Dive Brief:
Payment models developed by the CMS Innovation Center generally aren’t saving Medicare money, according to an analysis by consultancy Avalere Health. But some models were successful at reducing costs, suggesting CMMI’s work could generate savings in future models.
One-third of the 18 models studied by Avalere produced savings for the federal government, while one-third generated substantial losses. Another one-third had a nominal impact, according to the analysis.
Quality impacts, like reducing emergency department visits or improving chronic disease management, were mixed. Four demonstrated improvement when it came to boosting care quality, while three showed minimal progress. However, another four had no statistically significant effect, and seven showed mixed reports.
Dive Insight:
The CMMI was created by the Affordable Care Act to test new payment and care delivery models in a bid to lower costs and improve quality in government healthcare programs. Most models have focused on Medicare, the major insurance program geared toward older Americans, according to Avalere.
However, contrary to its goal, the agency’s work has driven up federal spending. CMMI’s activities increased direct spending by $5.4 billion, or about 0.1% of the net spending on Medicare, between 2011 and 2020, according to a report from the Congressional Budget Office. And although models can be expanded nationwide if they’re shown to lower costs without hurting care quality, improve quality without increasing spending, or both, only four out of 50 models have been selected for expansion.
Under the Trump administration, the CMMI has become a target for government spending reductions. Last month, the CMMI said it would cancel four payment models early and halt two planned demonstrations before they could begin.
The latest report aims to offer a “more complete assessment” that includes quality and transparency metrics, as well as the experiments’ impact on government spending. The models were chosen for analysis because they were in use for at least two years with published evaluation reports, and affected more than 25,000 beneficiaries.
Overall, the models cost the federal government $7.7 billion, including $6.4 billion in model net expenses as well as $1.3 billion in model-specific implementation and evaluation costs. Not all models will produce savings, but they could help regulators learn new ways to advance value-based care, the report’s authors wrote.
Most of the models surveyed patients about their experience and health outcomes. Four models showed small improvement, while eight had no statistically significant changes and another two demonstrated mixed impact.
Additionally, the report noted “limited opportunities” for public input while the models were being designed. Half of the 18 models solicited comments through rulemaking or requests for information. None were endorsed by the Physician-Focused Payment Model Technical Advisory Committee, a decade-old group that aims to propose ideas for Medicare reform.
