What to Look For In BPCI Advanced

by Jonathan Pearce and Darcie Hurteau

Many current and potential bundled payment participants are eagerly awaiting the announcement by the Centers of Medicare and Management Services (CMS) of the revised Bundled Payment for Care Initiative (BPCI)" program. This “BPCI Advanced” program has been planned by CMS for many months, but announcement of the details has apparently been stalled somewhere within the government.

This raises concerns for the many current BPCI participants are hoping to continue their participation in an ongoing program, as well as the hospitals and physicians who didn’t participate in BPCI, but may now wish to join the program. The attractiveness of this program has yet to be seen, and it will depend on multiple factors. Here’s our list of the biggest programmatic questions to ask once the changes are published:  

1.   How are episodes defined? 

In the current BPCI program, episodes are defined by the diagnosis-related group (DRG) on the "index" admission, in some cases stratified by the presence of a fracture diagnosis for major joint replacement episodes. In the CMS Oncology Care Model (OCM), episodes are identified by the presence of the cancer diagnosis code on primary physician claims. And, while CMS hasn't yet implemented outpatient- based episodes, it’s possible that BPCI Advanced will include such episodes.

One issue of significant interest to hospitals is the ability to identify episode-eligible patients within their patient population if the episode definitions involve complex combinations of DRGs, diagnosis codes, and other factors. In the BPCI program it‘s relatively easy to identify participating patients, but as definitions become more complex the difficulty of identifying these patients may increase.

2.   Who are the participants?

One of the biggest points of contention in BPCI has been the question of who "owns" the episodes--who is financially responsible for surpluses and deficits? In the BPCI program, both physicians and hospitals could participate in "Model 2" episodes, which are initiated by a hospital stay, while post-acute providers could participate in "Model 3” episodes that are initiated by post-acute service. In the comprehensive care for join replacement (CJR) program, only hospitals participate, and hospitals were the only participants in the now-rescinded final rule for the Episode Payment Models (EPM)program.

Many observers, however, believe that physicians will be able to own episodes in the BPCI Advanced program. This belief appears to originate from the fact that BPCI Advanced is expected to be similar to BPCI. In addition, participation in a BPCI program enhances the ability of physicians to actively participate in advanced alternative payment models (AAPM), which improve their financial opportunities under MACRA.

An additional question, if physicians are allowed to own episodes, is whether non-provider organizations can act as "awardee conveners" as in BPCI. An awardee convener assumes the financial risk for the episode, and contracts with hospital or physician providers whose services initiate the episode. While most hospitals can absorb the financial risk of episodes in which they participate, many physician groups do not have the financial resources to withstand the significant variations in surplus and deficit that can occur in many episodes; therefore the convener creates a financial buffer between the physician and the actual financial results of the episode.

Finally, will a "Model 3" exist to allow participation by post-acute providers? Many post-acute providers in BPCI to may wish to continue participation in bundled payment initiatives, and this ability will be predicated upon allowing such providers to initiate episodes.

3.   How are targets established, and what are the effects for each provider?

How episode cost targets, from which surplus and deficit are derived, is critical information for potential participants. CMS has used two primary approaches to establishing targets in bundled payment programs: those based on a participant’s historical costs during a baseline period, and those based on the episode costs of other providers within a geographic region.

Historically-based targets reward lowering cost, while regional targets reward existing cost-efficiency; providers with historically high costs prefer historically-based targets, while those who are already cost-efficient prefer regional targets. The BPCI program used entirely historical targets, while the CJR program transitions from primarily historical targets to entirely regional targets throughout the five-year participation period. Since BPCI Advanced is expected to be voluntary, the derivation of targets will significantly affect the types of providers who will choose to participate.

4.   How will quality metrics be defined?

In the current BPCI program, quality metrics are not used to affect financial performance; however, in the CJR program, a participant’s quality scores determine the CMS discount applied to the episode targets. Participants with high quality receive a higher target, while participants with low quality may be denied payment of any surplus that they may have earned. The selection of quality metrics, and their applicability to payment, will be a significant factor in the BPCI 2.0 program.

One of the challenges participants in CJR and OCM know well is the data lag on their performance metrics. In the current models, participants receive information based on data that is three to six months old, which creates a lag in how they are adjusting performance and workflows, and means they may face penalties without being able to do anything about them. Hopefully, BPCI Advanced will improve on how quality metrics are defined and used—and set up more effective reporting timelines

5.   What will the gainsharing rules be—and how will they affect current BPCI participants who continue into BPCI Advanced?

In both BPCI and CJR, participating hospitals were allowed to share financial surpluses with physicians and other providers who were directly responsible for creating those surpluses. Where implemented, these gainsharing programs were often highly successful in creating the physician engagement necessary to effect change. Therefore, the presence of gainsharing programs may be a significant factor in the decision of a hospital to participate in BPCI 2.0.

With the BPCI program ending in 2018, participants would be faced with the discontinuation of gainsharing programs, the requirement to delete all of the data and derivative products from that program, and the loss of ongoing data. Thus, many participants may wish to continue the activities that they initiated as part of BPCI. The extent to which this is possible will be determined by rules governing gainsharing, data retention, and other aspects of the program. What will be very interesting to see is if participants can keep the data and continue in their existing episodes.

6.   What will the evaluation and participation periods be?

In the BPCI program, participants were allowed to receive their historical episode data and evaluate participation options for almost two years before being required to commit to participation in certain episodes. Following that, they were allowed an additional year during which they could add or drop episodes; after that date participants could drop episodes but could no longer add them.

In the BPCI Advanced program, participants will need sufficient time to receive their historical episode data, identify current care patterns and costs, and begin to develop participation strategies. Their ability to be flexible in participation, terminating episodes in which they cannot be successful and adding additional episodes, and applying experience in dealing with bundled payments, will be critical to their selection to participate in this program.

The initial term of the BPCI program was to be three years; however, CMS extended it to allow early participants to remain in the program through the second quarter of 2018. The CJR program runs for five years.

Regardless of how the BPCI Advanced changes play out in practice, it’s a fair bet that day-to-day operations for participants won’t be the same. Once CMS releases BPCI Advanced, keep a close eye on how these operational questions are addressed. Once the program is announced we’ll be posting our thoughts and opinions on how the above issues are implemented.

Jonathan Pearce is a Principal with Singletrack Analytics. Darcie Hurteau is a Director with DataGen. Singletrack and DataGen collaborate to create the BPCI 360 and CJR 360 family of analytics tools for Medicare bundled payment programs.