Key result
Adjusting for patient characteristics reduces the 30-day readmission gap between top and bottom hospitals by ~48%.
Why the study?
Does adjusting for a comprehensive set of patient characteristics account for differences in 30-day hospital readmission rates compared to standard Medicare adjustments in hospitalized Medicare beneficiaries?
Population
Health and Retirement Study participants enrolled in Medicare who were hospitalized from 2009 to 2012
Comparison
Adjustment for 29 additional patient… vs Standard Medicare adjustments of hospital…
Design
Cohort
Follow-up
30 days
Authors
Loading...
Broader patient characteristics explain nearly half of readmission differences; supports refining Medicare models but should not yet change practice.
Observational (n=8,067)
Yes
Does adjusting for a comprehensive set of patient characteristics account for differences in 30-day hospital readmission rates compared to standard Medicare adjustments in hospitalized Medicare beneficiaries?
Effect estimate: Reduction in difference -2.12 percentage points (95% CI -3.33 to -0.67)
Absolute Event Rate: 2.29% vs 4.41%
p-value: p=.003
Patient characteristics not included in Medicare's current risk-adjustment methods explain a large portion of the difference in readmission risk between hospitals, suggesting that hospitals with high readmission rates may be penalized largely based on the patient populations they serve.
Barnett et al. (2015) conducted an observational in Hospitalized Medicare patients (n=8,067). Comprehensive patient characteristics adjustment vs. Standard Medicare adjustments was evaluated on All-cause readmission within 30 days of discharge (Reduction in difference -2.12 percentage points, 95% CI -3.33 to -0.67, p=.003). Adjusting for comprehensive patient characteristics reduced the difference in 30-day readmission probability between highest and lowest quintile hospitals by 48% (reduction -2.12%; P=0.003).
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: