Abstract WP291: Predictors of 30-day Same Hospital Readmission After Ischemic Stroke
Background: Health systems across the US are focusing on preventable hospital readmissions. Patient care systems are often focused on shorter hospital length of stays and faster discharges, while renewed efforts to regulate this penalizes health systems for readmissions. We aimed to explore factors that predict 30-day same hospital readmission after stroke, the impact on inpatient hospital cost and if a lower length of stay increased readmissions.
Methods: We included patients discharged from UC Health System with the principle diagnosis ischemic stroke (ICD: 9 433.01-9; 346) from 10/1/2011 to 4/30/2015. We excluded all patients with inhospital death, discharged to hospice or who left against medical advice (AMA). Patients are grouped in: 1) readmitted within 30 days (all cause); 2) not admitted within 30 days. We compared demographics, pre-existing/comorbid conditions and health care metrics (length of stay (LOS), direct cost for the index admission and readmission. Severity of illness adjustments for LOS and cost were made using the University HealthSystem Consortium® (UHC) algorhthm. We used unpaired t-test for comparing means continuous variables) and chi-square to compare frequencies.
Results: There were a total of 770 discharges, Group 1: 91, Group 2: 679. The groups were balanced in age, sex, race, and ethnicity, direct care cost and insurance status. Patients with readmission had longer adjusted LOS (O/E), 1.15±1.16 versus 0.97±1.11 (p=0.024), more renal failure (33.0 versus 14.0%, p<0.0001), peripheral vascular disease (9.9 versus 4.4%, p=0.048) and chronic atherosclerosis (40.7 versus 24.4%, p=0.002). (Table) Direct cost for readmissions was substantial, but not higher than for the index admission (14,746±21,795 versus 11,830±14,415, NS).
Conclusion: There is no evidence that lower hospital length of stay contributed to readmission. We identified unique markers for readmission risk, which requires further review in larger samples across regions and health care systems to better understand what factors lead to readmission after stroke and to identify potentially preventable causes of readmission.
Author Disclosures: E. Hemmen: None. E.D. Botts: None. D.L. Paulson: None. A.G. Snyder: None. T.M. Hemmen: None.
- © 2016 by American Heart Association, Inc.