Changes coming to bundled payments and Direct Contracting, CMMI chief says

• A new mandatory bundled payment program for hospitals may start before 2023, the previously announced time frame.
• Shifting to a retrospective trend in 2021 is expected to save $1.1 billion in the current bundled payments program.
• Many Medicare ACOs are waiting out the first year of the Direct Contracting program, an adviser says.

Coming changes to Medicare value-based payment (VBP) models could include the launch of a new mandatory bundled payment program sooner than the previously announced 2023 start date, says the director of federal VBP efforts.

Brad Smith, director of the Center for Medicare and Medicaid Innovation (CMMI), said Oct. 16 at the Better Medicare Alliance’s annual Medicare Advantage Summit that his office is considering a faster rollout of a successor model to the voluntary Bundled Payments for Care Improvement Advanced (BPCI-A) program.

“The kind of feedback we got from folks is making us see if we can do that earlier,” Smith said. “Because what we heard from a lot of participants is, ‘Hey, if you’re going to move to mandatory, we’d like you to think about doing that sooner.’”

On Sept. 10, CMS notified participants in BPCI-A that a future mandatory model was coming and that the existing program would make a number of changes in 2021, including a shift of participants to a realized trend adjustment.

Reasons for the changes in Medicare’s VBP approach

The bundled payment changes are one result of a CMMI review, conducted in January through May, of the 54 models it’s launched during its 10-year existence. That review found only five of the models produced “statistically significant savings” after accounting for payouts to providers. Significant savings were found in the:

• Maryland All-Payer Model
• Repetitive, Scheduled Non-Emergent Ambulance Transport Model
• Home Health Value-Based Purchasing Model
• Pioneer Accountable Care Organization (ACO) Model
• ACO Investment Model

One of the worst CMMI performers was BPCI-A, which was on track to lose $2 billion if changes were not made, Smith said.

“While we’re learning a lot and we’re seeing some really great changes in care, especially people going home after surgeries versus going to skilled nursing facilities, that obviously meant — the $2 billion in losses — that the model was not sustainable,” Smith said.

The 2021 change from a prospective trend to a retrospective trend is estimated to increase savings by $1.1 billion.

“What that really means is that in the last two years of the model, we’ll end up breaking even,” Smith said.

Another 2021 BPCI-A change will require participation in clinical episode service groups rather than in individual episodes.

That change relates to another barrier to Medicare savings that CMMI found in its review: adverse selection.

As an example of that problem, Smith specifically cited the episodes from which hospitals in bundled payment models may select.

“People are rational actors,” Smith said. “They look for those areas where they have the most advantageous benchmarks, and those are the places they participate.”

Another barrier to savings for Medicare is the additional payments offered for care coordination in primary care and oncology care models, Smith said.

Savings also were prevented by inaccuracies in the benchmarks or prices of models, he said.

“What we’ve found is that’s pretty hard to do,” Smith said about accurately predicting prices. “Through a number of the lessons that we’ve learned looking at retrospective trends, thinking about how we do regional adjustments, thinking about how we do risk adjustment, we’ve learned a lot about benchmarks that we can work on and continue to refine [in] the models in the future. A larger number will produce savings in the future.”

Direct Contracting model to start next spring

The highly anticipated Direct Contracting (DC) is slated to start in April. That model allows physician groups to take 50% to 100% risk on patients. CMMI recently posted the benchmark and risk adjustment methodology data, which providers were waiting to examine before deciding whether to join.

“The number of applicants is higher than we anticipated,” Smith said.

However, one adviser to ACOs, which are expected to comprise a large share of DC participation, said those groups are planning to opt out of the 2021 start in favor of a second start in January 2022.

“They want to see how this will play out because new programs always have hiccups,” said Noah Champagne, a consulting actuary for Milliman.

His group concluded that the discount percentage for the DC Global track was high and would require a group to generate larger gross savings than in the Professional track to make participation worthwhile.

“Some very high-performing ACOs could make Global work, but I don’t know how many ACOs are really going to go for it,” Champagne said.

Such concerns may lead to more changes to the Global track, such as the type of discount-related changes that occurred after the launch of the Next Generation ACO program.

Still under development is a third DC option, called the Geographic track, which will allow health plans or large providers to take risk on Medicare fee-for-service (FFS) beneficiaries across a region. Participants could offer more benefits, lower premiums, lower out-of-pocket costs and more flexibility on components like care management.

Participants in the Geographic model also could build preferred provider networks to move providers off Medicare FFS rates, but their enrollees could still use any Medicare provider.

Smith said the increased focus on helping participants derive savings is part of an effort to get private health plans to join in on the DC models.

“We know that it is very hard for providers across the country to have one payer paying one way and another payer paying a different way,” Smith said.