Should CMS continue mandatory bundled payment models?

December 2017 | Update: CMS has officially cancelled the hip fracture and cardiac mandatory bundled payment models. What are the implications? Read More.

The Centers for Medicare and Medicaid Services (CMS) proposed a new rule that would cancel two mandatory bundled payment programs slated to begin in January 2018—the Advancing Care Coordination through Episode Payment Models (EPMs) and Cardiac Rehabilitation Incentive (CRI) Payment Models—in addition to reducing the number of mandatory geographic areas participating in the existing Comprehensive Care for Joint Replacement (CJR) model by (nearly) half and allowing hospitals to opt-out of the program.

Canceling these models would eliminate the expansion of bundles into new areas of care including heart attack treatment, bypass surgery, and surgical hip and femur fracture treatment, and allowing participants to opt-out of the CJR model may have sweeping implications on the continued velocity of bundled payment adoption.

Promise of Bundled Payments

CMS has been experimenting with bundled payments for decades, despite administration pendulum swings, the Affordable Care Act and other reforms, and there’s positive early evidence for their effectiveness.


In February 2017, a study in JAMA Internal Medicine found that the CJR bundled payment model was proven to save millions of dollars annually—and, if scaled up, could save billions—without impacting the quality of patient care. The 800 participating hospitals saved an average of 8 percent, while San Antonio’s Baptist Health System saw average Medicare episode expenditures decrease by 20.8 percent for joint replacement surgeries.

The New England Journal of Medicine identified that the higher-quality, more collaborative care encouraged by bundled payments could shave anywhere from 7-35 percent off the costs associated with treating patients with chronic conditions and frequent hospitalizations.

A 2016 study by McKesson and ORC International predicted that bundled payments will grow 6% over the next five years, making them the fastest-growing of all payment models that will account for 17% of medical reimbursement payments in five years. But, only half of payers and 40% of healthcare organizations report a readiness to implement bundles. And it’s that readiness question that has industry opinions on these proposed cancellations so mixed.

A Mixed Bag

Several groups, like the Federation of American Hospitals, have argued CMS doesn’t have the authority to force healthcare organizations to participate in any payment model. Others, including Presbyterian Healthcare Services (PHS) cited that an “undue burden” is being placed on “providers who lack familiarity, experience, or proper infrastructure to support care redesign efforts into a bundled payment system.”

However, the American Hospital Association (AHA), among others, supported delaying the bundles until 2018 (they were initially meant to go into effect in July 2017) so that healthcare organizations had time to prepare the necessary infrastructure and processes, and expressed concerns that "cancellation of the...programs may be disruptive to providers who have expended valuable resources to put these programs in place." The AHA continues to support bundled payment models, calling them “a win for physicians and patients alike."

Healthcare payment reform is a knotty issue, but as the single largest payer for health care in the United States, should CMS force healthcare organizations to evolve into a value-based world?


Voluntary vs. Mandatory Participation

The real question is this: will giving health care organizations the flexibility to voluntarily test bundled payments, rather than making them compulsory, make it more difficult to collect reliable results and measure their real impact?

Some argue (including the current secretary of Health and Human Services, Tom Price) that making bundles payments mandatory constitutes federal overreach. However, the Pennsylvania-based Geisinger Health System, brings up an interesting point: voluntary models would produce limited results compared to what a broader, mandatory model could provide.

“Participants in voluntary models tend to be early adopters whose performance may be systematically different than hospitals that choose not to participate—limiting the generalizability of program results.” A mandatory program “limits participants’ ability to ‘game the system’ by selectively referring or transferring complex patients to nonparticipating hospitals.”

CMS currently has two voluntary bundled payment pilot programs: Bundled Payments For Care Improvement Initiative (BPCI), slated to end in 2018, and the Oncology Care Model (OCM), which will conclude in 2021. While there isn’t enough evidence to definitively point towards BPCI’s success, early indicators show positive financial results.

Like any other innovation gaining traction, there are early adopters and majorities followed by late majorities and laggards. If bundled payment adoption becomes purely voluntary, participants will be the hand-raising early adopters and reform will be sandbagged, perhaps indefinitely.

Let the Data Decide

So, how do we cut through this grey area around reliably testing the effectiveness of bundled payments when a small sample size of healthcare organizations who are already efficiently delivering care can self-select into the program, or non-participants can decide to stay with the status quo (fee-for-service payments), should it seem more lucrative at the time?

CMS should try mandatory bundled payments, in addition to other APM initiatives until there is enough credible evidence one way or the other.

  • Letting the data speak will show the efficacy (or lack thereof) for bundled payment models. Data-driven change is much less risky and much more of an incentive for healthcare organizations to change, should results be promising. And, even a negative or neutral impact will help inform the structure of other APMs and payment reform initiatives that may be more effective.

  • Mandating these programs would set an even playing field across all healthcare organizations, with shared benchmarks that could drive larger improvements in the efficiency of care and patient health outcomes.

  • With a wide spectrum of participating organizations, policymakers would have a better understanding of how bundled payments affect healthcare organizations of all types and scales, not just early adopters.

Regardless of how CMS proceeds with this decision, as healthcare organizations continue to prioritize their focus towards achieving the Quadruple Aim, it’s clear that bundled payments, with other value-based reimbursements, are on course to becoming a significant piece of the reform puzzle.