Here We Go Again
Newtown, CT - January 12, 2018
Several years ago we offered some guidance to CMS about key design elements that would increase the success of an APM. It’s time to revisit them – with the announcement of BPCI Advanced, Medicare just sent a strong signal that alternative payment models, and especially those that physicians participating in MACRA can opt for, are here to stay. That’s certainly consistent with our own analysis drawing from our experience in the private sector, and good news for all those fretting that market momentum for APMs would wither.
More than five years ago we published an Issue Brief that described some of the flaws in the original BPCI design, including, chief among them, a lack of severity adjustment. When CMS published proposed rules for expanding the mandatory CJR and introducing new mandatory cardiac bundles, we raised the same issues. And here we go again.
BPCI Advanced has some positive aspects. First, it expands the program to outpatient procedures, thereby delinking the program from acute care facilities and recognizing the importance of moving this type of APM to other sites of care. Second, it provides physicians with the opportunity to be episode initiators and to control the episode, thus shifting the locus of control, consistent with comments made by Seema Verma in a WSJ editorial this past September.
BPCI Advanced also has some flaws, and chief among them is the continued lack of adjusting for patient characteristics, even though the current documents do refer to patient case-mix adjustments. This will now be all the more important that, according to the description of the program posted on the CMS website, all professional services (Part B) will be included in the post-acute or post-procedure period, apart for a few exceptions. As such, the impact of patient co-morbidities that we had clearly demonstrated more than five years ago will be even more exacerbated. For example, patients with cancer who are included in any of the targeted episode would have their cancer costs included even though those costs could be completely irrelevant to the management of the episode that was initiated. Of course, all of the historical costs for these patients would be included in the baseline calculations, and that leaves the door wide open for cherry-picking. In fact, it would be shocking if participating providers didn’t do their very best to cherry-pick patients while participating in BPCI Advanced. It’s the easiest way to win. Making these programs mandatory solves that problem but, as we’ve said before, mandating a poorly designed program is not a viable solution. And while case-mix adjustment will capture a small amount of reasons for resource use variation, it isn’t an adequate substitute for a robust severity-adjustment model.
There are other flaws in the BPCI Advanced program. Including AMI as a target episode could encourage clinicians to hospitalize patients with mild MIs instead of encouraging medical management. Target prices will apparently be calculated for each facility or physician group based on their historical performance. And it’s not clear that those who are most efficient today will get credit for that greater efficiency or would, to the contrary, be penalized for already being better.
Some of these issues will get clarified in the weeks to come, but there isn’t much time for applicants to make sure that they will have the protections they deserve if they’re to take on significant financial risk, given that applications are due in 60 days. Not only does this seem rushed, it also seems set up for failure, at least for those who won’t be smart enough to figure out how to stack the odds in their favor. APMs shouldn’t be about gaming, they should be about better quality care at a lower total cost. Yet how can one focus on the real objective when the potential random variation of uncontrollable patient costs can wipe you out? We’ll be looking closely at the details as they come out, as others will, and we’ll be hoping that these details alleviate our concerns. But if they don’t, you’ll know soon enough.