Hospital Variation in Home-Time After Acute Ischemic Stroke

Abstract

Background and Purpose— Stroke survivors identify home-time as a high-priority outcome; there are limited data on factors influencing home-time and home-time variability among discharging hospitals. Methods— We ascertained home-time (ie, time alive out of a hospital, inpatient rehabilitation facility, or skilled nursing facility) at 90 days and 1-year post discharge by linking data from Get With The Guidelines-Stroke Registry patients (≥65 years) to Medicare claims. Using generalized linear mixed models, we estimated adjusted mean home-time for each hospital. Using linear regression, we examined associations between hospital characteristics and risk-adjusted home-time. Results— We linked 156 887 patients with ischemic stroke at 989 hospitals to Medicare claims (2007–2011). Hospital mean home-time varied with an overall unadjusted median of 59.5 days over the first 90 days and 270.2 days over the first year. Hospital factors associated with more home-time over 90 days included higher annual stroke admission volume (number of ischemic stroke admissions per year); South, West, or Midwest geographic regions (versus Northeast); and rural location; 1-year patterns were similar. Lowest home-time quartile patients (versus highest) were more likely to be older, black, women, and have more comorbidities and severe strokes. Home-time variation decreased after risk adjustment (interquartile range, 57.4–61.4 days over 90 days; 266.3–274.2 days over 1 year). In adjusted analyses, increasing annual stroke volume and rural location were associated with significantly more home-time. Conclusions— In older ischemic stroke survivors, home-time post discharge varies by hospital annual stroke volume, severity of case-mix, and region. In adjusted analyses, annual ischemic stroke admission volume and rural location were associated with more home-time post stroke.