Newtown, CT – August 28, 2015
It might be time to refresh It's The Incentives, Stupid! – Three perspectives collided this week to cause federal officials to take pause and consider their approach to payment innovation. In a hard-hitting Health Affairs blog David Himmelstein and Steffie Woolhandler repudiate CMS' claims that the readmission penalty is having a real effect and that, instead, hospitals are simply getting pretty good (or better) at cheating. In a more thoughtful NEJM Perspective, Rob Mechanic exposes some of the significant deficiencies in Medicare's proposed Comprehensive Care for Joint Replacements program. And in a KevinMD blog post, Ashish Jha exposes what appear to be serious flaws in the CMS Compare star rating of hospitals based solely on HCAHPS. While Himmelstein and Woolhandler aptly remark that a program's design can introduce bad incentives, they of course fail to suggest alternatives because, if it were up to them, we'd continue on the FFS road all the way to bankruptcy. To the contrary, both Rob and Ashish point out design problems with current or proposed Medicare incentives program and suggest some alternatives. All three pieces contribute more evidence to the importance of how incentives can and should be designed. And rule #1, just like in medicine, is First Do No Harm.
What this means to you – We know that the current payment models dominating the landscape are riddled with bad incentives, especially fee-for-service. And we also know that many quality measurement programs create a "teaching to the test" effect. Therefore, it should stand to reason for anyone introducing new models that the primary objective should be to reduce the negative impact of current programs. And yet time and again what come out of most payer organizations, including Medicare, are designs that simply fail to consider the negative incentive introduced in the package. And by negative I mean a reaction of the organizations targeted by the change that creates a potential outcome that is worse than prior to the introduction of the program. Taking Ashish's post as the first example, while HCAHPS are helpful and useful feedbacks for hospitals, it seems as if those who can afford to create a perception of good quality by providing a good experience – think about nice ambiance, friendly service – get rewarded with 5 stars. That's fine if the stars came with a strong label indicating that they only apply to the perceived experience and not the actual quality of care delivered. But they don't. Flipping next to Rob's piece, it's quite incomprehensible that Medicare would rush to introduce a program with very significant design flaws and that will almost assuredly lead to various forms of gaming from those who feel they are being unfairly treated. But there it is. And finally, the all-cause 30-day readmission penalty, while it certainly sounds good, also creates a perverse incentive because, after all, should the hospital be the only one held financially accountable for the mistakes of the entire delivery system? Does that sound fair? Does it sound fair that hospitals who spend all their capital on improving core processes of care and not customer service get a three star rating? Does it sound fair that hospitals be subject to losses on bundles simply because of a bad luck of the draw? Of course not, but these are the bad incentives being implemented or proposed by Medicare and no one should be surprised when they backfire. It doesn't mean that changing incentives is wrong, quite the contrary, but it's all in the design of the incentives, stupid!