Performance in the Medicare Shared Savings Program by Accountable Care Organizations Disproportionately Serving Dual and Disabled Populations

The growth of accountable care organizations (ACOs) and other alternative payment models has prompted concern about whether value-based payment will disadvantage providers who serve vulnerable populations. Understanding the impact of new payment models on these providers is critical as models are scaled up, but to date there has been limited evidence in this area. We examined performance in the Medicare Shared Savings Program (MSSP) by ACOs that serve a high proportion of dual or disabled beneficiaries from 2014 through 2016, and found that these ACOs had slightly lower quality and similar or higher baseline spending than other ACOs, but achieved greater savings and were more likely to earn shared savings.

We found that high-dual and high-disabled ACOs had slightly lower quality scores in 2014, but were equally or more likely to save and to share in savings relative to other ACOs. Both high-dual and high-disabled ACOs had greater savings per beneficiary than other ACOs ($212 vs. $51 for high-dual ACOs, p=0.04; $241 vs. $44 for high-disabled ACOs, p=0.012). Further, these ACOs were equally or more likely to earn shared savings; just over 30% of high-dual ACOs earned shared savings compared to 25% of non-high-dual ACOs (p=0.35) and 38% of high-disabled ACOs earned shared savings compared to 23% of non-high-disabled ACOs (p=0.013). Third, trends in quality and savings over time showed that these high-social risk ACOs closed the “quality gap” and had higher savings than other ACOs with more time in the program, providing suggestive evidence that these ACOs gained experience in managing their complex patient population.

One factor driving the finding of greater savings in MSSP among ACOs serving at-risk groups may be program design. The design of MSSP explicitly recognizes that certain populations are likely to have unique spending and utilization patterns, which would in turn affect benchmarks and, ultimately, provider performance in the program. Thus, cost targets are calculated separately for several vulnerable populations, including dual eligible and those who are disabled. Further, as of 2014, ACO benchmarks were based solely on each ACO’s own historic spending, which implicitly controls for social risk factors and unmeasured confounders. Use of a historical performance baseline may thus favor providers who care for vulnerable populations, who might have more spending that is due to unmet care needs and thus amenable to intervention. It is also possible that savings in high-risk ACOs reflect the success of interventions by these organizations to improve care coordination, communication, and disease management, which may disproportionately benefit patients with complex medical and social needs.

Our findings are in contrast with previous studies of other Medicare value-based payment programs which found that providers serving at-risk patients tended to have negative financial consequences under these payment models. Our results suggest that alternative payment models can have positive financial outcomes for providers who serve vulnerable populations. Lessons from MSSP may be useful for policymakers in efforts to reduce disparities when designing alternative payment models.