[Skip to Navigation]
Comment & Response
March 2018

Evaluating Readmission—Need for More Clarity on Methods

Author Affiliations
  • 1Yale School of Medicine, New Haven, Connecticut
  • 2Clover Health, Jersey City, New Jersey
  • 3Harvard Medical School, Boston, Massachusetts
JAMA Cardiol. 2018;3(3):265. doi:10.1001/jamacardio.2017.5321

We have several questions about the study by Gupta et al1 that are central to interpreting their evaluation of the Hospital Readmissions Reduction Program (HRRP). The Medicare Hospital Quality Chartbooks,2,3 based on national Medicare data, showed that the 30-day mortality of heart failure (HF) rose annually as early as 2006 (before implementation of the HRRP) before leveling off after 2012. Thus, understanding their sample is important. It would be useful to know how many American Heart Association Get With The Guidelines (GWTG)–HF hospitals were continuously enrolling throughout the study period and how their characteristics compare with the nation’s hospitals. Moreover, the authors included 115 245 HF hospitalizations from 416 hospitals over 9 years1—on average, 2 to 3 hospitalizations per hospital per month. Because many hospitals in the GWTG-HF program, a voluntary registry, are large, this small number per hospital is unexpected. Therefore, it would also be helpful to know the extent to which included GWTG-HF hospitals (and also among those continuously enrolling) were enrolling all of their patients with HF into the registry. Specifically, what percentage of the Medicare fee-for-service patients hospitalized with HF at these GWTG-HF hospitals (as documented in Medicare Provider and Analysis Review files) were entered into the GWTG-HF program during the study period? This information differs from the percentage of GWTG-HF patients matched to Medicare data (also of interest).

Add or change institution