MA plans have taken a different approach to value-based insurance design than did earlier VBID models—and only about one-third of enrollees are interested, according to an early analysis.


Jan. 21—Medicare is expanding its value-based insurance design (VBID) model and creating a new drug plan model.

The VBID expansion is the means through which the Centers for Medicare & Medicaid Services (CMS) will test innovations in the Medicare Advantage (MA) program. A separate Part D model will test whether plans can find ways to help lower Medicare drug spending.

“I think both models are good ideas that hold a lot of potential,” said Adam Finkelstein, JD, counsel with Manatt Health. “The details of each model's implementation really matter, though, and it is hard to gauge the prospects of success without seeing how CMS plans to implement them in practice.”

The CY20 VBID initiatives will add to the existing VBID program already operated by the Center for Medicare and Medicaid Innovation (CMMI). The MA VBID model—launched in January 2017—aims to cut Medicare spending while enhancing care quality through reduced cost sharing or additional supplemental benefits for enrollees with select chronic conditions.

The voluntary expansion will be open to MA plans, Regional Preferred Provider Organizations, and all special-needs plans. The CY20 VBID application period is open through March 1, 2019.

The new VBID models will focus on one or more of the following:

  • Condition- or socioeconomic-based designs
  • Rewards and incentives programs
  • Telehealth networks
  • Wellness and healthcare planning

As part of telehealth approaches to be tested in the model, MA plans can use telehealth services instead of in-person visits to meet network adequacy requirements. MA plans must ensure that enrollees can still make an in-person visit if they prefer.

“CMS expects that this will provide MA plans with an opportunity to enter into underserved markets, including rural areas where there may be few to no MA plan choices,” a CMS fact sheet stated.

Generally, telehealth may comprise up to one-third of the required access to in-network specialists.

The VBID model will be expanded in CY21 to test the Medicare hospice benefit among MA plans.

CMS also extended the performance period of the VBID model by an additional three years, through 2024.

Drug Plan Model

In January 2020, CMMI plans to launch the Part D payment modernization model to test whether incentives can be used in prescription drug plans (PDPs) or MA-PDPs to steer plans, patients, and providers to choose drugs with lower list prices.

Under the voluntary, five-year model, plans will take two-sided risk for CMS’s federal reinsurance subsidy (80 percent of catastrophic-phase liability). Plans in the model can create new rewards and incentives to increase engagement by enrollees and promote better enrollee understanding of their benefits, out-of-pocket costs, and other drug options.

CMS said the model aims to address overall Part D spending, which almost doubled (to $146.1 billion) from 2010 to 2016—despite a proliferation of generic drug options. CMS’s evaluations blamed the increase on high list prices of new specialty and branded medications for cancer, Hepatitis C, rheumatoid arthritis, and other conditions.

“However, while payments to Part D plan sponsors to administer the benefit have more than doubled from 2006 to 2017, the portion of the Part D benefit that plan sponsors are liable for managing, termed the direct subsidy, has decreased,” a CMS fact sheet stated.

After one plan year, CMS will retrospectively create a spending target benchmark that represents the federal reinsurance subsidy (80 percent of Part D catastrophic-phase costs after rebate) that CMS projects would have been paid to participating organizations if they were not participating in the model.

If a plan’s federal reinsurance subsidy spending is lower than its spending target, then the parent organization will receive performance based-payments based on the total percent saved. If the spending is higher than the benchmark, participating parent organizations will have to repay 10 percent of the difference.

Plans can use “clinically based drug utilization management techniques that make prescription drugs with lower list price available while also ensuring appropriate beneficiary access,” CMS stated.

A 2018 HealthMine survey of 781 MA beneficiaries found that just 25 percent said their plan offered a digital tool to predict drug costs.

The request for application to join the model for CY20 will be available on the model’s website, and applications will be accepted through March 1, 2019.

Finkelstein, a former CMMI official, warned that the application time frame could cause problems.

“One thing that strikes me is that it could be that there simply won't be enough time for most plans to pull together an application this spring,” he said in an interview. “If that's the case, I'd expect many potential applicants to skip this cycle and aim for a Jan. 1, 2021, start.”

Early VBID Results

CMS also recently released an evaluation of the first-year results for the MA VBID program, which targeted enrollees with COPD, congestive heart failure, diabetes, and hypertension. The voluntary program for enrollees provided VBID benefits to 61 percent of the 96,000 eligible beneficiaries. However, most 2017 MA plan data were not complete in time to include a full impact analysis for this first report, according to CMS.

The analysis found VBID MA plans had no changes in enrollment, which wasn’t surprising since they were not allowed to market their VBID benefits, said Christine Eibner, an author of the analysis and a senior economist at the RAND Corporation. CMS later relaxed those marketing restrictions.

The analysis also found no difference in Medicare bids by VBID plans. But that finding also was not a surprise to Eibner, she said in an interview, because savings from VBID that would justify lower bids were unlikely in the first year.

The VBID approaches used by the MA plans differed from earlier employer-plan VBID designs, Eibner said. Many earlier VBID models focused on incentivizing better pharmaceutical use, while most MA VBID plans focused on incentivizing care management. For example, enrollees in the MA plans had their cost sharing reduced when they participated in care management activities.

“Many of the plans see the care management itself as a key component of their VBID intervention,” Eibner said.

However, only 30 percent of patients in plans with a care management approach engaged with the care management program.

“Maybe for the subset that did engage it had more benefits, but that remains to be seen,” Eibner said.

Another difference from earlier VBID plans was the MA plans’ use of rebate checks for out-of-pocket spending, instead of reduced cost sharing, as an enrollee incentive.

The MA VBID model hopes to improve on the employer experience with VBID, which Eibner said improved process-related outcomes (like medication adherence) but had a decidedly mixed record on driving reduced overall spending.

“Potentially, there is more bang for the buck for implementing VBID in Medicare Advantage because people are costlier” than in employer plans, Eibner said.

The VBID model started in seven states and now is in 18 states. The Bipartisan Budget Act of 2018 expanded the model to all states and territories by 2020.


Rich Daly is a senior writer/editor in HFMA’s Washington, D.C., office. Follow Rich on Twitter: @rdalyhealthcare 

Publication Date: Wednesday, January 23, 2019