Bundled Payments for Care Improvement Advanced (BPCI Advanced)

Analysis updated November 9, 2026

Strategic questions for your team

Given that BPCI Advanced requires downside financial risk from the outset — with reconciliation payments capped at +/-20% of the volume-weighted sum of final Target Prices across all Clinical Episodes at the EI level, and a 30-day Post-Episode Spending Monitoring Period that can trigger additional repayments — does our organization have sufficient cash reserves and financial buffers to absorb worst-case repayment scenarios without materially impairing operating margins?
Across the 8 Clinical Episode Service Line Groups encompassing 29 inpatient, 3 outpatient, and 2 multi-setting Clinical Episode categories, which specific CESLGs represent our highest episode volume, and do our historical Medicare FFS expenditures for those episodes run sufficiently above peer-group benchmark spending to create a realistic margin between our actual costs and the Target Price after the CMS discount (2% for medical episodes, 3% for surgical episodes) is applied?
The BPCI Advanced final evaluation confirms that the primary driver of the model's ~$800 million in net Medicare savings was a 5 percentage point reduction in PAC facility discharges for surgical episodes and approximately 3 fewer SNF days per episode — do we have the discharge planning infrastructure, care navigator capacity, preferred PAC network relationships, and physician engagement needed to replicate these care redesign strategies at scale within our episode volume?
Because Target Prices are calculated using a rolling baseline period that shifts forward each Model Year (MY8 baseline: October 1, 2017–September 30, 2021), and because participants that previously reduced PAC utilization under BPCI, CJR, or ACO programs report being 'penalized' by lower preliminary target prices, has our organization already captured the low-hanging fruit in PAC reduction, and can we credibly identify further efficiency opportunities that will keep actual episode spending below the discounted target price year-over-year?
Given that Clinical Episode attribution follows a strict precedence hierarchy — first to the PGP whose attending physician NPI appears on the claim, second to the PGP whose operating physician NPI appears, and third to the ACH where the Anchor Stay or Anchor Procedure occurred — have we mapped exactly which physicians in our network are affiliated with competing PGP participants, and do we understand which episodes would be attributed away from our facility, reducing our volume and potentially distorting our risk profile?
The model excludes Medicare Advantage enrollees, ESRD beneficiaries, United Mine Workers beneficiaries, and beneficiaries for whom Medicare is not the primary payer — and beneficiaries aligned to an ACO REACH entity cannot trigger BPCI Advanced Clinical Episodes — so what percentage of our actual Medicare FFS inpatient discharges and outpatient procedures in our target CESLGs will be excluded under these rules, and is the remaining eligible episode volume large enough to generate statistically meaningful financial results and justify the administrative investment?
BPCI Advanced requires the use of Certified Electronic Health Record Technology as one of the three criteria for Advanced APM qualification under the Quality Payment Program — and quality performance via the Composite Quality Score (CQS) can adjust reconciliation payments by up to 10% — do we have the EHR infrastructure, data analytics capabilities, and registry reporting capacity to select and perform competitively on the Administrative or Alternate Quality Measures Sets (up to 5 measures per Clinical Episode), and have we modeled the financial impact of our likely CQS score on our reconciliation amounts?
The final evaluation found that 80.9% of BPCI Advanced hospitals and 76.7% of PGPs participated under a convener — with conveners bearing or sharing financial risk and providing data analytics and case management support — should we participate as a Non-Convener Participant bearing full financial risk independently, or does our organization's data infrastructure, care management maturity, and risk tolerance make partnering with a convener the more prudent structure, and what contractual protections and financial arrangements would we require from a convener?
Given that hospital ownership of IRFs creates a directly conflicting financial incentive — the evaluation explicitly notes that participants owning IRFs struggled to reduce IRF utilization and some avoided selecting the neurological care CESLG specifically because of high IRF discharge rates — how will we govern the tension between BPCI Advanced's incentive to minimize costly IRF stays and our IRF's need to maintain daily census, and have we quantified the net financial impact of each scenario on a consolidated basis across both entities?
With the BPCI Advanced model period of performance ending December 31, 2025, and the mandatory Transforming Episode Accountability Model (TEAM) — which covers five high-volume surgical procedures and launched January 1, 2026 — already underway for hospitals in selected geographic areas, does participation in the final years of BPCI Advanced (or prior participation history) provide meaningful operational learning, physician alignment, and PAC network infrastructure that will position us competitively under TEAM or any successor mandatory bundled payment model the Trump administration's CMS may advance?

Traps and concerns

Benchmark Drift Penalizes Early Efficiency Gains
BPCI Advanced's baseline period advances forward every model year — from Jan 2013–Dec 2016 for MY1-2, all the way to Oct 2017–Sep 2021 for MY6 — meaning participants who aggressively reduced PAC utilization in early years effectively wrote their own lower target prices into future reconciliation cycles. The 7th Annual Evaluation confirmed this dynamic explicitly: participants that had already lowered spending prior to joining (e.g., prior BPCI Initiative or CJR participants, ACOs) struggled to find further efficiencies and exited rather than risk owing repayments to CMS. This 'fighting against yourself' trap contributed to substantial participation attrition following the MY4 design changes, with surgical CESLG participation among hospitals dropping from 57.6% to 41.3% by MY8. Organizations that achieved deep early savings should model a multi-year 'target price decay' scenario and stress-test whether their residual efficiency headroom — particularly in SNF and IRF utilization, already reduced by ~$500/episode and ~$883/surgical episode respectively by MY4-6 — is sufficient to cover the mandatory 3% discount on surgical and 2% discount on medical episodes before committing to later cohorts.
CESLG Forced Bundling Exposes Unprofitable Episode Mix
Beginning in MY4, CMS eliminated the ability of participants to cherry-pick individual clinical episodes and instead required selection of entire Clinical Episode Service Line Groups (CESLGs), each containing multiple related MS-DRG or HCPCS-triggered episodes. A participant selecting the Cardiac Care CESLG, for example, becomes accountable for AMI, Cardiac Arrhythmia, and Congestive Heart Failure simultaneously — regardless of whether all three offer comparable margin opportunity. The 7th Annual Evaluation documented that this mandatory bundling 'forced organizations to take on financial risk for clinical episodes for which they could not achieve financial gains,' directly corresponding to post-MY4 attrition. Organizations entering MY7 or MY8 must conduct granular episode-level contribution margin analysis across every clinical episode within each CESLG they select — not just the anchor episode — because the stop-loss limit of ±20% of volume-weighted aggregate target prices operates at the EI level, meaning losses in one episode cannot be fully offset by gains in another beyond that cap.
Post-Episode Monitoring Window Creates Hidden Liability
Beyond the 90-day episode window, BPCI Advanced imposes a 30-day Post-Episode Monitoring Period during which any Medicare FFS expenditures that exceed an empirically titrated risk threshold must be repaid to CMS by the participant. This liability is calculated and collected during the first true-up Reconciliation on a semi-annual basis — meaning organizations may not know their true post-episode exposure until months after the episode ends. Participants focused narrowly on managing the 90-day window may underinvest in care transitions and follow-up protocols that extend into the post-episode period, only to face unexpected clawback amounts at reconciliation. This risk is compounded for medical episodes involving complex, high-comorbidity patients where post-discharge instability is less predictable than in elective surgical episodes. Operational teams must implement monitoring protocols that extend through day 120 (90+30) and build actuarial reserves accordingly.
Attribution Conflict Between PGP and ACH Creates Revenue Leakage
BPCI Advanced uses a strict three-tier attribution precedence rule: (1) the PGP billing the attending physician's NPI gets the episode first; (2) the PGP billing the operating physician's NPI gets it second; (3) only if neither applies does the ACH where the anchor stay occurred get attribution. In markets where both a PGP and an ACH are participating in BPCI Advanced — a common scenario given that over 1,100 hospitals and 800 PGPs participated across the model — ACHs can systematically lose attribution of surgical episodes to PGPs, particularly orthopedic and cardiac episodes. This directly affects which entity bears financial risk and which earns any NPRA reconciliation payment. ACHs that selected surgical CESLGs without mapping their employed or affiliated physicians' NPI relationships to participating PGPs risk systematically losing episode volume to PGP participants, distorting their projected episode counts and undermining the volume assumptions underlying their target price calculations. Finance teams must audit NPI-to-entity assignments before each performance period.
ACO REACH Overlap Silently Erodes Surgical Episode Volume
Under BPCI Advanced overlap rules, beneficiaries aligned to an ACO REACH entity cannot trigger BPCI Advanced Clinical Episodes — they are excluded from reconciliation entirely. Participants also reported during MY7 site visits that the loss of episodes for ACO REACH-aligned patients, combined with the migration of healthier surgical patients to ambulatory surgery centers (which cannot serve as EIs under BPCI Advanced, as only ACHs and PGPs qualify), has created a systematic adverse selection effect: the remaining episode pool for surgical CESLGs skews toward higher-acuity, higher-cost patients, compressing margins against a target price derived from a broader historical baseline. This is particularly acute for orthopedic CESLGs where outpatient MJRLE and MJRUE procedures can now be triggered in HOPDs, but freestanding ASC volume escapes the model entirely. Participants should quantify what percentage of their prospective surgical episode volume is at risk of ACO REACH exclusion and model the case-mix acuity shift on their expected episode spending relative to preliminary target prices.
IRF Ownership Conflicts Undermine PAC Cost Reduction Strategy
The primary mechanism through which BPCI Advanced generated its ~$800M in net Medicare savings was reduction in PAC utilization — specifically SNF discharges reduced by ~5 percentage points for surgical episodes (MY4-6) and SNF length of stay shortened by 2.66 to 3.50 fewer days. However, the 7th Annual Evaluation identified that health systems owning Inpatient Rehabilitation Facilities faced a direct structural conflict: maintaining IRF daily census to meet the 60% compliance threshold (requiring that at least 60% of IRF patients have one of 13 qualifying conditions) creates financial pressure to increase IRF referrals that directly opposes BPCI Advanced's cost-reduction incentives. One evaluated participant reported being 'the largest provider of IRF beds in the state,' explicitly limiting its ability to reduce PAC expenditures. Governance and contracting teams must assess whether their organization's IRF revenue dependency structurally prevents achieving the PAC utilization reductions needed to stay below target prices, and whether IRF ownership should factor into CESLG selection decisions — particularly for neurological care (stroke patients have high IRF utilization) and complex orthopedic procedures.
Quality Score Adjustment Offers Limited Financial Buffer Against Repayment
BPCI Advanced's Composite Quality Score (CQS) adjusts reconciliation payments by up to 10%, which sounds material but in practice provides a narrow financial cushion. The 7th Annual Evaluation quantified the impact directly: for a participant owing CMS $500 per episode, scoring at the 95th percentile on quality reduced repayments by only $41 per episode, while scoring at the 5th percentile reduced repayments by only $13 per episode — a difference of $28 per episode between best and worst quality performers. With the stop-loss limit capping total exposure at 20% of the volume-weighted aggregate target price, participants that are already above target price cannot rely on quality performance to materially reduce their financial liability. Separately, participants must choose between the Administrative Quality Measures Set (7 exclusively claims-based measures, available since MY1) and the Alternate Quality Measures Set (up to 5 claims-based and registry-based measures, available since MY4) — and the registry-based measures in the alternate set require active data submission infrastructure. Organizations that select the Alternate Quality Measures Set without establishing reliable registry reporting workflows risk data submission failures that could suppress their CQS and reduce their reconciliation payment by up to 10% relative to potential, a meaningful sum when episode volumes are high.