Did GPDC Save Medicare Money or NOT?

Breaking down the PY2022 GPDC (Global and Professional Direct Contracting) financial reconciliation results and model evaluation 

Reconciling the savings paid and losses recouped as part of an Innovation Center model with evaluation findings that differ has often been painful—but important. Learning to unpack success and statistical significance, and taking the best lessons forward requires effort.  In this Q&A with Coral Co-Founder Melissa Cohen, we explore what happened, what conclusions are safe to draw, and what we can all learn from it.  

Did GPDC save Medicare $371.5 million, or did it have no impact on Medicare spending? These seem mutually exclusive—both can’t be true, can they? 

Melissa: The first thing to call out is that even though these announcements were released back-to-back, they are speaking to different years.  Looking to the 2022 financial reconciliation results, these are based on an agreed upon methodology in a participation agreement between a DCE and the CMS Innovation Center.  The participation agreement holds DCEs accountable for a total cost of care target or benchmark. Savings are paid or losses recouped according to that published methodology which bakes in a lot of policy considerations.  CMS just published individual DCE financial results and when they were all tallied together, net savings to Medicare was $371.5 million. CMS defines Medicare net savings as gross savings minus the amount paid out to the DCEs in shared savings.  Gross savings is the delta between the total of all DCE target benchmarks and the amount DCEs actually spent on their attributed population in PY2022.

The Innovation Center separately attempts to apply academic and research rigor to the question – how much did the attributed beneficiaries cost Medicare in this model and what would that cost have been absent the model and intervention? Because of the time it takes to perform this type of health services research, there is usually a longer lag before results are released.  There, the evaluation concluded there was no impact on Medicare spending due to the GPDC program in 2021.  It is important to note that 2021 was only a partial performance year beginning in April.  While this report covers 2021, the 2022 evaluation might conclude something similar.  Evaluation results do not always align directionally with financial reconciliations.  In part, we can see different answers because we ask different questions. But the directionality has always been hard – for everyone. Ideally, CMS isn’t paying out savings for a program that, when evaluated, finds the costs would have been the same without the intervention.

Okay, so financial methodologies and research evaluations are different, but do the sets of results and findings agree on anything? Are there results that stand out to you? 

Melissa: The evaluation and model results both point to reductions in utilization—hospitalizations and emergency department visits. The evaluation uses a mixed methods approach that looks not only at a difference in difference cost analysis, but also explores themes across interviews and site visits with DCEs. There, the DCEs speak for themselves, sharing the priorities and interventions they pursued, of which three were noted: avoiding unnecessary utilization, providing additional care management for populations with complex health needs, and improving primary care. In the model results, I’m struck by the difference between some of the aggregator-ACOs and newer DCE entities who seemed to have fared better in the methodology than long-time ACOs, that are often hospital and health system based. We should all be curious to learn more about that—are aggregators picking and choosing participating providers to optimize opportunities in the financial methodology (including risk adjustment) or are they operationalizing interventions and changes in care delivery that are yielding different care outcomes and utilization patterns?  

What now? What does CMS do with information like this, and what about the ACOs? 

Melissa: This is nothing new for the Innovation Center. Failure to prove savings for Medicare in an evaluation means the program does not meet the statutory requirements to scale but CMS is accustomed to the idea that model savings take time. The two most recently announced Innovation Center models have decade long performance periods partly for that reason. CMS will apply the findings from the model operations, savings and evaluation into lessons learned as it continues evolving ACO REACH and any future Innovation Center ACO or total cost of care models.

For ACOs, news of results and evaluations land differently depending on how you did.  If your own ACO suffered losses, you may face tough questions around why and how when compared to the larger group that earned savings. If your ACO did well, your Board and executive team may have a hard time understanding the time-limited nature of the test of change—what do you mean the model ends, we saved Medicare a ton? Savvy Boards and investors will also be looking at the mix of results—if the program is a test of whether or not it’s saving Medicare overall, and evaluations say it’s not, they may wonder how much they can rely on this type of return and success over the long run.  

Go to the source(s): 

The CMS Innovation Center evaluation: Annual Report 1: Global and Professional Direct Contracting Model Evaluation (cms.gov) 

PY22 Model Results: gpdc-py2022-financial-results.xlsx (live.com) and Shared Savings and Losses and Assignment Methodology Specifications Version 10 (cms.gov)