Comparing Characteristics of Hospitals Participating in Medicare's BPCI With Nonparticipants

Medicare’s Bundled Payments for Care Improvement (BPCI) is a voluntary pilot program to assess bundled payments for multiple common cardiovascular conditions. A recent study aimed to determine whether participants in Medicare’s BPCI cardiovascular bundles are representative of US acute care hospitals.

The retrospective cross-sectional study published in JAMA Cardiology¹ involved hospitals participating in BPCI model 2 bundles for acute myocardial infarction (AMI), congestive heart failure (CHF), coronary artery bypass graft, and percutaneous coronary intervention and nonparticipating control hospitals, from October 2013 to January 2017.

The researchers identified the BPCI participants through the use of data from CMS. Controls for the study were identified using the 2013 American Hospital Association’s Survey of US Hospitals. In addition, the American Heart Association survey and CMS were used to obtain hospital structural characteristics and clinical performance data.

In total, 159 hospitals participating in BPCI model 2 cardiac bundles were compared with 1240 nonparticipating control hospitals and a multivariable logistic regression was estimated in order to identify the predictors of BPCI participation.

Following the analysis, BPCI participants were found to be larger, more likely to be privately owned or teaching hospitals, had lower Medicaid bed day ratios (ratio of Medicaid inpatient days to total inpatient days), and were less likely to be safety net hospitals, when compared with nonparticipants, according to the results. The BPCI participants also were found to have higher AMI and CHF discharge volumes, were more likely to have cardiac intensive care units and catheterization laboratories, and had lower risk-standardized 30-day mortality rates from AMI and CHF.

The multivariable analysis revealed that larger hospital size and access to cardiac catheterization laboratory were positively associated with participants, while being a safety net hospital was negatively associated with participation.

“Hospitals participating in BPCI model 2 cardiac bundles differed in significant ways from nonparticipating hospitals,” the authors concluded. “The BPCI outcomes may therefore have limited external validity, particularly among small and safety net hospitals with limited clinical cardiac services.”

In an accompanying editorial,² Karen E. Joynt Maddox, MD, MPH, noted that since the program was voluntary, there was significant selection bias, and the differences discovered in the paper mean the results of BPCI "will amost certainly not be generalizable to all hospitals nationwide."

However, BPCI's results are still important. Joynt Maddox wrote that while payment models should not be scaled based solely on comparisons of participating and nonparticipating hospitals, "we should not miss out on the opportunity to use this program to learn more about the potential for innovative care redesign efforts, or for unintended consequences, under a new paradigm of financial incentives."

References

1. Oseran AS, Howard SE, Blumenthal DM. Factors associated with participation in cardiac episode payments included in Medicare's bundled payments for care improvement initiative. [published onling Jun 27, 2018]. JAMA Cardiol. doi: 10.1001/jamacardio.2018.1736.

2. Joynt Maddox KE. What can we learn from voluntary bundled payment programs? [published online June 27, 2018]. JAMA Cardiol. doi: 10.1001/jamacardio.2018.1734.