Applied Radiology

When the Centers for Medicare & Medicaid Services (CMS) unveiled the initial accountable care organization (ACO) proposal in 2011, providers responded unfavorably. This cool reception prompted CMS to increase physician participation and initially provide 3 tracks for provider participation. These tracks include the ongoing original Medicare Shared Savings Program (MSSP) – with an immediate shared savings/loss model and an immediate shared savings (and later, shared loss) model – as well as a one-time offer of a Pioneer ACO Model, which contained special provisions designed to encourage early adopters.

Summarizing the economic incentives, risk sharing is based on a case-mix adjusted benchmark index of expenditures for Medicare patients nationally. In the one-sided, savings-only model, an ACO is paid 50% of the calculated savings; in the two-sided model, the ACO receives 60% of the calculated savings or losses. To ensure that savings are actual savings rather than statistical noise and random variation, these shares are paid only on savings or losses within a certain range of the overall expected spend. Nothing is paid until a 2% to 3.9% savings is achieved. Presumably to ensure that a plan doesn’t deny necessary care, maximum savings rates are capped at 10% and 15% for the one- and two-sided models, respectively. For the two-sided model, losses are capped at 5% to 10% of the overall spend.

By structuring payments in this fashion, CMS hopes to encourage primary care physicians (PCPs) to be more effective stewards in caring for their patient pool—in particular, to use resources more parsimoniously and with greater care coordination. The incentive payment levels are also gated by—and dependent upon—performance on a wide range of quality measures. These measures are designed to minimize the risk of PCPs offering insufficient care.

While some physician groups have embraced ACOs, most wait to see whether CMS will change any key provisions of these organizations. Of particular interest to these wait-and-see physician groups is the lack of any incentive for a patient to actually use his or her assigned ACO, complicating the cost-control efforts of primary physicians. Many existing regulatory requirements around quality measures and IT will generate significant infrastructure costs, effectively discouraging the participation of nonelectronic medical practices (driving health IT is a major current policy goal). The ACO rules effectively require a separate governance structure that doesn’t merge well with current structures of large integrated delivery networks (IDNs) or multispecialty group practices.

Perhaps the biggest issue regarding ACOs is the fact that most Medicare spending is driven by specialists who are excluded by design in the ACO governance. Since roughly 5% of Medicare spending is on primary care services, CMS is hoping it can afford to significantly increase primary care pay (via ACOs), while still generating large savings on the other 95%. Admittedly, this hope of CMS’s does challenge the notion of integrated care, particularly as patients remain free to see whichever specialists they choose.

In my final blog post, I’ll examine the efficacy of ACOs and speculate on their future.