CMS Administrator Seema Verma's recent op-ed in Modern Healthcare titled "Correcting the course of value-based care" misses the forest for the trees. While the administrator is right to support the movement to value-based care models, Health Care Transformation Task Force members believe she is wrong that the evidence does not suggest they are saving money. In fact, the cup is more than half-full.
America is undoubtedly moving toward value-based care, which is having a meaningful and positive impact on coordination of healthcare services, improved quality and reduced costs. Not all of this, however, is happening inside alternative payment models that can be counted by CMS model evaluators and actuaries applying narrow model review and model expansion parameters.
For this very reason, a better way to evaluate these models is needed. Policymakers should consider approaches used in the private sector, where each model tested contributes to building a body of knowledge and experience necessary for future success. The Health Care Transformation Task Force is hard at work reimagining approaches to appropriate model evaluations to propose to CMS.
Moreover, while we agree with some of the policy recommendations advanced by Verma, we are very concerned that her misinformed contention about a lack of savings has and will lead to policy decisions that will undermine the real and significant progress made to date.
CMS' model evaluations, for instance, are not able to capture increased participation in private-sector value-based programs, which are influencing utilization patterns in the comparison groups. Beyond model overlap among Medicare models, a 2019 study found that accountable care organizations cared for more than 44 million Americans nationwide (60% under commercial, 30% under Medicare and 10% under Medicaid contracts). One only needs to participate in a senior staff meeting of a health system or large physician group to understand that, whether the organization is in an alternative payment model or not, they are paying close attention and transforming their delivery of healthcare services to keep up with their competitors. The spillover, or "halo," effect is significant, making the pursuit of an accurate comparison group almost impossible.
Perhaps the only accurate studies were those early on in model development when markets were still learning of this evolution. A report from HHS' Office of Inspector General found that Medicare Shared Savings Program ACOs generated $1 billion in savings during the first three years of the program and improved performance on 82% of quality measures. Also, a 2019 MedPAC study found that beneficiaries consistently assigned to the same Medicare Shared Savings ACO from 2013 to 2016 experienced a 10% reduction in spending growth relative to the market average in their region.
The most important point has been lost in the calculations under the administrator's model portfolio evaluation: Overall health spending as a percentage of gross domestic product has slowed over the last decade for the first time since the start of Medicare and Medicaid. In 2010, CMS' Office of the Actuary's 10-year projection for healthcare spending was that in 2020 19.8% of GDP would be spent on healthcare. In fact, 18% of GDP is currently being spent. About $600 billion of projected spending has disappeared. This is the proverbial forest. The evaluators are looking at trees while also missing broader factors affecting them—a market shift to value-based care.
CMS evaluations should also consult other peer-reviewed research that demonstrates the transformation and innovation now underway in American healthcare. The Health Care Transformation Task Force has documented these studies in impact briefers.
Verma's recommendation to increase the use of mandatory models is understandable, however, we believe mandating narrow models focused on individual conditions is flawed. Instead, policy should encourage more comprehensive models in which the incentive is to not only improve the efficiency of care delivery but to avoid many procedures altogether by focusing on prevention in the first place.
Most importantly, Verma's suggestion about setting benchmarks lower and increasing discounts causes great concern. This will only serve to squeeze providers' upside in these models, which restricts capital reinvestment in care delivery and further innovation. Benchmarks should be based on a desired future rate of growth, factoring in population growth, aging and technological advances.
Continually evaluating performance based on historic accomplishment is driving a race to the bottom, undermining the very strength of the American healthcare system. Moreover, the policy choice to evaluate models based on the average performance of all participants is masking the true success of those high-performing participants who have fully invested in and committed to these models.
It is time for policymakers to recalibrate their perspective and appreciate the broader, positive advances being made, cheer them on, and give incentives for greater positive movement. We are on the right track, but poor policy decisions will undermine that progress and the greatness of the American healthcare system.
