The Centers for Medicare and Medicaid Services (CMS) recently announced the Transforming Episode Accountability Model (TEAM), a new mandatory five-year model starting January 1, 2026, covering five surgical episodes that include 30 days of post-acute care.

According to CMS, TEAM aims to “improve quality of care for people with Medicare undergoing certain high-expenditure, high-volume surgical procedures, reducing rehospitalization and recovery time, while lowering Medicare spending and driving equitable outcomes.”¹ The model (and its mandatory nature) is a key part of CMS’s specialty care strategy, which has set a goal of moving 100% of Medicare beneficiaries into accountable care arrangements by 2030.

CMS Episode-Based Models 101

The CMS Innovation Center was established by the Affordable Care Act to develop and implement novel payment and service delivery models intended to improve patient care and lower costs.

Episode-based or bundled payment models are a type of alternative payment model (APM) that set a fixed price for a predefined set of services (an episode), commonly consisting of a procedure and all related services or all care for a medical condition.

Participation in these models can be:

 •  Voluntary: Any organization that meets certain criteria can opt to participate in the model for a specific length of time.
 •  Mandatory: Specified organizations, typically those of a certain type (such as hospitals) in selected geographic regions, are required to participate in the model, usually for the entirety of the program.

Participant performance in the model is valuated against quality and financial benchmarks as one of the main goals of these models is to improve quality of care while reducing unnecessary spending.

Model design can be:

 •  Prospective, which involves setting prices up-front for an episode and concurrently identifying eligible patients.
 •  Retrospective, which involves identifying episodes based on paid claims and reconciling episode costs against targets after the episodes have ended.

Read more

Specifically, CMS is looking to provide continuity in bundled payment/episode-based models given that the Comprehensive Care for Joint Replacement (CJR) and Bundled Payments for Care Improvement Advanced (BPCI Advanced) models are scheduled to end December 31, 2024, and December 31, 2025, respectively. TEAM will be a blend of these models, building on the components of each model that CMS views as successful and refining program methodologies based on lessons learned from recent program experience. Additionally, TEAM will provide “support for safety net providers and underserved beneficiaries, ultimately helping foster a complementary relationship between accountable care organizations (ACOs) and hospital participants.”² Please refer to the Milliman article “What Are Bundled Payments and How Can They Be Used by Healthcare Organizations?” for more information on the specifics of bundled payment programs and episode-based models.³

TEAM was included in the fiscal year (FY) 2025 proposed Hospital Inpatient Prospective Payment System (IPPS) rule, CMS-1808-P.⁴ Unless otherwise noted, all information contained here describing TEAM is drawn from that proposed rule. Any comments on the proposed rule are due to CMS by June 10, 2024. We expect the final rule to be released in August 2024.

Who will be required to participate in TEAM and how will they be impacted?

Mandatory model and participation

CMS has in the past tested a variety of episode-based models—the majority of which have been voluntary. Voluntary models are easier to implement and more flexible because they can be implemented without rulemaking; however, they are harder to evaluate due to selection bias and provider attrition. On the other hand, a mandatory model can help prevent (or at least minimize the impact of) these issues while furthering CMS’s accountable care participation goals. Using the mandatory CJR model as a foundation, CMS will be expanding the model’s geographic scope with a focus on CBSAs. This structure will result in selection of a broader set of participants who are more representative of the U.S. population. CBSAs include both micropolitan and metropolitan statistical areas (mSAs and MSAs, respectively) and reflect urban core populations of at least 10,000. In CJR, CMS relied on MSAs, which reflect urban core populations of at least 50,000. CBSAs will be stratified into 17 cohorts based on the following criteria and assigned a selection percentage:

• Average historical episode spending
• Number of hospitals
• Number of safety net hospitals
• Past exposure to CMS bundled payment models

This stratification methodology will allow CMS to “oversample CBSAs with low past exposure to CMS’ bundled payment models and CBSAs with a high number of safety net hospitals” (pg. 36411). Each CBSA will have a 20% to 50% chance of selection, which will result in approximately 25% of the 800 eligible CBSAs being selected for participation in TEAM.

CMS will require all ACHs within the selected CBSAs to participate, including safety net, rural, Medicare-dependent, sole community, and essential care access hospitals. Historically, these types of hospitals have participated in value-based payment arrangements at lower rates due to a lack of resources or concern with taking on substantial financial risk (pgs. 36388, 36390, 36393).

As only ACHs can initiate episodes and all ACHs in the selected CBSAs (defined based on its CMS Certification Number) will be TEAM participants, episode attribution will be simpler than voluntary models such as BPCI Advanced, which allowed both ACHs and Physician Group Practices (PGPs) to sign up as participants and initiate episodes. The episode will be attributed to the ACH that performs the episode-initiating procedure. However, participants will be able to seek administrative or convener support if they choose.

As one goal of TEAM is to expand the reach of value-based care to providers that have yet to participate, CMS acknowledges that these providers will need time to adapt their infrastructures and become familiar with the components of the model. Therefore, CMS will be providing a one-year glide path that will ease participants into taking on downside financial risk in TEAM, while allowing certain eligible participants to opt for limited downside financial risk for the remainder of the model.

All TEAM participants will have the option to choose between Tracks 1 and 3 in the first performance year (PY), and eligible participants will have the option to choose between Tracks 2 and 3 throughout the rest of the program, as illustrated in Figure 1.

Participants will be required to notify CMS of their requested track prior to the start of a PY, with the default tracks being Track 1 for PY 1 and Track 3 for PYs 2 to 5. All participation tracks will qualify as a Merit-Based Incentive Payment System (MIPS) APM under the Quality Payment Program, and Tracks 2 and 3 will also qualify as an Advanced APM.

Why this matters: CMS will be mandating participation in more geographic regions under TEAM than in CJR, so the model will impact more ACHs, both those familiar and unfamiliar with value-based care. However, the mandated ACHs will have one year to prepare before the model starts in 2026 and the PY Track 1 option will help providers new to value-based care get up to speed before taking on downside financial risk.

Episode selection and length

CMS reviewed the clinical episodes included in the BPCI Advanced model to determine which episodes to test in TEAM. While BPCI Advanced includes both surgical and medical episodes, for TEAM CMS chose to focus on surgical episodes as they “are time-limited with well-defined triggers, have clinically similar patient populations with common care pathways, and have sufficient spending or quality variability, particularly in the post-acute time period, to offer participants the opportunity for improvement” (pg. 36413).

After evaluating 2021 BPCI Advanced experience, CMS chose to include the following five episodes in TEAM:

• Lower extremity joint replacements (LEJR)
• Coronary artery bypass grafting (CABG)
• Surgical hip and femur fracture treatment (SHFFT)
• Spinal fusion (FUSION)
• Major bowel procedures (BOWEL)

CMS believes these episodes reflect the best opportunities for improvement as they are high-volume and high-cost, with post-acute spend ranging from 27% to 63% of total episode costs.

Figure 2 illustrates the proportion of spending during the procedure that initiates the episode (anchor) versus the 30-day post-acute period for the 2021 BPCI Advanced episodes.

Using the same 2021 period, CMS estimated national volume and average costs for all TEAM episodes, shown in Figure 3. Altogether, these episodes represent approximately 1.9% of the total Medicare fee-for-service (FFS) spending (Parts A and B only) in 2021, whereas the corresponding 90-day BPCI Advanced episodes accounted for approximately 2.1%.

Similar to both the BPCI Advanced and CJR models, these episodes are triggered by a surgical procedure that takes place during either an inpatient stay or an outpatient encounter, as identified using the Medicare severity diagnosis-related group (MS-DRG) and Healthcare Common Procedure Coding System (HCPCS) codes listed in Figure 4.

Performing one of these procedures will initiate an episode, whereby the participant ACH will be held accountable for the Part A and B costs associated with the surgery (anchor) and the subsequent 30 days of post-acute care. The first day of the post-acute period is the anchor discharge date for inpatient admissions or the day of the procedure for outpatient surgeries. Like BPCI Advanced, TEAM episodes will include nearly all care that occurs in the post-acute period—there are only a handful of excluded services, including admissions for oncology, trauma, and organ transplantation, as well as high-cost Part B drugs and hemophilia clotting factors. This may mean that participants are exposed to risk that is not related or only loosely related to the initiating procedure.

The 30-day post-acute period of TEAM episodes will be significantly shorter than the 90 days included in both BPCI Advanced and CJR. CMS will be shortening the episode length in TEAM because it has found, through analyzing BPCI Advanced episodes between 2020 and 2022, that 75% of post-acute episode spend occurs in the first 30 days of the post-acute period while 90% occurs in the first 60 days. Shorter episodes will also help limit spending for unrelated and chronic conditions that tend to occur in the later stages of longer episodes and minimize potential ACO overlap.

Enhancing the relationship between episode-based models and ACOs appears to be a deciding factor in the episode length, with CMS noting “A 30-day episode would position the specialist as the principal provider near the anchor event with a hand off back to the primary care provider for longitudinal care management and we believe that ACOs are better equipped to address the population health needs of Medicare beneficiaries” (pg. 36419).

Why this matters: TEAM participants will be held responsible for surgical episodes spanning multiple specialties with varying volumes, care patterns, and areas for efficiency. The broader scope and limited number of excluded services from the episodes will allow for more combined potential savings in terms of episode mix; however, the shorter episode length will limit post-acute savings opportunities.

How is performance measured under TEAM?

As a central goal of TEAM is to improve the quality of care while reducing unnecessary spending, participant performance will be evaluated against both quality and financial benchmarks.

Quality measures

Through the quality component of episode-based models, participants have the opportunity for better financial rewards by focusing on their quality performance or quality improvement. CMS will be selecting the following quality measures, which highlight care coordination, patient safety, and patient-reported outcomes (PROs) and will be relevant for the applicable episodes:

• All TEAM episodes
  • Hybrid Hospital-Wide All-Cause Readmission Measure With Claims and Electronic Health Record Data (CMS Measure Inventory Tool [CMIT] ID #356)
  • CMS Patient Safety and Adverse Events Composite (CMS Patient Safety Indicator [PSI] 90, CMIT ID #135)
• LEJR Episodes
• Hospital-Level Total Hip and/or Total Knee Arthroplasty (THA/TKA) Patient-Reported Outcome-Based Performance Measure (PRO-PM, CMIT ID #1618)

These quality measures will be consistent with those used in current and ongoing models, while being meaningful metrics for the acute surgical procedures in TEAM.

Individual measure performance will be combined during reconciliation to create a composite quality score (CQS), which will adjust the reconciliation payment up to 10% or repayment amount up to 15% (depending on participation track) with higher quality resulting in better financial outcomes.

Why this matters: CMS will include quality measures consistent with other models to reduce the burden on participants. While there will be an additional measure that is only applicable to LEJR episodes, the measure will align with the upcoming 2027 mandatory reporting for THA/TKA PRO-PM, which will begin prior to the start of TEAM.⁵

Financial

CMS has tested multiple payment methodologies in prior episode-based models, ranging from simple to complex. The goal of the TEAM payment methodology is to combine the most successful components of past models to provide predictable, accurate, and interpretable prospective target prices.

Target prices

The foundation of an episode-based payment methodology is the benchmark data used for the creation of the target prices against which participants will be held, in this case a historical baseline period. There are challenges in choosing an appropriate baseline period. An older, fixed baseline may fail to accurately represent updated performance care and spending patterns, but updating the baseline period over time may cause target prices to decrease too quickly, potentially penalizing high-performing participants. TEAM will balance these challenges by using a three-year weighted baseline period that rolls forward annually. Each individual year of the three-year baseline period will be weighted differently, with higher weights for the more recent years that may better predict performance spending.

The first baseline period will include episodes starting between January 1, 2022, and December 31, 2024, and reflect the weighting illustrated in Figure 5.

Average spending in the three-year weighted baseline period at the MS-DRG/HCPCS and U.S. Census Division levels, capped at the 99th percentile, will result in 216 standardized, regional target prices.

The following factors will be applied to the standardized, regional target prices to create the preliminary target prices, which will be provided to participants prior to the start of each PY in November:

• A prospective trend factor, calculated as the percentage difference between the average regional MS-DRG/HCPCS episode type spending in the last year of the baseline period compared to the first year of the baseline period
• A prospective normalization factor to remove the overall impact of the risk adjustment variables on the national target price based on the last year of the baseline period, with limited retrospective modification at reconciliation to account for the observed case mix in a PY
• A discount factor of 3%

Risk adjustment

The preliminary target prices will provide a solid foundation for predicting episode spending in the performance period; however, they will require risk adjustment to account for variation due to each beneficiary’s unique characteristics. Built upon the risk adjustment methodology used in the CJR three-year extension, TEAM will employ a risk adjustment model based on the following beneficiary characteristics:

• Age bracket: Based on the beneficiary’s age as of the first day of the episode.
• Hierarchical Condition Category (HCC) count: Based on a 90-day look-back period (similar to BPCI Advanced).
• Social risk: Based on dual eligibility, area deprivation index, or Part D low-income subsidy eligibility.

The resulting risk adjustment multipliers calculated using the baseline period will also be broken out at the MS-DRG/HCPCS episode level. They will be provided prospectively and remain fixed at reconciliation, so participants will be able to proactively estimate the finalized target prices prior to reconciliation.

Reconciliation

CMS has tested various reconciliation processes through prior models and has found that there is a need to balance timely feedback on performance, while limiting the administrative burden. Based on its experience, there will be a single annual reconciliation process for TEAM. Mirroring the current CJR process, episodes in a PY will be reconciled once annually in the fall, allowing for six months of claims run-out following the end of a PY. CMS has found through analysis that most claims are received and final within six months of the date of service.

Why this matters: CMS’s learnings from prior models and participant feedback translates to a payment methodology that it believes focuses on predictability and accuracy. Mostly prospective target prices with an easily interpretable risk adjustment model will likely provide participants with more stable targets as they strive for financial success in TEAM, according to CMS. With target prices built upon regional spending and risk adjustment to account for patient acuity differences, some ACHs may already be below the regional target without making changes while others may be well above the regional target and will have to make significant changes to achieve the required savings. Additionally, purely regional trending means that regions that perform well may see their target prices decrease in future PYs. Finally, the 3% discount factor for the target price equates to between 6% and 18% of post-acute spending nationally, depending on the episode type. This presents a challenge for participants, as post-acute spending is the primary avenue for reducing episode spend within the TEAM specifications.

Downstream provider financial arrangements

TEAM will be a flexible model, allowing ACH participants to engage in financial arrangements to encourage success in the program. Participants will be allowed to share their reconciliation payments or repayments with providers and suppliers, including Medicare ACOs, which contribute to the ACH participant’s performance. These sharing arrangements will require documentation and be subject to several criteria including a quality provision.

Why this matters: While the ACH participant will ultimately be responsible for the cost of an episode, it may not have significant authority over care in the post-acute period. Financial arrangements could allow participants to incentivize downstream providers to focus on optimizing care patterns and reducing unnecessary spending to achieve better model performance.

What support should potential participants consider for TEAM?

Potential TEAM participants, those both familiar with and new to value-based care, should carefully consider their internal capabilities and seek outside support where they see resource limitations. Succeeding in TEAM will not be solely about the participating ACH’s quality of care—it will require managing relationships with other providers and understanding how their performance compares to other hospitals.

Potential TEAM participants should consider reaching out to organizations that can provide the following support:

• CMS bundled payment expertise: CMS continues to build upon its prior models, incorporating and improving past methodologies and measures. Organizations well-versed in these models can help participants navigate the process while providing insights for achieving success.
• Analytics: CMS models provide raw episode- and claim-level data to participants on a regular cadence. Analytical support can help transform the detailed CMS data into reports or interactive tools that provide visibility into program performance in a user-friendly format.
• Benchmarks: As key components of the target price methodology are derived by CMS regionally and nationally to achieve policy objectives, it is beneficial to have access to post-acute utilization benchmarks to determine where your hospital ranks among your peers and to identify potential opportunities and risks.

Closing thoughts

TEAM will impact a wide array of ACHs in about a quarter of CBSAs across the United States. These organizations, both those familiar with and new to episode-based models, might not be able to fully realize opportunities and synergies that can be leveraged within the model.

Data alone will not change behavior or solve any inherent issues within a procedure or an ACH. Understanding the data, uncovering possible opportunities, and firmly committing to instituting change are all necessary components of success. Milliman has extensive subject matter expertise and data-driven tools for episode-based programs to help highlight an organization’s areas for quality improvements and cost savings. For more information, contact your Milliman consultant.

---

¹ CMS. Transforming Episode Accountability Model: Overview Fact Sheet. Retrieved May 28, 2024, from http://www.cms.gov/files/document/team-model-fs.pdf.

² Health Affairs Forefront (April 2, 2024). The CMS Innovation Center’s Strategy to Support Person-Centered, Value-Based Specialty Care: 2024 Update. DOI: 10.1377/forefront.20240328.868596.

³ Bazell, C., Alston, M., Pelizzari, P.M., & Sweatman, B.A. (March 27, 2023). What Are Bundled Payments and How Can They Be Used by Healthcare Organizations? Milliman. Retrieved May 28, 2024, from http://www.milliman.com/en/insight/what-are-bundled-payments-and-how-can-they-be-used-by-healthcare-organizations.

⁴ The full text of the proposed rule, 89 FR 35934, Is available at http://www.federalregister.gov/documents/2024/05/02/2024-07567/medicare-and-medicaid-programs-and-the-childrens-health-insurance-program-hospital-inpatient.

⁵ CMS. THA/TKA PRO-PM. Retrieved May 28, 2024, from http://qualitynet.cms.gov/inpatient/measures/THA_TKA.