A sizeable step forward but miles still to go: CMS’ Next Generation ACO model


CMS has issued a “Request for Applications” describing its Next Generation (NG) ACO. The model makes progress on three issues that have generated plenty of analytical handwringing from MedPAC and the broader ACO community. It also signals a strategy to set ACOs up to compete more directly with Medicare Advantage (MA).

(1) Enhancing predictability
The Medicare Shared Savings Program (MSSP) and Pioneer ACO models had different approaches to solving the same business parameters. With NG, CMS has generally picked the ones which enhance simplicity and predictability (see table). For example,

  • Retrospective beneficiary reconciliation used by MSSP meant that an ACO didn’t know which patients were included in their baseline or what their real cost target was until after the fact; a prospective approach reduces the uncertainty
  • “Matched population” approach to risk adjustment and trending used for Pioneer — a black box for most participating ACOs – is avoided and the tried and trusted HCC model used instead.

Notably, these choices also bring the ACO program closer to how MA operates, thus taking a step towards harmonization of Medicare payment strategies across programs (a key MedPAC goal).

(2) Moving towards sustainability
Both MSSP and Pioneer use past performance as a cost baseline and ratchets to make payouts progressively harder to earn. More efficient providers faced a disincentive to participate initially and the incentives only got weaker as performance improves.

The NG model mitigates (but does not fix) the issue. The provider’s past performance still serves as a baseline for setting the cost target. However, ACOs which are more efficient than others in its region could see – via a benchmark discounting mechanism — their cost target increased by up to 2%; those operating in more efficient counties vs. the national average could see a further net 1% increase. Taken together, efficient ACOs operating in efficient counties could get an effective 3% “subsidy” relative to their very inefficient peers operating in very high cost counties. Further, while the specific terms of how the benchmark will be reset in the second contract period (which resulted in a ratchet for MSSP and Pioneer) are not specified, CMS promised that recent ACO cost experience will be “eliminated” or “further deemphasized” in resetting the benchmark.

(3) Opening up new terrain for performance improvement
The NG model provides ACOs with three critical levers for managing costs and quality:

Member engagement. NG ACOs can offer members financial rewards (up to $50/year) for staying in the ACO network – modest, but enough to persuade a patient to get educated, and, on the margin, stay within the network’s bounds. In addition, a voluntary alignment process (whereby beneficiaries can elect to be part of an ACO patient panel even without a past record of seeing its doctors) encourages more direct discussions with members regarding the ACO benefits.

Targeted benefit enhancements. NG ACOs can implement three targeted benefit enhancements (easier access to skilled nursing facilities, telehealth benefits outside of rural areas, post-discharge home visits even if not homebound). These enhancements will allow ACOs to swap out or avoid care at more expensive settings or with more expensive providers. They also make patient participation in an ACO more attractive than straight Fee-For-Service (FFS) – and feel more like MA.

Expanding the provider partnership. NG ACOs can contract with other providers as “preferred providers” and “affiliates.” “Preferred providers” qualify as “in-network” for members to earn the $50/year bonus. “Affiliates” can negotiate alternative reimbursement arrangements (i.e., not tied to the FFS structure) with an NG ACO using the capitation payment option. As a result, the ACO can expand the footprint of care over which it can have a direct impact (favoring high performing providers and contracting accordingly) without necessarily expanding the provider set within the ACO.

Creating a market among Medicare eligibles for the ACO-enhanced benefit
Since the emergence of ACOs, there has been a lot of discussion about the end game – are ACOs here to stay or just a temporary stepping stone towards vertical integration? Many accountability-seeking providers have run the scorecard on ACO vs. MA and found the ACO model wanting. Hence the burgeoning interest in provider-payer collaborations and providers launching their own MA plans.

The NG ACO iteration model is an improvement but CMS has given itself plenty of latitude on specifics (especially around the benchmarking process and how the second contract period will be managed) and some key issues remain unaddressed (e.g. the complex array of quality metrics which burden MSSP and Pioneer are largely retained with a little swapping in and out on specific metrics). It remains unclear whether this improvement will be enough to keep providers focused in FFS vs. running fast towards MA.

In the end, however, providers are only one of the decision-makers. ACOs will thrive if they can secure a market to serve. With NG, CMS has opened the door to a clearer value proposition for the patient. While NG takes just a few steps past that doorway, a brisk walk further down that corridor (e.g. further benefit enhancements, integration of Part D, larger patient rewards for staying in network – options to which CMS already alludes) and ACOs might become an attractive new option for Medicare beneficiaries (FFS, MA and now ACO/FFS).

The real test for NG (and ACOs) may be less which providers sign up in the first round and more how Medicare eligibles react to the value proposition – with indifference, confusion or interest. A tall order… 

Interestingly, the move may also only strengthen provider interest in Medicare Advantage because NG ACOs will need access to the capabilities MA plans typically have, for example, strong, cost effective ways of communicating with patients not just about clinical matters but also about benefit designs, networks etc. and actuarial skills to assess whether the enhanced benefit options will help reduce or just add to cost. Finally, ACOs with allied MA plans may well already have a preferred Part D model which can be folded into the ACO once CMS creates that option. Anticipate that providers pursuing NG ACOs which don’t have an affiliated MA plan will be looking for plan partners.

* * *

A few other thoughts:

  • The NG model may accelerate provider restructuring by removing the ratchet (synergies are more valuable if not subject to a ratchet) and enabling network strategies in FFS (creating more opportunity for sophisticated winners)
  • Center of Excellence (COE) strategies delivering high value episodes should get more traction as a result of the NG model. The “affiliate” contracting mechanism means COE providers don’t have to wait for CMS reimbursement to catch up (via bundled payments or other structures) to serve FFS patients and the telemedicine benefit may allow for easier remote engagement by the COE with patients before and after the acute “core” of a episode.
  • MA plans will want to think strategically about how they play with NG ACOs. Providers want to operate on an all-payer basis. CMS has gone part of the way down harmonization (e.g. using HCCs for risk adjustment). Further harmonizing the contracting (e.g. payment models, quality metrics) could be a valuable negotiation “give” for MA plans; but such moves also make the NG model incrementally less costly for a provider to pursue. ACOs have always competed vs. MA for lives indirectly. But the NG model removes some of the handcuffs on ACOs on patient engagement and makes the competition more direct.