Abstract DP338: Medicare Advantage seems to promote less intensive post-acute care utilization after stroke than Traditional Medicare

Background: Stroke is a leading cause of long-term disability among older adults in the United States. Approximately 70% of Medicare beneficiaries discharged for acute stroke use Medicare-covered post-acute care (PAC). American Heart Association/American Stroke Association Stroke Rehabilitation and Recovery Guidelines recommend stroke survivors who qualify for and have access to an Inpatient Rehabilitation Facility (IRF) receive treatment in an IRF over a Skilled Nursing Facility (SNF) due to improved outcomes for IRF-treated patients (Level IB). It is unclear how Medicare coverage type—Traditional Medicare (TM) or Medicare Advantage (MA)—influences stroke PAC utilization. We examine if TM or MA coverage is associated with (1) percentage of beneficiaries discharged to home, Home Health, IRF, SNF, or Other and (2) 30-, 60-, and 90-day hospital all-cause readmission following acute stroke. Methods: We used Center for Medicare&Medicaid Services 20% nationally representative 2018-2019 Medicare Beneficiary Summary Files and 2018 administrative data. We implemented inverse probability of treatment weighting using propensity scores (PS) using age group (65-74, 75-84, 85+), sex, race, Elixhauser Comorbidity Index, and urban residence. We estimated multinomial models for PAC discharge and linear probability models for all-cause readmission. All regressions were doubly robust. Results: Of 43,538 unique beneficiaries, 54% (n=23,352) were in TM and 46% (n=20,186) in MA. Relative to TM beneficiaries, MA beneficiaries were 47% more likely to be discharged home compared to the IRF (relative risk ratio 0.53 [95% confidence interval 0.50-0.57]), 28% more likely to be discharged home compared to SNF (0.72 [0.68-0.76]), and 35% more likely to be discharged home compared to Other (0.65 [0.59-0.71]) after PS matching. MA beneficiaries were 8.4%, 9.2%, and 9.7% less likely to be readmitted within 30 days (percentage point difference -0.084 [-0.093, -0.075]), 60 days (-0.092 [-0.11, -0.082]), and 90 days (-0.097 [-0.11, -0.088]) respectively. Conclusion: MA beneficiaries were more likely to be discharged to less intense PAC settings and less likely to be readmitted to the hospital following acute stroke. MA may steer stroke beneficiaries to less intensive PAC settings. Follow-up studies with more robust methods are in progress to adjust for confounders such as selection bias while exploring the clinical impact of less intensive PAC for some stroke survivors in Medicare.