BPCI-Advanced: The Race Is On for Episode Selection

Sg2 Princpals Kevin Lieb and Chris McBride contributed to this post.

As we say goodbye to summer and gear up for a busy autumn, CMS’s recent release of claims data related to the Bundled Payments for Care Improvement Advanced (BPCI-Advanced) program has increased anxiety levels across the country. Providers currently participating in, or considering entering into, the BPCI-Advanced program for Model Year 3 face the daunting task of culling 3 years of claims data in search of opportunities to reduce Medicare spend over 90+ days of patient care.

With BPCI-Advanced classified as a voluntary alternative payment model, these data are provided to help participants determine which episodes they wish to enter into. That decision is critical, principally because BPCI-Advanced is a risk-based program inclusive of both upside and downside risk. Episode selection significantly influences financial performance and means the difference between receiving a positive NPRA (net payment reconciliation amount) payment from Medicare or writing a check to CMS to cover spend above the target price. Hence, careful and informed analysis of these data is essential for a participant’s decision to enter the program. This is true whether the participant is “going it alone” or is “under the wings” of a convener. When working with a convener, hospitals and physicians should have an unbiased eye toward episode selection. Such objectivity enables meaningful, informed discussions with the convener, who may recommend an episode that makes the hospital or physicians uneasy for quantitative or qualitative reasons.

Two major data challenges exist relative to BPCI-Advanced.

1. The sheer amount of data provided for analysis: specifically, 3 years of claims data and target pricing sheets encompassing both peer and risk adjustment factors

2. The extremely short time frame between the receipt of data and the episode selection deadline. CMS’ BPCI-A Model Team has not yet provided the 60-day window for clinical episode selections. Providers must complete their build-out of the data files in short order to ensure ample time for discourse among all parties.

Analysis Paralysis

Sg2’s recommended approach to data analysis entails starting at a macro level for each episode by breaking down Medicare spend in each area. As many providers will attest, reducing spend in episodic models such as BPCI-Advanced starts by focusing on care coordination after the patient is discharged from the hospital. Therefore, prime episodes for consideration are those where at least 50% of the overall spend occurs outside the 4 walls of the hospital. Sg2 recommends deselecting episodes where the majority of spend is directly related to the MS-DRG or procedure payment (for outpatient episodes)—a variable you can’t reduce.

Next, align the remaining episodes around surgical/medical conditions or, if your organization is service line–centric, around specific service lines (eg, cardiac, orthopedics, spine, general surgery). Then, break down the episodes into concise buckets of spend (again, setting aside the MS-DRG payment). Data analysis will highlight specific areas of opportunity—as well as risk—regarding key performance indicators, including all post-acute spend buckets that CMS will be holding BPCI-Advanced participants accountable for—of which there are many.

  • For surgical cases, focus on key areas of care and spend, such as readmissions and causes, skilled nursing facility (SNF) utilization and length of stay, and use of inpatient rehab (see episode data sheet below for an example).
  • For chronic conditions, be certain to include emergency department use for follow-up care vs alternative care settings, patient follow-up with the primary care providers or care managers within 24 hours to 7 days of discharge, and social determinants of health relevant to patients and their families.

HH = home health; IRF = inpatient rehabilitation facility; LTCH = long-term care hospital; TP = target price.

Time Is Not on Your Side

CMS’ BPCI-A Model Team has established December 1st as the deadline for clinical episode selections and participation terms, which means participants have little time to comprehensively analyze the data and make decisions. These data must be downloaded, compiled and returned to providers in a digestible format within days to enable as much time as possible for discourse. Opportunity for thorough deliberation allows the organization and physicians to commit to episodes for which risk has been carefully evaluated and the likelihood of success is high. This quantitative analysis must then be paired with the qualitative aspects of the program, including key components such as:

  • Performance across the quality metrics
  • Accountable care organization interaction
  • Hospital-physician alignment
  • Patient attribution across the market
  • Value-based care infrastructure

Due to this complexity, a digestible set of data should be provided to leadership within 72 hours of receipt—something that Sg2 can commit to and provide to your organization.

Final Thoughts

If your organization is considering entering into BPCI-Advanced as a system, or minimally with multiple facilities as episode initiators, a careful evaluation of the data is essential to establish where common, effective approaches to reducing Medicare spend can be deployed, as well as where significant differences in care approaches exist across markets or the system. Moreover, the data provide insight into how various post-acute providers perform, thereby enabling the initial exploration of post-acute care networks to help ensure BPCI-Advanced success.

Sg2 has significant experience in looking at data such as these from a system-level perspective and helping participants make sense of the data to highlight opportunities—as well as potential deficiencies—at both an individual and group level. If you have any questions or you would like to discuss options for how Sg2 can support your organization in this extremely important process, please contact me or

Additional BPCI-Advanced Resources

  • Share
  • Follow Sg2 on Twitter
  • Connect with Sg2 on LinkedIn

Tags: , , , , , , , , , ,