Clinical question: Are there patient characteristics not currently measured by the Medicare readmission program that account for differences in hospital readmission rates?
Background: The Medicare Hospital Readmissions Reduction Program (HRRP) financially penalizes hospitals with higher than expected 30-day readmission rates. During 2014, more than 2,000 U.S. hospitals were fined $480 million for high readmission rates. HRRP accounts for differences in patient age, gender, discharge diagnosis, and diagnoses identified in Medicare claims over the previous 12 months; however, the impact of other factors is uncertain.
Study design: Survey data from the Health and Retirement Study, with linked Medicare claims.
Setting: Community-dwelling U.S. adults, older than 50 years.
Synopsis: Investigators analyzed more than 33,000 admissions from 2000 to 2012. They found 22 patient characteristics not included in the HRRP calculation that were statistically significantly predictive of hospital-wide, 30-day readmission and were more likely to be present among patients cared for in hospitals in the highest quintile of readmission rates. These characteristics reduced by 48% the differences in readmission rate between the highest- and lowest-performing quintiles. Examples include patient ethnicity, education level, personal as well as household income level, presence of prescription drug plan, Medicaid enrollment, cognitive status, and numerous others.
Bottom line: Patient characteristics account for much of the difference in readmission rates between high- and low-performing hospitals, suggesting that HRRP penalties reflect who hospitals treat as much as how well they treat them.
Citation: Barnett ML, Hsu J, McWilliams JM. Patient characteristics and differences in hospital readmission rates. JAMA Intern Med. 2015;175(11):1803-1812.