Newtown, CT – Aug 31, 2012
Ready? Set?….Whoa, hold on, and think before you sign on the dotted line – Two months ago, at the end of June, hospitals, health systems and other organizations submitted applications to the CMMI for the Bundled Payment for Care Improvement (BPCI) pilot. The program was launched a year ago, and, since then, hundreds of potential applicants have received two years' worth of Medicare claims for their market area, and performed various analyses to determine if they want to file a formal application. Many walked away after pouring through the numbers, but many stuck through the process. Today we release an important report that summarizes significant issues in the design of the BPCI. Taken together, they can lead to applicants or Medicare winning or losing simply based on the luck of the draw. The amount of probability risk introduced in this pilot, if the design is not modified (or certain conditions and terms aren't negotiated in the agreements between CMS and the applicants), might very well cause evaluators on the back end to conclude that the pilot's effects on reducing costs of care are uncertain. That wouldn't simply be a shame; it would be a phenomenal setback for the value-based purchasing movement. And it doesn't have to be.
What this means to you – our report identifies several problem areas. First, case mix shifts can cause providers or Medicare to realize gains or losses, irrespective of any clinical improvement. That's because, so far, CMMI has refused to allow applicants to distinguish the bundled price based on the underlying diagnosis of the patient, or based on the specific procedure performed. And that's contrary to the principles of bundled payments and all the implementations done in the private sector. Second, high and low cost outliers are included in the bundled price, and a shift in the number of those cases can also lead to gains or losses irrespective of clinical improvement. There are easy solutions to mitigate the outlier problem, and we propose a few that are pretty standard in private sector bundled payment contracts. Third, and unique to academic medical centers, the exclusion of mission-related add-ons from the episode price can cause a net reduction of these subsidies as the centers reduce the number of readmissions and improve cost and quality. This one falls into the category of "no good deed goes unpunished," and can also be mitigated.
So the good news is that these design problems are all solvable if CMMI is willing to solve them. As such, applicants should insist on fixes before signing on the dotted line, and officially joining the pilot. We've said before that this pilot has the potential to fundamentally shift the value of US health care in a positive direction, and we say it again…provided the design flaws are corrected.
Francois de Brantes
Health Care Incentives Improvement Institute, Inc.