Wins and participation rates after first dropout deadline

Almost a year ago, the Centers for Medicare & Medicaid Services (CMS) launched the Bundled Payments for Care Improvement (BPCI) Advanced model as part of its ongoing bundled payment initiative through the Center for Medicare & Medicaid Innovation (CMMI). The voluntary, episode-based program ties payments to quality performance and requires that participants bear financial risk and use certified electronic health record technology. Participants who meet the requirements can earn an Advanced alternative payment model (APM) incentive.

AAOS welcomed the new physician-led specialty opportunity within the Advanced APM track and has actively advocated for the program ever since. Through comments and discussions with both CMMI and Congress, AAOS has pushed for improvements to various aspects of the model’s design. This article offers an update on those efforts.

Delayed dropout date

One of AAOS’ initial concerns was that the application calendar was too fast-paced. Potential participants would not have adequate opportunity to consider performance data by the Oct. 1, 2018, launch. CMS recognized this challenge and decided to allow participants to drop out of the model without additional penalty through March 1. This encouraged participants to opt in with the assurance that they could later drop out of the program if they were not satisfied.

Reversed cap on physician practice reimbursement

Another issue with the model’s initial design was that it placed a cap on physician reimbursement. In the BPCI Classic model, CMMI instituted an incentive payment cap for physician and nonphysician practitioners at 50 percent of the Medicare Part B payments for the episodes. However, the cap did not extend to the physician group practice (PGP) that employed the practitioner. The PGP itself could receive net payment reconciliation amount (NPRA) savings above the aggregate cap and utilize that to offset direct and indirect costs related to the program as well as general overhead for the group.

In the BPCI Advanced model, CMMI elected to prohibit NPRA savings from flowing to a PGP working under a convening entity so its physicians could be reimbursed only up to the 50 percent cap. AAOS strongly advocated to reverse that interpretation in meetings with CMMI and by initiating a letter to CMS signed by 10 members of Congress. Then just days before the March 1 dropout deadline, CMMI announced that it would revert to its interpretation of the cap under BPCI Classic. This will allow excess NPRA funds to flow to PGPs participating in the new model.

Inclusion of outpatient TKA episodes

AAOS was also successful in getting CMS to include outpatient total knee arthroplasty (TKA) as a new outpatient clinical episode for model year three, which begins 2020. Effective Jan. 1, 2018, TKA was removed from the Medicare inpatient-only list. The policy change has economic impacts on BPCI Advanced participants, as lower-extremity joint replacement (LEJR) procedures are only considered inpatient episodes in the first round of the BPCI Advanced model.

New participation rates

Following the March 1 dropout deadline, CMS announced that about 250 participants left the program and 1,295 still remain (715 acute-care hospitals and 580 PGPs). The high retention rate has been attributed to the amendments, which are a huge win for AAOS legislative and regulatory advocacy. The BPCI Advanced model remains the only fully voluntary, physician-led Advanced APM for AAOS members and helps promote value-based care.

Ongoing concerns and advocacy

Despite the program’s improvements, AAOS remains concerned about the BPCI Advanced model’s interaction with other existing CMMI models. For example, the Comprehensive Care for Joint Replacement (CJR) model has precedence—or priority—over BPCI Advanced. This continues to be a major problem for PGPs that successfully led the LEJR episodes in BPCI Classic.

AAOS understands that CMS gave CJR precedence in the mandatory metropolitan areas to have an adequate sample size for a hospital-led mandatory model evaluation. Without any indication for policy change, however, AAOS has identified several reasons this may not be beneficial for Medicare beneficiaries or the overall success of BPCI Advanced.

Because CJR is a hospital-led model, the precedence of this model over physician-led BPCI Advanced limits specialty physicians’ opportunities to participate in an Advanced APM. Anecdotal evidence suggests that hospital systems are buying orthopaedic physician groups in several markets so they are well-positioned to lead the BPCI Advanced LEJR episodes. Of note, only hospitals and physician practices can be episode initiators in BPCI Advanced. Thus, precedence of hospital-led CJR over BPCI Advanced will be a big leadership challenge for physicians, especially on trauma cases.

The precedence also hampers quality of care for Medicare beneficiaries.

Additionally, as discussed previously, many former/BPCI Classic PGPs are in CJR markets and must participate in the CJR model. Their valuable learning, collaboration, and work to coordinate care (e.g., preoperatively optimizing patients, redesigning care plans, working with post-acute care providers, etc.)—which resulted in reduced costs to Medicare and improved outcomes for beneficiaries—will be wasted. This is likely to weaken one of the key objectives of CMMI’s value-based payment models.

For example, in Texas (a state with several CJR mandatory metropolitan statistical areas), physicians will have limited leadership opportunities to produce meaningful change in the mandated CJR model. The physicians operate in many different hospitals and health systems across their geographical area. Specifically, physicians are concerned that each hospital’s individual CJR program will create a fractured delivery system for patients. Because the surgeons in the CJR program have no risk, they are less engaged in creating meaningful change for the Medicare program.

Furthermore, because the BPCI Advanced model creates precedence for hospitalists/attending physicians over operating surgeons, Medicare beneficiaries’ care decisions are redirected to physicians whom patients did not initially choose.

Additionally, BPCI Advanced is an Advanced APM, but the qualifying thresholds are very difficult to attain. This year, qualifying participants in Advanced APMs must either receive 50 percent of their Medicare Part B payments or see 35 percent of their patients through an Advanced APM. It will be extremely difficult for specialists to reach those thresholds, and the problem is exacerbated by the aforementioned BPCI Advanced model design issues.

AAOS will continue to advocate for these concerns. The Academy is encouraged, however, by the ongoing dialogue with CMMI on these pertinent issues and its willingness to engage with the association. Through regular listening sessions, discussions, and opportunities to collaborate, AAOS seeks to improve various aspects of BPCI Advanced and other APMs.

Shreyasi Deb, PhD, MBA, is senior manager of health policy in the AAOS Office of Government Relations.