Readmission Rates are Limited in Measuring Hospital Quality
BACKGROUND:
Recent national policies intended to reduce readmission rates primarily measure performance at the hospital level (i.e., through hospital readmission rates), which is logical because of hospitals' central role in inpatient care and the discharge process. However, this approach remains controversial in light of conflicting evidence on the relationship between readmission and quality of care, the poor predictive ability of most readmission risk models, and the potential for negative unintended consequences. This study assessed readmission rates as a hospital quality measure. Using Centers for Medicare and Medicaid Services' Hospital Compare data from 2009 and 2011 on hospitals' performance on process and outcomes measures for myocardial infarction (MI), congestive heart failure (CHF), and pneumonia, investigators sought to answer three specific questions: 1) How much do quartile rankings of hospitals based on readmission rates change over a two-year period? 2) Do changes occur in a pattern that suggests changes in quality or random variation? 3) Are readmission rates correlated with commonly used indicators of hospital quality, including mortality, volume, teaching status, and performance on process measures? Data used in this study represent 2,247 hospitals for MI, 3,758 hospitals for CHF, and 3,940 hospitals for pneumonia. Hospital teaching status and patient volume also were examined. In addition, hospitals were ranked from lowest readmission rate (quartile 1) to highest readmission rate (quartile 4), separately for each year.
FINDINGS:
- The change in readmission rates between 2009 and 2011 was inversely related to readmission rates in 2009: Hospitals with higher readmission rates in 2009 tended to improve by 2011, while hospitals with lower readmission rates in 2009 tended to worsen by 2011. On average, readmission rates for hospitals in quartile 4 ("worst" performers) in 2009 decreased over time by between 2% and 4%, depending on the condition, while readmission rates for hospitals in quartile 1 ("best" performers) in 2009 increased by between 3% and 7%.
- Readmission rates were higher in teaching hospitals and were weakly correlated with the other indicators of hospital quality.
LIMITATIONS:
- Readmission rate point estimates were compared for two years. Longer data panels could allow for more consistent identification of hospital performance.
- Quartile ranges of composite process-measure scores were close together, which may have limited the ability to detect a relationship between readmission rates and process-measure performance.
- Readmission rates were not compared to direct indicators of the quality of care transitions.
IMPLICATIONS:
- Results suggest that policymakers should consider augmenting the use of readmission rates with other measures of hospital performance during care transitions, and should build on current efforts that take a community-wide approach to the readmissions issue.
AUTHOR/FUNDING INFORMATION:
Dr. Volpp is part of HSR&D's Center for Health Equity and Promotion, Philadelphia, PA. The study was funded by the Commonwealth Fund.
Press M, Scanlon D, Ryan A, Zhu J, Navathe A, Mittler J, and Volpp K. Limits of Readmission Rates in Measuring Hospital Quality Suggest the Need for Added Metrics. Health Affairs June 2013;32(6):1083-1091.
