In their first year, Medicare's Pioneer Accountable Care Organizations (ACOs) generated moderate spending reductions, with greater savings for ACOs that started out spending more, but the model may hurt future efforts to save money, according to a study published in the New England Journal of Medicine.
Researchers, led by J. Michael McWilliams, M.D., Ph.D., of Harvard Medical School, compared Medicare spending on Pioneer ACO beneficiaries and other beneficiaries, using data covering the period before and after Pioneer ACO contracts began. They found the ACOs reduced spending by $118 million in the first year of the program, exceeding the $76 million in bonuses CMS paid ACOs, generating net savings of $42 million.
The study echoes earlier research that showed 11 Pioneer ACOs earned shared savings in their second year and increased their average quality score to 85.2 percent.
When McWilliams and his team conducted a closer analysis, they found significantly higher savings for ACOs that started out with higher baseline spending, suggesting expenditures were easier to reduce when they began with more of them.
"There is growing evidence that alternative payment models can curb spending while improving or at least maintaining quality of care," McWilliams said in a statement. "But building on the early success of these models in Medicare will depend on getting the incentives right."
While the current ACO model provides incentives for reduced spending, McWilliams said, it is insufficient to offset those ACOs' financial investments in quality improvement. The ACOs that left the Pioneer program after 2012 had generated comparable savings to those that remained within the program, indicating the payment model may drive ACOs away.
McWilliams called for ending the provision of the model that lowers benchmarks for ACOs that reduces Medicare spending, essentially disincentivizing them from saving, according to the study. Researchers also called for redesigned benchmarks that continue to encourage higher-spending organizations to participate, but without benchmarks giving them a disadvantage compared to more efficient organizations.