Center for Medicare Advocacy Submits Comments to CMS about Direct Contracting, Geographic-Based Model

The Affordable Care Act created the Centers for Medicare & Medicaid Innovations (CMMI), which is tasked with testing demonstration programs aimed at delivery system reform. As discussed in a previous Weekly Alert in May 2018, among the models being developed is a suite of “Direct Contracting Model Options” described on the CMMI website. According to CMMI,

Direct Contracting (DC) is a set of three voluntary payment model options aimed at reducing expenditures and preserving or enhancing quality of care for beneficiaries in Medicare fee-for-service (FFS). The payment model options available under DC create opportunities for a broad range of organizations to participate with the Centers for Medicare & Medicaid Services (CMS) in testing the next evolution of risk-sharing arrangements to produce value and high quality health care. Building on lessons learned from initiatives involving Medicare Accountable Care Organizations (ACOs), such as the Medicare Shared Savings Program (MSSP) and the Next Generation ACO (NGACO) Model, the payment model options available under DC also leverage innovative approaches from Medicare Advantage (MA) and private sector risk-sharing arrangements.

As discussed in a previous Alert, the Center submitted comments to the initial request for information, including the overarching comment that the proposal was so ambiguous that it was difficult to provide meaningful feedback and specific recommendations without more substance offered.
CMMI recently sought input on one of the DC options, the Geographic Population-Based Payment Model. This model, among other things, “would offer an opportunity to take total cost of care (TCOC) risk for all Medicare [fee-for service, or those in traditional Medicare] beneficiaries in a defined target region.”
Drafted in collaboration with other beneficiary-focused organizations, the Center recently submitted comments to CMMI. In our “General Comments” we stated the following:

Overall, we reiterate our previous concerns about the lack of detail, and the posing of very specific questions in the RFI, about a concept that still remains largely abstract.  The lack of details of this proposal make it difficult to provide thoughtful comments and adequately troubleshoot problems Medicare beneficiaries might face. In the absence of a more clearly formed proposal, we offer the following general and specific comments.

On the one hand, we are encouraged that CMMI is exploring ways to address social determinants of health (SDOH) outside of the Medicare Advantage setting, and urge CMMI, and CMS more broadly, to continue this exploration beyond the direct contracting models. We also appreciate that the current RFI notes that traditional Medicare beneficiaries aligned to [entities] participating in the Geographic PBP model option “would retain all of their Original Medicare benefits, including freedom of choice of any Medicare provider or supplier, even if the provider or supplier does not have an arrangement with the [participating entity].”

On the other hand, any model that relies on capitation must have rigorous oversight and a heavy focus on beneficiary-reported outcomes and satisfaction to ensure beneficiaries are receiving the care they need. Given our current concerns about oversight of the Medicare Advantage program, we are uneasy about how this proposed model seems to invite even less regulatory oversight. Stinting on care would remain a constant threat for beneficiaries with chronic or complex conditions. It is unclear how [participating entities’] incentives would work, how [such entities] would achieve savings, and what authority [they] would have to control costs (e.g., would they employ utilization management?). More broadly, it is unclear, exactly, what current problems or issues within the Medicare program the Geographic PBP is attempting to resolve. In other words, there are significant details missing that would certainly inform our, and other consumer advocates’, input.