MPFS 2026: the headline narrative on ortho/spine reimbursement is wrong.

The 2026 Medicare Physician Fee Schedule final rule, released October 31, 2025 and effective January 1, 2026, raised the non-QP conversion factor from $32.3465 to $33.4009 — an increase of 3.26%, the first positive conversion factor update since 2020. The trade press, the major physician advocacy organizations, and the healthcare investment banking community have received the change as a relief from six consecutive years of physician payment cuts.¹

For office-based primary care, the framing is approximately correct. For facility-based surgical specialties — including every orthopedic and spine procedure performed in a hospital, ASC, or HOPD — the framing is wrong. CMS simultaneously finalized two structural changes that more than offset the conversion factor increase for facility-based codes:

First, an efficiency adjustment of −2.5% applied to the work RVUs of nearly every non-time-based procedure, with reapplication every three years in perpetuity. The adjustment cuts work RVU values across essentially the entire surgical CPT inventory. CMS justified the adjustment by asserting that procedures become more efficient over time and that current work RVUs do not capture this efficiency. The American College of Surgeons published empirical work from the National Surgical Quality Improvement Program showing that operative times for 90% of CPT codes were longer or unchanged from 2019 to 2023, directly refuting the assumption underlying the adjustment.² CMS finalized the policy anyway.

Second, a reallocation of the indirect practice expense RVU. Beginning January 1, 2026, the portion of facility PE RVUs allocated based on work RVUs is reduced to half the allocation applied to non-facility services. CMS’s rationale: physicians performing services in facility settings no longer bear the indirect costs that the historical PE allocation was designed to cover, because those costs have shifted to facilities under employment and integration arrangements. AAOS analysis of CMS’s own Table 92 estimates that the change produces a −9% cut to facility-furnished services, partially offset by a +5% increase to office-furnished services, and that the net impact on orthopedic surgery total allowed charges is −3%. 66% of orthopedic surgeons will see RVU reductions of at least −1%, and 30% will see cuts of −5% or more.³

The net effect on the highest-volume ortho/spine codes is captured in Table 1. The conversion factor goes up. The work RVU goes down. The facility PE RVU goes down. The realized payment drops 7–8% relative to 2025 for the major joint and spine procedures.

The decline is approximately uniform across the major ortho/spine codes. A deal model that priced 2% Medicare growth for 2026 — the modal assumption in current LBO models — is overstating the Medicare-correlated revenue base by roughly 10% for the highest-volume codes in year one alone.

The trade press framing of 2026 as “physician payment relief” reflects the conversion factor in isolation. The deal-relevant framing requires the conversion factor, the efficiency adjustment, and the PE reallocation considered together. The first frame is the headline. The second frame is the truth.

The 2026 changes are not an anomaly. They are the continuation of a structural decline in Medicare reimbursement for facility-based ortho/spine procedures that has run consistently across every administration and every regulatory regime since 2000.

Palmer et al. (2025), examining Medicare PFS data from 2000 through 2024, found that unadjusted physician reimbursement for primary TKA declined 22.3%, primary THA declined 17.7%, revision TKA declined 9.0%, and revision THA declined 13.0%. After adjustment for inflation to 2024 dollars, the same procedures declined 56.9% (TKA), 54.3% (THA), 50.8% (spine fusion), 49.4% (total ankle arthroplasty), 43.6% (femoral ORIF), and 38.4% (total shoulder arthroplasty).⁴ The mean inflation-adjusted decline across all nonarthroplasty orthopedic procedures was 44.0%.

The pattern is not specific to arthroplasty. Haglin et al. (2024) examined the five most common spinal procedures from 2000 to 2020 and found inflation-adjusted reimbursement declines of −11.05% for ACDF, −28.38% for posterior cervical fusion, −7.85% for TLIF, −28.17% for lumbar microdiscectomy, and −31.88% for lumbar laminectomy. Spinal instrumentation codes declined 33.43% on average.⁵

The trajectory has three features that matter for deal underwriting.

First, the decline is consistent across administrations. Republican and Democratic CMS leadership have produced similar trajectories. The 2018–2020 IPO list elimination effort under the first Trump administration produced negative reimbursement effects through site-of-service migration; the 2022 Biden reversal slowed but did not reverse the decline; the 2026 efficiency adjustment under the second Trump administration accelerates it. Modeling the decline as cyclical or administration-dependent is empirically wrong.

Second, the decline is consistent across procedure categories. Hip, knee, shoulder, ankle, spine, and trauma codes all show similar inflation-adjusted decline trajectories over 25 years. The decline is not specific to procedures that have become more efficient, nor to procedures that have lost case complexity, nor to procedures that have migrated outpatient. It is structural to how CMS values facility-based physician work.

Third, the decline is accelerating rather than plateauing. The efficiency adjustment establishes a recurring −2.5% reduction every three years. The PE reallocation continues to compound. The conversion factor is structurally tied to the One Big Beautiful Bill Act’s 2.5% temporary increase, which expires December 30, 2026 — meaning the 2027 conversion factor begins from a lower base, before any further policy changes.

The implication for LBO modeling is direct. A deal that assumes 2% annual Medicare growth on facility-based ortho codes — the modal assumption we have observed in PE underwriting — is overstating the Medicare trajectory by approximately 4–5% per year against actual realized declines. Across a five-year hold, that overstatement compounds to roughly 20–25% overvaluation of the Medicare-correlated revenue base. For a typical ortho/spine PE asset with 30–40% Medicare exposure, this is an 8% overvaluation of total revenue. Through standard ortho EBITDA margins of 18–22%, this translates to roughly a 30% overvaluation of EBITDA growth — which is the variable that drives multiple expansion at exit.

The deal that looks like a 10x EBITDA entry at $400M is more like an 11x entry once the Medicare trajectory is correctly modeled. The cost of getting this wrong is quantifiable, and it is the cost most current models are paying.

The 2026 OPPS/ASC final rule, released November 21, 2025, finalized the elimination of the Medicare Inpatient-Only list over three years, beginning January 1, 2026 with the removal of 285 procedures — the substantial majority of them musculoskeletal — from the list.⁶ This is the second attempt at IPO list elimination, after the 2018–2020 effort that removed TKA (2018) and THA (2020), was reversed by the Biden administration in 2022 over safety concerns. The 2026 rule resumes the phase-out with full elimination scheduled for January 1, 2029.

The trade press, the ASC industry, and the major PE-backed ortho platforms have positioned the change as a substantial positive for facility-fee capture in PE-owned ASCs. The framing is correct in direction. The framing is wrong in magnitude, in timing, and in the structure of realized economics.

Three specific divergences from common deal-model assumptions:

The migration capture is not 100%. Procedures removed from the IPO list become eligible for outpatient settings — they do not automatically migrate there. Migration depends on the asset’s ASC capacity utilization, anesthesia coverage, surgeon preference, patient acuity profile, and payer authorization patterns. We have not seen a PE-backed ortho/spine platform achieve full migration of newly-eligible procedures within a three-year hold; realistic capture is 30–50% for the most-eligible high-volume procedures (knee/hip arthroscopy, simpler arthroplasty cases), lower for higher-acuity procedures (revision arthroplasty, multi-level fusion). LBO models that assume 80–100% migration capture are systematically overstating the synergy.

The HOPD-to-ASC payment delta is materially below the gross facility fee. The 2025 OPPS payment for TJA in the HOPD setting was $12,866.82; the 2025 ASC payment was approximately $9,449.04 (THA) and $9,255.60 (TKA). The PE-owned ASC captures the facility fee that previously went to the hospital — but the per-procedure facility revenue is approximately $3,400 lower in the ASC setting than in the HOPD setting. The net economic capture for the platform depends on whether the platform was previously getting any economics from the hospital setting (typically not, except via joint-venture arrangements) and what the platform’s ASC capacity profile is (typically constrained on case-day throughput). The modal deal model assumes the platform captures the gross HOPD facility fee on migrated cases. The realistic capture is the gross ASC facility fee minus the ASC’s direct operating costs minus any commercial pass-through reductions.

Commercial pass-through lags Medicare by 12–24 months and is incomplete. Commercial payers are not required to follow Medicare’s IPO list. They have their own coverage policies, prior authorization protocols, and site-of-service restrictions. As Medicare opens site-of-service flexibility, commercial payers will follow — but slowly, inconsistently across markets, and with negotiation friction. The commercial pass-through that delivers the full ASC migration synergy materializes over 18–36 months on a payer-by-payer basis, not on day one. LBO models that underwrite full commercial pass-through in year one are systematically wrong on timing.

The combined effect of these three divergences is that the realistic ASC migration synergy is approximately one-third to one-half of what most current deal models assume, and the realization timeline is 2–3 years longer. For a deal that’s underwriting site-of-service migration as the largest single synergy lever, the corrected math often shifts the IRR from headline 14–15% to realistic 8–10%.

The regulatory and structural changes have different impacts on different ortho/spine asset profiles. We model three representative asset types that approximate the structural form of most current PE-backed musculoskeletal platforms.

Profile A: General orthopedic platform. 40–50 locations, 150–200 providers, mixed sub-specialty (sports medicine, hand, foot/ankle, general ortho). High volume on arthroscopy (29881, 29827), moderate volume on primary TJA (27447, 27130), low spine exposure. Payer mix 35% Medicare, 50% commercial, 15% other.

Under correct 2026 assumptions: PFS revenue declines 6–7% on the surgical book (efficiency + PE reallocation), partially offset by office-based procedure increases (+5%). Net professional fee impact: −3% to −4% on the surgical revenue base in 2026.

ASC migration upside is real but limited — arthroscopy is already largely ASC-based, so the IPO list elimination has marginal incremental benefit for the surgical book. Site-of-service synergy: real but modest, 1–2% of revenue over a three-year hold.

Profile B: Spine-heavy practice. 15–25 locations, 50–75 providers, significant fusion volume (22551, 22612, 22633), substantial inpatient case mix, ASC capacity for outpatient spine. Payer mix 30% Medicare, 55% commercial, 15% other.

Under correct 2026 assumptions: PFS revenue declines 7–8% on the surgical book. Spine codes are heavily facility-based, so the PE reallocation hits hard. Commercial pass-through is delayed.

ASC migration upside is substantial in theory — single-level spine procedures are among the 285 removed from the IPO list, and the platform’s ASC capacity is the binding constraint on capture. But realistic three-year migration is 30–40% of newly-eligible procedures, with commercial pass-through lagging 18–24 months. Site-of-service synergy: 3–5% of revenue over a three-year hold, materializing in years 2–3 rather than year 1.

Profile C: Mixed musculoskeletal platform. 30–40 locations, 100–150 providers, balanced sub-specialty mix, owned ASC infrastructure with 2–4 surgery centers, established commercial contracting. Payer mix 35% Medicare, 50% commercial, 15% other. This is the structural form of the typical current PE rollup.

Under correct 2026 assumptions: PFS revenue declines 6–7% on the surgical book in 2026. ASC migration produces meaningful but bounded upside — the platform’s existing ASC capacity captures some incremental volume, but capacity expansion (additional ORs, anesthesia coverage, post-op infrastructure) is required to fully realize migration upside. Capacity expansion is itself a capex item that current models often understate.

Net professional fee impact in 2026: −3% to −5% on the surgical revenue base. ASC migration upside: 2–4% of revenue over three years. Combined impact relative to a model assuming 2% Medicare growth and full ASC migration: approximately −8% on revenue trajectory in year one, gradually closing to −3% to −5% by year three as ASC migration partially compensates.

For a Profile C asset at $400M purchase price, this is approximately $30–50M of equity value misallocation under the modal LBO assumptions, before considering any other variance findings.

Four specific synergy and trajectory assumptions appear in essentially every ortho/spine PE LBO model we have reviewed. Each diverges from the computable answer in ways that materially affect deal economics.

Assumption: Medicare grows at low single digits, commercial passes through with a modest lag. Both halves of this assumption are wrong for facility-based ortho/spine. Medicare is structurally declining for facility-based ortho codes — the 25-year trajectory is unambiguous, the 2026 changes accelerate it, and there is no policy mechanism currently in place to reverse the direction. Commercial pass-through is incomplete and lagged in both directions — commercial payers do not fully pass through Medicare increases, and they also do not fully pass through Medicare decreases in the near term. The realistic modeling assumption is that Medicare declines at 1–2% annually for facility-based ortho, commercial reimbursement is approximately flat in nominal terms (declining in real terms), and the realized blended rate trajectory is −0.5% to −1.5% per year. Current models assuming +1% to +2% blended growth are overstating the revenue trajectory by 200–300 basis points annually.

Assumption: Contract migration to platform contracts produces immediate uplift in line with the platform’s average rates. The math of contract migration is computable claim-by-claim. The diligence team that models it on a portfolio-average basis (target’s revenue × ratio of platform rates to target rates) systematically overestimates the uplift, because the actual uplift depends on the rate differential by procedure by payer, weighted by the target’s specific volume mix. The most commercially valuable codes in the target’s mix are often the codes where the rate differential is smallest, and the codes with the largest rate differential are often the codes with the lowest volume. Realistic computed uplift is typically 40–60% of what the portfolio-average method produces.

Assumption: Site-of-service migration to owned ASCs captures full facility fee delta on day one. The composite of issues developed in section 3 above. Realistic migration capture over a three-year hold is 30–50% of eligible procedures, the per-procedure facility delta is approximately $3,400 lower than the HOPD-to-ASC gross differential, and commercial pass-through lags 12–24 months. Combined effect: realistic synergy is one-third to one-half of common model assumptions.

Assumption: Provider productivity normalization to platform standards produces the modeled synergy. Provider-level analysis requires claim-by-claim computation of each provider’s actual coding patterns against the platform’s coding standards. The gap between average platform productivity and the acquired asset’s productivity is almost always smaller than the diligence team’s initial estimate, because the diligence estimate is based on average wRVU per provider rather than on the procedural mix that drives the average. Realistic computed synergy is typically 30–50% of what the platform-average method produces, and the realization timeline is 12–18 months rather than the 6-month timeline that most operating partners assume.

Each of these divergences alone is material to deal economics. Combined, they typically produce a 10–15% overstatement of EBITDA growth assumptions in current LBO models, which translates to a 15–25% overvaluation of equity value at exit.

The diligence methodology that produces these divergences is not specific to the firm performing the analysis. It is structural to a methodology that relies on portfolio-average projections rather than claim-level computation. The 2026 regulatory changes amplify the magnitude of the error because the changes shift the relative economics of specific procedures, payers, and sites of service in ways that average-based modeling cannot capture.

Three implications follow for ortho/spine deals currently in market or under early-stage evaluation.

The comparable transactions that anchor current valuations are stale. Comparable musculoskeletal PE transactions from 2024 and the first three quarters of 2025 were priced before the efficiency adjustment, before the PE reallocation, and before the IPO list elimination was finalized. The multiples reflect a regulatory environment that no longer exists. Underwriting against these comps without adjustment for the 2026 changes systematically overprices the asset.

The Medicare trajectory in the LBO model should be recomputed under correct 2026 assumptions. The headline conversion factor is not the relevant variable. The relevant variable is the realized payment trajectory by procedure, after the efficiency adjustment, after the PE reallocation, and accounting for the recurring −2.5% adjustment scheduled every three years. The realistic trajectory for facility-based ortho codes is −1% to −2% per year nominal, not +1% to +2%.

The site-of-service migration synergy should be re-priced under realistic capture and timing assumptions. Migration is real but bounded; ASC payment is materially below HOPD payment; commercial pass-through lags Medicare. Models assuming full migration in year one are overstating the synergy by a factor of 2–3x.

For deals in active diligence, the corrections to the LBO model are computable claim-by-claim against the target’s actual historical data. The 2026 changes apply to the target’s exact procedural mix, payer composition, and site-of-service distribution. The work of correctly pricing the deal under current regulatory conditions is mechanical once the analytical infrastructure exists.

The deal that’s getting underwritten this quarter against stale comps, headline regulatory analysis, and average-based synergy projections is mispricing the asset by 15–25% of equity value. Across the wave of first-cycle PE exits coming over the next 18–24 months in musculoskeletal services, the aggregate mispricing across the sector is substantial.

This note documents the structural divergences. The corrections are available to any buyer willing to perform claim-level analysis under correct 2026 assumptions.