost Among Stroke Survivors: A Retrospective Analysis of Medicare Beneficiaries.

Purpose: In the US, stroke is a leading cause of disability (Centers for Disease Control and Prevention, 2017). Stroke survivorship is expected to rise significantly in the coming years (Feigin et al., 2024; Martin et al., 2025), resulting in a large number of people with disabilities who rely on Rehabilitation-Related Durable Medical Equipment (RR-DME) for daily living. However, there is limited published research on equipment access and costs.

The overall objective of this study was to identify predictors of access to and costs of RR-DME among stroke survivors. This project was a collaboration with a research lab specializing in large-scale healthcare data analysis to influence data-driven rehabilitation policy and practice.

Methods: This was a retrospective cohort analysis of 23,604 nationwide Medicare beneficiaries who survived a hospitalized stroke. Newly-hospitalized stroke survivors were tracked for 365 days post-hospital discharge, and RR-DME Medicare costs were aggregated. Unadjusted analyses included unadjusted post-hospitalization RR-DME costs, including total expenditure, median costs, and mean/median ratio. For our primary objectives, we used a two-step model: 1) a logistic regression to explore predictors of any RR-DME claim and 2) a generalized gamma model with a log link for adjusted cost estimates. Covariates included demographic, regional, and clinical characteristics.

Results: Of those who received RR-DME, there was a mean/median ratio of 2.83, indicating a right-skewed distribution of costs. Black individuals (p < 0.01), those residing in the south region (p < 0.01), inpatient rehab discharges (p < 0.01), and SNF discharges (p < 0.01), and higher severity strokes (p < 0.01) were all more likely to receive RR-DME. Traditionally aged Medicare participants, full-dually eligible (p < 0.01), and individuals with higher levels of diagnosed comorbidities (p < 0.01) prior to their stroke were less likely to receive RR-DME.

In accordance with higher costs: men (< .001), individuals discharged to long-term care (< .001), higher severity strokes (p < .001), and those who had higher pre-stroke RR-DME costs (p < .001). Characteristics associated with lower costs included individuals >85 (p < .001), and individuals residing in the northeast region (p < .001).

Conclusion: This study identified variables associated with RR-DME accessibility and cost, including region, race, age, discharge destination, and stroke severity. This study provides foundational knowledge about DME access that can influence future projects.

Download

Share

COinS