CMS issued the Calendar Year 2027 Medicare Physician Fee Schedule proposed rule on July 14, 2026, published in the Federal Register on July 16. The comment period closes September 14, 2026 — 23 days from today. This is the single most consequential open regulatory item for physician practices right now, and unlike the OPPS rule that dominated July, this one hits every practice that bills Part B.
| Track | CY 2026 | Proposed CY 2027 | Change |
|---|---|---|---|
| Qualifying APM participant | $33.5675 | $33.1693 | −$0.40 (−1.19%) |
| Non-qualifying APM participant | $33.4009 | $32.8409 | −$0.56 (−1.68%) |
The driver is statutory, not discretionary. Public Law 119-21 provided a one-year 2.50% conversion factor increase for CY 2026 only. That increase expires December 31, 2026, and current law requires the −2.50% unwind. Everything else in the rule — budget-neutrality adjustments, the efficiency adjustment, practice expense reform — moves the number back up partway, netting to the figures above.
The provision generating the loudest response is a proposed 50% payment reduction on E/M visits billed with modifier 25 on the same day as a procedure carrying a 0-, 10-, or 90-day global period. Under the proposal, only the highest-valued affected service that day — whether the E/M or the procedure — is paid at 100% of the fee schedule. Every additional affected E/M or global procedure furnished that day is paid at 50%.
Critically, the reduction applies even when the E/M is properly supported as significant and separately identifiable. This is not an anti-fraud measure aimed at improper modifier 25 use. CMS’s stated rationale is that same-day E/M work overlaps with the pre-service and post-service work already bundled into any global-period procedure, creating what CMS characterizes as duplicate payment.
The specialties carrying the heaviest modifier 25 volume — dermatology, podiatry, ophthalmology, orthopedics, OB/GYN, family medicine, and urgent care — absorb the bulk of the impact. For a practice where a same-day injection, lesion removal, or fracture care commonly follows an evaluation, this is a structural revenue change, not a coding-hygiene issue.
The CY 2027 PFS comment period closes September 14, 2026, with the final rule expected in November. Modifier 25 is the provision most likely to be modified if the comment record is strong. Pull your trailing-12-month volume of E/M visits billed with modifier 25 alongside a global-period procedure, multiply the allowed amount by 50%, and put that number in a comment letter.
CMS proposes a 2.5% reduction to work RVUs and intra-service time for most non-time-based services, on the theory that procedures become more efficient over time. Time-based codes are exempt: E/M visits, care management, behavioral health, and telehealth-list services are carved out. The practical effect is a systematic transfer of RVU value away from proceduralists and toward cognitive and longitudinal care.
CMS proposes to phase out the Indirect Practice Cost Index over a two-year transition, on grounds that it anchors practice expense values to the AMA’s 2007 Physician Practice Information Survey — data now nearly two decades old. Indirect PE allocation would be re-based to align with work RVUs and clinical labor inputs. To dampen the volatility, CMS proposes a PE stabilization adjustment capping annual PE RVU movement at ±5% for most existing services.
McDermott+ has flagged an important caveat: because the ±5% cap compresses year-one movement, the proposed 2027 specialty impact tables understate where several specialties actually land once the transition completes. Practices modeling only the published 2027 percentages are modeling the floor of the change, not its full extent.
CMS proposes converting HCPCS G2211, the visit-complexity add-on for longitudinal care, from a flat-rate add-on code into a two-digit modifier appended to the base E/M code, increasing payment on the associated E/M by 16%. Because it is a percentage rather than a flat dollar amount, the benefit scales with visit level. Clinicians participating in Medicare Shared Savings Program ACOs and the Long-Term Enhanced ACO Design (LEAD) model would receive an enhanced 32% adjustment — double the standard rate.
The Consolidated Appropriations Act, 2026 continued geographic and originating-site waivers through CY 2027. CMS proposes to extend the corresponding flexibilities — geographic restrictions, originating sites, audio-only, and the tele-mental-health in-person requirement — through December 31, 2027, and to pay RHCs and FQHCs for non-behavioral-health telecommunication visits. One significant tightening: beginning January 1, 2027, CMS proposes that clinical staff furnishing RPM or RTM services must be direct employees of the billing practitioner or practice — contracted third-party staffing would no longer support the codes.
Layered with a 7% specialty-level RVU decline, orthopedic surgery faces an estimated 8% to 9% reduction in total PFS allowed charges — the fourth consecutive year of net physician payment decline. The conversion factor is the headline; modifier 25 and the efficiency adjustment are where the money actually moves.
2026 is the year prior authorization reform stopped being a press release and became a compliance obligation. Three CMS-0057-F provisions are now operative, and practices that have not re-tooled their PA workflow around them are leaving both time and money on the table.
Decision clocks. Impacted payers — Medicare Advantage organizations, state Medicaid and CHIP programs, Medicaid managed care plans, and Qualified Health Plan issuers on the federal exchanges — must issue standard prior authorization decisions within seven calendar days and expedited/urgent decisions within 72 hours. This is enforceable now, not aspirational.
Specific denial reasons. Payers must provide a specific reason for every denied prior authorization request for a medical item or service. “Does not meet medical necessity” without further specificity no longer satisfies the requirement — and a non-specific denial is itself an appeal argument.
Public metrics. The first mandatory public prior authorization reporting was due March 31, 2026, covering calendar-year 2025 data: approval rates, denial rates, decision turnaround times, and appeals outcomes, posted on each payer’s public website. This is the first year practices can benchmark a payer’s stated behavior against their own experience — and it is the single most underused negotiating asset in contracting right now.
Still ahead. The API build-out is the long pole. Impacted payers have until January 1, 2027 to meet the FHIR-based Patient Access, Provider Access, Payer-to-Payer, and Prior Authorization API requirements. CMS has also issued the 2026 Interoperability Standards and Prior Authorization for Drugs proposed rule (CMS-0062-P), which would extend electronic PA, shortened decision timeframes, and transparency requirements to drug prior authorization — the category CMS-0057-F explicitly excluded.
Gold carding is a look-forward exemption tied to a provider’s approval rate on specific procedures, and virtually every program permits revocation mid-period if the approval rate falls below threshold. Gold card status is a performance state to be maintained, not a credential to be earned once. Assign a named owner to monitor each program you participate in.
Reform has not yet reversed the underlying trend. In MBC’s 2026 denial management analysis across 240 specialty practices, prior authorization denials now represent 34% of all first-pass claim denials — up from 22% in 2023. Imaging is the sharpest edge: prior authorization accounts for roughly 35% of radiology denials, and imaging denial rates across studied Medicare Advantage procedures run about 4.94%, with materially higher rates for spine MRI and PET.
Standard manual PA submissions still take 3 to 7 days to turn around. Automated submissions run 1 to 3 days, and expedited urgent cases must resolve within 72 hours under the 2026 rules. The gap between manual and automated submission is now, on average, a four-day difference in cash conversion per authorized encounter.
Pull the March 31 public PA reports for your three highest-volume payers. Compare their published denial rate and turnaround time against your own observed numbers for the same period. Where your experience is materially worse than their published figure, you have a documented, payer-authored benchmark to bring to your next contract or joint operating committee discussion.
The 2026 mid-year AI data tells a consistent story across every survey: adoption is nearly universal, deployment is partial, and realized return is rare. The gap between those three numbers is where this year’s competitive advantage lives.
Four out of five organizations are in motion. Roughly one in seven has anything to show for it. The failure mode is not model quality — it is integration. AI deployed as a parallel tool alongside an unchanged workflow produces a demo, not a return.
The most commonly prioritized AI functions are improving denial management and appeals (57%) and documentation and coding accuracy (56%). But intent and deployment diverge sharply: while 41% of providers now run denial rates of 10% or higher, only 14% use AI for denial reduction. That is the largest single mismatch between problem severity and tool application anywhere in the revenue cycle.
Ambient documentation has matured past the pilot phase, and the 2026 evidence base is now substantial enough to underwrite a business case. Clinicians using ambient AI tools spend 8.5% less total time in the EHR than matched controls, with a greater than 15% decrease in note-composition time specifically. Family and internal medicine practices report returning 60 to 90 minutes per day per clinician, largely by eliminating evening documentation.
A three-group pragmatic randomized clinical trial published in NEJM AI found reduced documentation time, improved clinician satisfaction, and meaningful reduction in reported burnout symptoms across all 14 specialties tested. One study documented mean burnout scores falling from 2.7 to 2.0 on a five-point scale after adoption.
On the revenue side: San Diego–based Sharp HealthCare reported a 3.5% to 6% increase in wRVUs per encounter. Published results from The Permanente Medical Group, Sutter Health, Mass General Brigham, Emory Healthcare, and St. Luke’s Health System show documentation-burden and burnout reductions commonly in the 20% to 40% range, with early single-site evidence of roughly a 5.8% gain in weekly RVUs.
For independent practices, the ROI arithmetic is different from a health system’s and generally more favorable: the return shows up in clinician retention and in the capacity to add one or two visits per day without adding burnout — not in an FTE reduction line.
The organizations converting AI to ROI are not the ones with the best model. They are the ones that redesigned the workflow around the model — retiring the manual step the AI replaced rather than running both. If your denial-management AI produces a worklist that a human still fully re-reviews before action, you have added a step, not removed one.
The global AI in revenue cycle management market was $20.63 billion in 2024 and is projected to reach $70.12 billion by 2030, a 24.16% CAGR.
Two code-set events are converging on practices this cycle, and they require different preparation tracks. ICD-10-CM hits in 40 days. CPT 2027 hits in 132 days but requires far more workflow redesign.
CMS released the FY 2027 ICD-10-CM diagnosis code files with 190 new codes, 30 deletions, and 4 revisions, applicable to encounters on and after October 1, 2026 and running through September 30, 2027. The additions cluster in three themes: greater anatomical specificity for cancer and infection sites; dedicated codes for genetic and rare-disease diagnoses that previously required unspecified-code workarounds; and expanded surveillance codes for emerging public health concerns, including illicit substance adulterants and environmental exposures.
| Code | Description | Who Should Care |
|---|---|---|
| Z86.17 | Personal history of Clostridioides difficile infection | Primary care, GI, infectious disease, long-term care |
| E89.830 / E89.838 | Postprocedural hypoglycemia following bariatric and other surgery | Endocrinology, bariatric surgery, primary care |
| Z68.18 | Body mass index 18.4 or less, adult | Primary care, nutrition, oncology, behavioral health |
| Z68.19 | Body mass index 18.5–19.9, adult | Primary care, nutrition, geriatrics |
The specialties with the heaviest review burden are maternal-fetal medicine, oncology, cardiology, and toxicology — coders and clinical documentation staff in those areas should be working descriptor and inclusion-term reviews now, not in late September.
The AMA has announced the first major restructuring of maternity care CPT coding in over 30 years, effective January 1, 2027. The global obstetric billing model is being unbundled into a granular structure that separately reports antepartum care, labor management, delivery, and postpartum services. Antepartum care moves to standard E/M codes, and new labor management codes 59080–59083 report inpatient labor management services.
For OB/GYN practices this is not a code-swap. It is a rebuild of the charge capture, contracting, and patient financial estimate workflow — global maternity packages are frequently contracted as a single case rate with commercial payers, and those contracts will need renegotiation. Practices should be opening that conversation with payers in Q4 2026, not Q1 2027.
58 new Category III codes, plus revised language for three existing codes. Most of the new Category III codes cover hypothermic oxygenated perfusion for liver allografts, mirroring the existing normothermic perfusion codes. The adaptive behavior (ABA) code set receives its largest revision since 2019 — six new codes, revised descriptors for all eight Category I codes, revised guidelines, and deletion of both Category III codes, all effective January 1, 2027. The CPT 2027 Professional Edition with final descriptors is expected later in 2026.
Run a payer-by-payer confirmation that your clearinghouse and top five payers will accept FY 2027 codes on October 1 date-of-service claims. Split-cycle testing failures at the clearinghouse level are the most common October 1 revenue disruption and the easiest to prevent. Practices still coding from the FY 2026 set into October encounters typically see the denial pattern surface two to three weeks later — when the cash impact is already baked in.
The denial environment deteriorated through the first half of 2026, and the aggregate numbers now make the automation business case arithmetic rather than aspirational.
| Capability | Manual Review | With Automated Scrubbing |
|---|---|---|
| First-pass acceptance rate | 78–82% | 92–95% |
| Typical denial rate | 11–15% | 5–7% within 60–90 days |
| Front-end denial reduction from accurate intake alone | — | up to 30% |
A practice at 82% first-pass acceptance submitting 2,000 claims a month is reworking roughly 360 claims monthly. At $57 per rework, that is $20,520 per month — $246,240 annually in pure rework cost, before counting the delayed cash and the write-offs on claims that time out. Denied amounts are also rising, not just denial counts: average denied amount is up 14% in hospital outpatient and 12% in inpatient settings versus 2025.
| KPI | Target | Watch Zone | Action Threshold |
|---|---|---|---|
| Days in A/R | ≤ 35 days | 35–45 days | > 45 days |
| Clean claim rate | ≥ 95% (top performers 99%) | 90–95% | < 90% |
| First-pass claim denial rate | < 5% | 5–10% | > 10% |
| A/R over 90 days | < 10% of total receivables | 10–15% | > 15% |
| Net collection rate | ≥ 96% | 93–96% | < 93% |
| Prior auth turnaround | ≤ 2 days (automated) | 3–5 days | > 5 days |
| Cost to collect | ≤ 3% of net patient revenue | 3–5% | > 5% |
HFMA’s standing guidance places days in A/R at 30–40 days, with high performers under 35, denial rates below 5%, and A/R over 90 days below 10% of total receivables.
Patient responsibility has grown from 1–2% of provider revenue to 10% or more, driven by a 65% increase in HDHP enrollment and a 69% rise in out-of-pocket spend over the past decade. Patient payments now account for roughly 35% of practice revenue, up from about 5% in 2000. Average collection rates for commercially insured patients fell to 34.4% in 2024.
Meanwhile 62% of patients prefer online payment options and 93% say the billing experience affects whether they will return. The practices recovering 80%+ of patient responsibility share a common pattern: upfront estimates, point-of-service collection, digital payment channels, and structured payment plans — deployed together, not selectively.
Front-end automation carries the highest ROI density in the entire revenue cycle. Sixty to seventy percent of denials originate before the claim is ever coded. If you can only fund one automation project this quarter, fund real-time eligibility and benefits verification at scheduling — not denial appeals.
Two developments this quarter point in opposite directions and are best read together: the vendor evaluation data says platform consolidation is working, and the capital markets data says the business model underneath it is being repriced.
KLAS Research released its 2026 Revenue Cycle Management Suites report in July, naming Waystar the top-rated suite vendor for organizations pursuing revenue cycle consolidation. Waystar earned an overall “A” grade — the highest in the report. KLAS attributed the rating to four specific drivers, and each is a useful evaluation criterion regardless of which vendor a practice is considering:
That fourth point is the tell. KLAS is now scoring vendors on whether customers actually use the AI capability, not whether it appears on the feature matrix. Practices running vendor evaluations should ask for adoption percentages by module, not capability lists.
PitchBook’s Q2 2026 report — bluntly titled “AI Kills the RCM Star” — argues private equity should be cautious on RCM deals because agentic AI is compressing take rates. The scale of exposure is significant: $42.9 billion in PE deals in RCM since 2017, $22.5 billion in exits, and a recently announced $12 billion acquisition of Ensemble Health Partners.
The sponsor thesis has been to layer agentic AI onto a labor-heavy back office, cut offshore staffing, and expand gross margins. PitchBook’s counter-argument is that if AI genuinely reduces the labor required to work a claim, the percentage-of-collections pricing model that RCM outsourcers depend on compresses toward the cost of the software — and the EHR incumbents, notably Epic, can bolt RCM onto platforms they already own and treat the revenue as pure incremental profit. Market pricing reflects the concern: Waystar, a high-quality near-AI-native vendor still growing organic revenue around 11%, is down roughly 43% year-to-date.
If percentage-of-collections pricing is under structural pressure, leverage in outsourced RCM contracts is shifting toward the buyer. Shorten your terms — a five-year percentage-of-collections agreement signed in 2026 locks you into today’s pricing model through a repricing cycle. And price the AI separately: if your RCM partner is deploying automation that reduces their labor cost, that saving should be visible in your rate, not absorbed into their margin. Ask for it in writing at renewal.
The CAQH Index frames the addressable opportunity: routine administrative transactions cost the U.S. health system roughly $90 billion annually, with about $20 billion in identified savings available. Providers shoulder 97% of the roughly $83 billion spent on staff time conducting those transactions.
Three compliance regimes crossed from “announced” to “enforced” in 2026. Practices with hospital affiliations, network participation, or exposure in the six WISeR states should be auditing against all three now.
The CY 2026 OPPS/ASC final rule tightened hospital price transparency requirements effective January 1, 2026, with CMS delaying enforcement to April 1, 2026 to allow a transition window. That window has closed. The “estimated allowed amount” data element is replaced by more granular metrics: hospitals must disclose the median allowed amount and the 10th percentile allowed amount, rely on standardized data sources rather than internal estimates, and satisfy stronger attestation and NPI encoding requirements in the machine-readable file.
For physician practices, the direct obligation is limited — but the derived opportunity is not. Competitor and affiliated-hospital machine-readable files now contain actual median and 10th-percentile allowed amounts by payer. That is real contract intelligence, published under federal mandate, and it is the strongest external benchmark most independent practices will ever get for free ahead of a rate negotiation.
Under the CMS Health Tech Ecosystem initiative, 21 networks and companies — including athenahealth, Carequality, Innovaccer, and Surescripts — voluntarily pledged to meet CMS interoperability framework criteria and become CMS Aligned Networks. By July 4, 2026, participating networks were required to expose data via modern FHIR APIs aligned with the US Core Implementation Guide and USCDI v3, with terminology compliance — LOINC for labs, RxNorm for medications, SNOMED for conditions — and a full FHIR capability statement. As of April 2026, 37 providers had signed on as Health Systems and Provider Early Adopters.
More than 500 million records had been exchanged through TEFCA as of February 2026. CMS’s stated position is that TEFCA is “the floor, not the finish line” — CMS Aligned Networks sit on top of it with tighter criteria around patient access, provider access and delegation, and data availability.
Why a practice should care: ask your EHR or HIE vendor in writing whether they are a CMS Aligned Network participant and whether they met the July 4 criteria. That answer determines whether you will be able to pull payer-side clinical and coverage data automatically when the January 1, 2027 payer API deadline arrives — or whether you will still be faxing.
The Wasteful and Inappropriate Service Reduction (WISeR) Model launched January 1, 2026 as a CMS Innovation Center pilot in six states: New Jersey, Ohio, Oklahoma, Texas, Arizona, and Washington. WISeR adds prior authorization requirements to 17 services in Traditional Medicare fee-for-service — a program that historically had almost none. CMS contracts with third-party technology vendors that apply artificial intelligence and clinical review to assess requests against existing Medicare coverage policies. Covered services include skin and tissue substitutes, electrical nerve stimulators, and knee arthroscopy for knee osteoarthritis; the complete list is in the CMS Provider and Supplier Operational Guide.
For practices in New Jersey, Ohio, Oklahoma, Texas, Arizona, or Washington, if your prior authorization workflow was scoped to commercial and Medicare Advantage only, it now has a live gap. Extend PA intake, tracking, and follow-up to Traditional Medicare for the 17 WISeR service categories.
HHS OIG confirmed that several major Medicare Advantage plans were using AI to systematically downcode claims without adequate clinical review. 26 or more states introduced bills in 2026 targeting AI-driven downcoding, most focused on requiring licensed physician review before any automated code reduction is applied. Track your state’s bill status — where these pass, they create a documented appeal argument for retroactive recovery.
| Date | Deadline | Who |
|---|---|---|
| Sept 14, 2026 | CY 2027 PFS proposed rule comment deadline | All Part B billers |
| Oct 1, 2026 | FY 2027 ICD-10-CM effective date | All practices |
| Nov 2026 (expected) | CY 2027 PFS final rule publication | All Part B billers |
| Jan 1, 2027 | CPT 2027 effective — maternity restructure, ABA overhaul, 58 new Cat III codes | OB/GYN, ABA providers, all practices |
| Jan 1, 2027 | Payer FHIR API requirements under CMS-0057-F | Practices consuming payer APIs |
| Jan 1, 2027 | RPM/RTM direct-employee staffing requirement (if finalized) | Remote care programs |
| Dec 31, 2027 | Telehealth geographic and originating-site waivers expire (as proposed) | Telehealth billers |
The independence data released this year is the starkest yet, and the CY 2027 PFS proposals land squarely on the practices least equipped to absorb them.
The structural driver has not changed: Medicare still pays, on average, two to four times more for many identical outpatient procedures performed in a hospital outpatient department than in a physician’s office. Every year that differential persists, the acquisition math favors the acquirer.
MGMA’s 2026 Regulatory Burden Report, drawn from leaders representing more than 230 medical group practices nationwide, quantifies what independent administrators have been describing anecdotally:
| Cost Category | Five-Year Increase |
|---|---|
| Median total operating cost per FTE physician (overhead) | +29.3% |
| Median total support staff cost | +19.3% |
| Median total general operating cost | +32.7% |
Set that against a CY 2027 conversion factor proposed at −1.68% and a fourth consecutive year of net physician payment decline. Overhead is compounding at roughly 6% annually while payment declines. That is the entire independence problem in two numbers.
Algorithmic downcoding is the pressure point where independent practices lose the most money with the least visibility, because it never appears as a denial.
Because a downcode posts as a payment rather than a rejection, it never enters the denial worklist. Most practices are not tracking it at all. That is precisely why 68% of it goes unrecovered — not because the appeals are hard, but because nobody is counting.
| Proposal | Health System Impact | Independent Practice Impact |
|---|---|---|
| Modifier 25 50% cut | Absorbed across a diversified service mix; hospital-based E/M often billed separately | Direct hit on the office-visit-plus-minor-procedure model that defines primary care, dermatology, podiatry, and orthopedics |
| 2.5% efficiency adjustment | Diluted by facility-fee revenue | Full exposure — no facility fee to offset |
| IPCI phase-out + ±5% PE cap | Large groups can model and forecast; dedicated reimbursement analysts | Few independents have the analytic capacity to model two-year PE transition impacts |
| G2211 → 16% modifier (32% for ACO participants) | Health-system ACOs capture the 32% tier | Independents outside an ACO capture 16% — a real gain, but half the available upside |
| RPM/RTM direct-employee requirement | In-house staffing already common | Contracted third-party RPM staffing — the model most small practices use — would no longer qualify |
The G2211 line deserves emphasis because it cuts the other way. An independent primary care practice that joins an MSSP ACO or the LEAD model captures a 32% E/M uplift on longitudinal-care visits instead of 16%. For a practice with high chronic-care volume, that spread is one of the few 2027 proposals that materially rewards staying independent — provided you are in an accountable care arrangement.
| Metric | Industry Target | Where Most Independents Are | Priority |
|---|---|---|---|
| Days in A/R | ≤ 35 days | 42–55 days | High |
| Clean claim rate | ≥ 95% | 82–90% | High |
| First-pass denial rate | < 5% | 10–14% | Critical |
| Front-end eligibility automation | 100% of scheduled encounters | Spot-checked or manual | Critical |
| Downcoding detection | Tracked as a distinct KPI | Not tracked | Critical |
| A/R over 90 days | < 10% | 18–25% | High |
| Cost to collect | ≤ 3% of net patient revenue | 5–8% | Medium |
| AI/automation in ≥1 RCM function | Deployed and integrated | Piloted or absent | Medium |
| G2211 capture rate on eligible visits | > 60% of eligible E/M | < 20% | High |
| Patient responsibility collected at point of service | > 50% | 15–25% | High |
1. Stand up downcoding detection this month. Run a paid-claims report comparing submitted E/M level to paid E/M level by payer for the trailing 12 months. Anything with a systematic delta is recoverable at a 68% success rate — found money requiring no new software. 2. File a CY 2027 PFS comment before September 14. Independent practices are structurally underrepresented in the comment record. 3. Audit your G2211 capture rate. Under-capture is near-universal, and the 2027 percentage modifier makes it worth more. 4. Evaluate ACO participation on the G2211 spread alone — 32% versus 16% changes the business case for chronic-care-heavy practices. 5. Automate front-end eligibility before anything else — 60–70% of denials start there.
The CY 2027 PFS proposed rule produces the widest specialty dispersion in several years. The conversion factor change is nearly uniform; the RVU and policy changes are not.
| Specialty | Direction | Key Driver This Week | Action |
|---|---|---|---|
| Primary Care / Family Medicine | Favorable on policy, exposed on modifier 25 | G2211 converts to a modifier adding 16% to the base E/M (32% for MSSP ACO and LEAD participants); heavy modifier 25 volume creates offsetting exposure | Audit G2211 capture rate; model modifier 25 exposure on same-day injection and procedure visits; evaluate ACO entry for the 32% tier |
| Cardiology | ~+1% overall | Conversion factor falls to $33.1693 / $32.8409; diagnostic imaging including echocardiography exposed to the 2.5% efficiency adjustment; CAD antiplatelet, beta-blocker, and AFib anticoagulation quality measures removed starting the 2027 performance period | Run the ACC 2027 payment calculator against your own service mix; re-map MIPS measures now that three cardiology measures retire |
| Orthopedic Surgery | −7% specialty-level; −8% to −9% all-in | Specialty-level RVU decline from work, PE, and malpractice changes; facility-based services −8%, nonfacility −5%; the modifier 25 50% cut hits same-day injections, fracture care, and minor procedures directly. AAOS formally opposing | File comments before Sept 14; model the modifier 25 revenue line specifically; reassess site-of-service mix |
| Oncology | Mixed — radiation favorable, drug economics pressured | Radiation oncology +3% in CMS estimates; 340B reimbursement pressure continues after the CMS hospital acquisition-cost survey (Jan 1–Apr 7, 2026) showed sizable 340B vs. non-340B differentials; FY 2027 ICD-10-CM adds cancer-site specificity codes | Review new ICD-10-CM oncology site codes before Oct 1; model 340B exposure if affiliated |
| Radiology | +2% to +3% | Diagnostic radiology +2%, nuclear medicine +2%, interventional radiology +3%, radiation oncology +3%. Countervailing pressure: PA drives ~35% of radiology denials; MA imaging denial rate ~4.94%, higher for spine MRI and PET | Automate imaging PA — 60% of new imaging PA software embeds approval-likelihood prediction; target the 3–7 day manual window |
| Mental / Behavioral Health | Favorable | Telehealth geographic, originating-site, audio-only, and in-person waivers proposed through Dec 31, 2027; audio-only parity retained; LMFTs and LMHCs bill Medicare independently at 75% of psychologist rates; final year of the four-year upward adjustment for timed behavioral health codes, extended to smoking cessation and SBIRT | Verify the in-person requirement compliance trail — a missed 12-month in-person visit produces an automatic denial with no appeal path |
| Gastroenterology | Mixed | Office-based procedures gaining while the same services in HOPD and ASC settings decline; 43889 carries a 90-day global with routine follow-up bundled; 99454 now requires automatic device transmission; 99470 lowered the threshold to allow reimbursement for the first 10 minutes of data review | Audit post-op billing against the 43889 global period; confirm RPM device transmission compliance |
| Dermatology / Podiatry / Ophthalmology | Adverse | Among the highest modifier 25 volume in medicine; dermatology and otolaryngology specifically flagged for significant decreases driven by PE methodology, modifier 25, and IPCI removal | These specialties have the most to gain from a comment filing; quantify same-day E/M-plus-procedure volume as comment evidence |
| Wound Care | Adverse on supply economics | Medicare overhauled skin substitute reimbursement effective Jan 1, 2026, paying as “incident-to supplies” at $127.28/cm² with no reimbursement for discarded amounts; MA downcoding actively affecting wound care revenue | Rebuild product utilization and waste tracking — the no-discard rule makes sizing a margin decision |
| OB/GYN | Structural change ahead | First maternity CPT overhaul in 30+ years effective Jan 1, 2027 — the global obstetric package unbundles into antepartum E/M, new labor management codes 59080–59083, delivery, and postpartum | Open payer contract renegotiation in Q4 2026; global case rates will not survive the code restructure |
Two proposals cut across every specialty in this table. Modifier 25: if your visit pattern routinely pairs an evaluation with a same-day 0/10/90-day global procedure, the 50% reduction is a larger line item than the conversion factor change — and it is the proposal most likely to be modified in the final rule if the comment record is strong. The ±5% PE stabilizer: because it caps year-one movement, the published 2027 specialty impact percentages are a partial view. Practices in specialties with large indirect PE components should model both the year-one capped number and the uncapped steady state.
Ten moves for the week of August 21, 2026 — each tied directly to something in this issue.