They shouldn’t have.
Aside from the fact that the results of virtually all of the demonstrations had previously been published, the failure to reduce Medicare spending is exactly what should have been expected.
Let’s take a look at the three payment models used by CMS for the demonstrations:
1. Regular Medicare reimbursement plus a guaranteed no-risk fee or bonus for participating
2. Regular Medicare reimbursement plus a fee or bonus dependent on performance
3. Bundled payment for demonstration services
Guess what? The demonstrations subject to the first payment model were the least successful in reducing costs, with service reimbursement savings generally insufficient to balance the bonus payments, typically resulting in net spending increases. Apparently, with no financial incentive for cost reductions, the demonstration sites focused on other issues. (It’s fair to note that improving the quality of care was a key objective of almost every demonstration.)
Demonstrations subject to the second model were generally slightly more successful, with some service cost reductions, but generally not enough to compensate for the additional fees or bonus payments, even though most were at least partially tied to service savings.
Only the third payment model resulted in real cost savings, and even there a caveat is in order. The one demonstration using a bundled payment approach (for coronary bypass surgeries) resulted in a commendable ten percent reduction in spending with no apparent effect on patient outcomes. However, for the seven participating hospitals, CMS provided a significant incentive: each was named a Medicare Participating Heart Bypass Center, potentially boosting its volume of bypass surgeries. In other words, the hospitals were given the chance to increase their revenues (albeit at the expense of competitor facilities), by agreeing to accept the lower bundled payments.
The all too obvious conclusion? Providers who participate in Medicare demonstrations don’t want to experience reductions in revenue—and would much prefer to see revenues increase. CMS has clearly understood this, and has provided a sugar coating of incentives for every demonstration “pill” that might otherwise cut provider income. The result: providers who see the incentives as offering a chance of more revenue may participate, while the others won’t bother, leading to the inevitable outcome of an overall increase in Medicare spending.
As DC consultant Bob Laszewski has noted, none of this bodes well for the new Accountable Care Organization demonstrations which, just like most of the projects reviewed by the CBO, hope to cut Medicare costs by improving coordination of care, while depending on financial incentives as “sweeteners” to encourage participation. In fact, CMS’s decision to modify its initial ACO proposal to eliminate much of the financial risk, after providers complained that it was too great for the potential return, has already increased the likelihood that the ACO demonstrations also will lead to higher Medicare spending.
ACO supporters may argue that cutting provider risk is worthwhile in order to show the potential of the model, and that the Affordable Care Act allows CMS to mandate universal adoption of successful demonstration approaches. The reality is that unless the financial outcomes of the ACO demonstrations are very different from the earlier projects, ACOs will not be counted as successful, and even if they do—unexpectedly—result in significant cost savings, providers will fight tooth and nail to prevent their broader implementation. (When was the last time a Medicare demonstration approach that cut spending was implemented program-wide?)
The bottom line: Medicare policymakers’ optimism will almost always be trumped by provider self-interest.