Abstract 78: Neighborhood Socioeconomic Deprivation Is Associated With Increased 30-day Mortality After Stroke: Results From A Nationwide Study Of 755,899 Medicare Patients
Lusk, J; Hoffman, MN; Clark, AG; Mac Grory, BC; Luedke, MW; Xian, Y; Bae, J; Hammill, BG
Published in: Stroke
The relationship between neighborhood socioeconomic deprivation- a reflection of the general economic and environmental conditions of a neighborhood- and outcomes after stroke is not known.
We performed a retrospective study of 100% nationwide Medicare inpatient claims from 2017-2019. We categorized admissions for cerebrovascular diseases using principal ICD-10-CM codes as follows: Ischemic Stroke (I63), Subarachnoid Hemorrhage (I60), Intracerebral Hemorrhage other than SAH (I61 and I62); Transient Ischemic Attack (G45); Precranial stenosis without ischemia (I65). The primary outcomes of interest were 30-day mortality and 30-day unplanned readmission (per CMS definition). The exposure of interest was the Area Deprivation Index (ADI), calculated at the census block group level. Adjustment covariates included beneficiary demographics (age, sex, and race), Medicaid-dual eligibility status (as a proxy for individual SES), end-stage renal disease status, discharge year, and the 29 Elixhauser index comorbid conditions. Generalized estimating equation methods were used to estimate logistic regression models.
A total of 755,899 patients were included (mean age 79.4 years (SD 8.3), 47.1% male, 18.9% dually eligible for both Medicare and Medicaid) and the median Elixhauser comorbidity index was 17.0 (SD 14.1). After multivariable adjustment, a strong, dose-dependent association was observed between neighborhood deprivation and 30-day mortality for all conditions studied (Figure). No adjusted association was observed between ADI and 30-day unplanned readmission.
For the first time, we show that neighborhood socioeconomic deprivation is associated with increased 30-day mortality for patients admitted with common cerebrovascular disorders, potentially in a dose-dependent fashion, after adjusting for age, sex, race, comorbidity burden and individual socioeconomic status.
