Abstract WP401: Influence of Hospital-Level Practices on Readmission after Ischemic Stroke
Introduction: The Center for Medicare and Medicaid Services intends to publicly report hospital-level readmission rates after ischemic stroke to enable comparisons of hospital quality. The influence of hospitals and hospital-level practices on readmission rates is unknown.
Methods: Adult subjects were entered into this cohort study when hospitalized for ischemic stroke (principal ICD-9-CM 433.x1, 434.x1, 436) in 6 states from 2003-2009 from the State Inpatient Databases. The primary outcome was any non-procedural readmission within 30 days. 26 hospital level practices of interest (utilization of diagnostic testing, procedures, ICU, tPA, and therapeutic modalities) were identified using a combination of ICD-9 procedure codes, diagnosis-related groups (DRGs) and Health Cost and Utilization Project utilization flags. Multilevel logistic regression was used to estimate the association between mean hospital-level practices and readmission after accounting for demographics, vascular risk factors, comorbidities, socioeconomic status and whether a practice was implemented in an individual patient.
Results: Hospitals accounted for 3.7% of the variance in the probability of readmission, intraclass correlation coefficient 0.037 (95% CI 0.031-0.043). Only three practices were associated with readmission: higher use of occupational therapy and acceptance of transfers were associated with lower readmission rates, while higher use of hospice was associated with higher readmission rates. (Table)
Conclusions: Hospitals are responsible for a small proportion of readmission variance. These findings suggest ways that readmission rates may be reduced and illustrate potential susceptibility of a publicly-reported readmission measure to manipulation.
- © 2012 by American Heart Association, Inc.