Hospitals Testing Cardiac Bundled Payments Not Representative

Debra L Beck

July 06, 2018

Selection bias may limit the validity of Medicare's pilot bundled payments for cardiac episodes of care, according to a recent study published in JAMA Cardiology.

Hospital-level structural and performance characteristics differ significantly in 159 of US hospitals that volunteered to test out the new Centers for Medicare & Medicaid Services (CMS) payment model compared with a sample of nonparticipating hospitals.

"This research is important because the next step from a pilot program is to try to make it more broadly applicable or make it mandatory, but we should think twice about doing that before we understand how it will impact hospitals that are underrepresented in this pilot," said cardiologist and senior author, Daniel M. Blumenthal, MD, MBA, from Harvard Medical School, Boston, Massachusetts.

The CMS launched Bundled Payments for Care Improvement (BPCI) in 2013, initially making participation mandatory but later switching it to voluntary.

Blumenthal and colleagues did a retrospective, cross-sectional comparison of the BPCI model 2 bundles for percutaneous coronary intervention (PCI), coronary artery bypass grafting (CABG), congestive heart failure (CHF), and acute myocardial infarction (AMI).

The bundled payments covered hospitalization costs and, in most cases, up to 90 days of post-acute care, including nursing home care, complications, and rehospitalizations. Diagnosis-related groupings cover only what happens in the hospital, while bundled payments cover the entire 90-day episode in most cases.

They identified 159 hospitals that voluntarily enrolled in the pilot program between October 2013 and January 2017 and compared them with 1240 control hospitals in the same referral regions with at least 1 BPCI participant – identified via the America Hospital Association's Survey of US Hospitals from 2010 to 2013.

Participants in the pilot were larger, higher-volume facilities and were more likely to be privately owned or teaching hospitals. They were also more likely to have cardiac catheterization laboratories, cardiac intensive care units, and transplant units than nonparticipants and had significantly lower risk-standardized 30-day mortality rates for acute myocardial infarction (AMI) and congestive heart failure (CHF).

Importantly, BPCI participants also had significantly lower Medicaid bed day ratios (the ratio of Medicaid inpatient days to total inpatient days), and were less likely to be safety-net hospitals, which the researchers defined as those having a disproportionate share of payments from Medicare to serve underserved populations in the top 10% of hospitals nationally.

Only 4 (2.5%) BPCI participating hospitals were classified as safety-net hospitals compared with 121 (9.8%) control hospitals (P < .001).


"A voluntary payment experiment is not a clinical trial, nor is it a substitute for one," writes Karen Joynt Maddox, MD, PhD, Washington University, St Louis, Missouri, in an  accompanying editorial.

"We should consider BPCI, and other voluntary models like it, to be the equivalent of an open-label pharmaceutical or device study: important and hypothesis generating, despite their limited generalizability," she suggests.

"Any efforts by CMS to expand bundled payments to a broader range of hospitals will require either that the pilot program be made mandatory or that there will need to be mechanisms put in place to support hospitals that are reluctant to enrol to ease their transition to value-based care," Blumenthal said in an interview.

"Either we'll need to change the risk adjustment models or in some other way support them as they try to put in place the care management programs they'll need to drive improvements in spending and quality."

BPCI Advanced Launch

The performance of the new payment models has particular importance, said Joynt Maddox, a cardiologist and health policy researcher, because CMS has "doubled down" on bundled payments with BPCI Advanced, which will include bundled pay for a list of 29 inpatient clinical episodes along with some outpatient episodes, including PCI and defibrillator implantation.

"BPCI Advanced launches in October 2018 without any great data suggesting the program works terribly well, but there is a widespread belief among policy makers that bundled payments are a good mechanism to move us towards a more value-driven system," Joynt Maddox said in an interview with | Medscape Cardiology.

"We'll have to wait and see, but I think the same hospitals that joined the old voluntary program will join the new one — better organized and resourced hospitals are going to join, but a safety-net hospital that doesn't have enough social workers or nurses to be able to keep beds open is not going to have the wherewithal to invest in a new payment model."

"We should not expect that the performance data we get from this voluntary program will be broadly generalizable to a diverse cohort of acute-care hospitals," noted Blumenthal. "I don't think it will tell us how smaller hospitals and safety-net hospitals will perform under bundled payments for cardiovascular conditions."

A Good, Not Simple Idea

Blumenthal and Joynt Maddox agree that the idea of using financial incentives to drive quality improvement is a good one, but one that requires careful consideration and input from clinicians.

"I think policymakers think that it's easier than it really is and, to be fair, why would a lawyer in DC understand how to make good health policy? I think we really need more clinicians and people with clinical knowledge involved in policymaking," Joynt Maddox said.

"The idea is to build the bridge between inpatient and outpatient care, by coordinating care better, coordinating transitions better, reducing unnecessary care, and avoiding complications and readmissions," she added.

An example might be to switch from automatically sending certain patients from the hospital to a nursing home for 30 days. "Maybe they only need 10 days or 1 week, or maybe they can just go home," she said, but to allow better transitions and lower costs, there needs to be "someone to strategically approach the issue, and a lot of hospitals don't have that ability."

"You could argue that all hospitals should have the ability, and I totally agree that we should be doing a better job of organizing across settings, but the problem is that realistically these voluntary programs aren't going to attract under-resourced hospitals, so this pilot will tell us what is possible in a well-resourced hospital but not much more," said Joynt Maddox.

To date, the only outcomes reported on the new payment models have been a few evaluations from the federal government. Joynt Maddox recently reported some preliminary outcomes showing a lack of "clinically meaningful changes in access, utilization, or clinical outcomes" with episode-based payment for AMI, CHF, and pneumonia. Her final findings will be published soon.

This work was supported by the John S LaDue Memorial Fellowship at Harvard Medical School. Blumenthal is the associate chief medical officer of Devoted Health, a Medicare Advantage Insurance Company. This position is unrelated to the contents of this paper. Joynt Maddox does contract work with the US Department of Health and Human Services. No other relevant disclosures were reported.  

JAMA Cardiol. Published online June 27, 2018. Full text, Editorial

For more from | Medscape Cardiology, follow us on Twitter and Facebook


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.