After big payouts, Medicare may throttle back bundled payment program

• Proposed changes would switch Medicare’s voluntary bundled payments program from using prospective to retrospective calculations.
• The changes would have resulted in $34 million in penalties over the first two years, according to an analysis, instead of the $650 million in bonuses that were paid.
• Some support the changes, while others warn they could drive some of the 1,700 participants from the program.

Medicare has proposed significant changes to its largest bundled payment programs, and the modifications could slash bonuses and drive hospitals out of the voluntary model.

CMS recently proposed changes to its voluntary Bundled Payments for Care Improvement Advanced (BPCI-A) model, the biggest of which would be switching from calculating savings prospectively to retrospectively.

 “When you use a retrospectively set a target [price], you get preliminary information up front, but you don’t know what you’re moving toward,” said Aisha T. Pittman, vice president of policy at Premier. “You might think you are managing to one price, but then the actual targets are set and you went from succeeding in the model to failing in a particular episode.”

The changes were proposed for year four of the program, which starts in January.

Based on an examination of available data for the program’s first two years, a June 26 analysis by Acumen concluded the change in price calculation would have resulted in $34 million in cuts, or $68 per clinical episode. Under the current methodology, hospitals received $650 million, or $1,289 per clinical episode, in bonuses.

CMS leaders in the spring raised concerns that BPCI-A was among models that are not saving Medicare money.

The proposed BPCI-A changes likely stem from CMS concerns over those big payouts, said John Kalamaras, business intelligence analytics manager of DataGen.

Winner and losers if the changes go through

The changes are expected to benefit participants in some episodes, such as joint replacement, but hurt performance in many of the 32 other episodes in the model, according to the Acumen analysis.

The analysis projected that five episodes would benefit from the price-setting changes:

• Back and neck except spinal fusion
• Cellulitis
• Cervical spinal fusion
• Major joint replacement of the lower extremity (MJRLE)
• Outpatient cardiac defibrillator

For instance, average payments for MJRLE would increase from $182 to $452 per clinical episode.

Dan Santmyer, chief actuarial officer with Cedar Gate Technologies, said the changes may be unfavorable in the short term to providers in geographies or peer groups where a significant number are just starting to use efficiency to drive down costs.

“But, in the long term, I think the proposed retrospective trend brings long-term stability to the program’s target price methodology,” Santmyer said. “Prospective trends become problematic as a market approaches maximum efficiency; under the current prospective trend methodology, providers in highly efficient markets would eventually be measured against trends that assume the environment still has potential for efficiency gains.”

• Episodes that would be most adversely affected by the change in methodology include:
• Combined anterior/posterior spinal fusion
• Disorders of the liver
• Congestive heart failure (CHF)
• Sepsis

For sepsis, retrospective trend adjustment (RTA) would result in an average $189 penalty per clinical episode, rather than the $2,085 bonus paid on average under the current methodology. For CHF, an average $450 penalty would replace an average bonus of $2,464.

Those two episodes are among the largest in the model, with participants over the first two years garnering bonus payments of $189 million for sepsis and $142 million for CHF.

“This brings more uncertainty of what to expect,” Kalamaras said about the proposed changes, which he added may drive some of the 1,708 provider organizations from the program.

Stakeholder feedback sought

CMS is accepting feedback on the proposed changes, which are pending right before the fall sign-up period for the 2021 performance year.

The large swings in performance that likely would result from switching to RTA could be alleviated with the type of caps used in the predecessor bundled payments program, Kalamaras said.

Additionally, he urged equal distribution of patients between different peer groups to avoid handicapping groups with more patients enrolled in either BPCI-A or the largely mandatory Medicare Comprehensive Care for Joint Replacement program.

Another alternative would use national comparative measurements instead of peer groups.

“This is important to fix because if they don’t, they can’t accurately measure the program,” Kalamaras said.

The uncertainty extends to the current year, during which CMS has allowed participants to choose from among several options to account for disruption stemming from the COVID-19 pandemic, including forgoing accountability for all episodes this year.

Pittman said participants should be allowed to also choose from among those options for the episodes they manage. That is because the pandemic has affected patient volumes differently for different episodes based on the urgency of the condition.