On September 21, 2026, CMS posted the preliminary CY 2027 Clinical Laboratory Fee Schedule (CLFS) rates — the weighted median private-payer rates that will reset what Medicare pays for clinical diagnostic lab tests starting January 1, 2027. The data comes from the second full reporting cycle required under the Protecting Access to Medicare Act of 2014 (PAMA), and CMS’s conclusion is blunt: Medicare has been paying roughly 16% more for lab services than private payers do.
Of the 1,947 HCPCS codes on the fee schedule, 1,528 had enough reported data to calculate a weighted median. Of those, 1,171 came in lower than the 2026 Medicare rate, 186 came in higher, and 169 were unchanged. The rates were computed from 6,304 laboratories after exclusions (6,411 submitted), using private-payer data collected January 1 through June 30, 2025 and reported May 1 through July 31, 2026.
The realized cut in 2027 will be smaller than the raw data implies, because 2026 federal budget legislation capped CLFS reductions at 15% per year for 2027 through 2029. That cap is doing a lot of work. Before it is applied, the category-level reductions look like this:
| Test category | Uncapped change vs. 2026 | What it means for 2027 |
|---|---|---|
| Genomic sequencing | −23% | Capped at −15% in 2027; remainder lands in 2028 |
| Molecular pathology | −22% | Capped at −15% in 2027; remainder lands in 2028 |
| Microbiology | −19.3% | Capped at −15% in 2027 |
| Immunology | −19.3% | Capped at −15% in 2027 |
| Chemistry | −16% | Capped at −15% in 2027 |
| Proprietary lab analyses | −2.4% | Passes through largely intact |
Read that table as a three-year glide path, not a one-year event. A code whose private-payer median sits 23% below its current Medicare rate takes 15% in 2027 and the balance in 2028. Any practice running an in-office lab — chemistry panels in primary care, urinalysis and cultures in urology, molecular testing in oncology, GI pathogen panels — should be modeling 2027 and 2028 now, because the 2028 number is already visible in this data.
CMS is accepting written comments on the preliminary rates for 30 days — roughly through October 21, 2026 — by email to CLFS_Annual_Public_Meeting@cms.hhs.gov. Final rates publish in November. If a code you bill at volume came in with a median that looks wrong for your market, this is the only window to say so with data.
The bipartisan RESULTS Act (H.R. 5269), cosponsored by 62 Republicans and 59 Democrats, would replace the 15% annual cap with a 5% cap. It has not moved. Until it does, budget against the 15% figure.
One thing practices do not have to plan for next week: a Medicare claims hold. Congress cleared a continuing resolution on September 1 — 370–48 in the House after a 90–6 Senate vote — and the President signed it on September 2, funding the government through December 11, 2026. Last year’s 43-day shutdown began on October 1 and briefly froze telehealth claims; this year, October 1 is a normal claims day. Medicare telehealth flexibilities, extended by the Consolidated Appropriations Act, 2026, run through December 31, 2027. The next funding cliff is 77 days away.
Last week this newsletter covered UnitedHealthcare’s October 1 removal of prior authorization on roughly 1,700 codes. This week the payer story splits three ways: one insurer consolidating authorizations, one insurer automating payment reductions, and one insurer shrinking its Medicare Advantage footprint. All three change what your front desk and billing team do in the fourth quarter.
Aetna is bundling separate prior authorizations for cancer treatment — chemotherapy, immunotherapy, and radiation oncology services — together with the associated imaging, including MRI and CT. Providers previously submitted an average of four separate prior authorizations per Aetna member with cancer. The pilot, launched last summer for lung, breast, and prostate cancers, saw nearly 25% of eligible members use the bundle.
The bundle became available to Medicaid members in eligible states on September 2, with the fall rollout covering Oklahoma, Illinois, Maryland, New Jersey, Virginia, Florida, Kentucky, and West Virginia. Full rollout to Medicare and commercial members is planned for the first half of 2027. Aetna covers 26 million members. Separately, insurers participating in the HHS prior authorization pledge have cut 11% of preapprovals since last summer.
A bundled authorization changes the unit of work, not just the count. An oncology authorization coordinator who used to work four requests per patient now works one larger, more clinically complete request — which means the documentation has to be assembled up front for the whole treatment plan. Practices that build the packet once, at the treatment-planning visit, will get the benefit. Practices that keep submitting piecemeal will not.
Cigna’s E/M Coding Accuracy Policy (R49), in effect since October 1, 2025, allows the payer to automatically reduce level 4 and 5 E/M claims by one level when its claims algorithm decides the submitted diagnoses do not support the complexity billed. The codes in scope are 99204, 99205, 99214, 99215, 99244, and 99245. On September 3, 2026, the American Academy of Otolaryngology–Head and Neck Surgery submitted formal opposition letters to Cigna challenging the policy. Several Blue Cross Blue Shield licensees have acknowledged similar recoding practices in specific situations.
The practical point: a downcoded claim is paid, not denied, so it rarely appears on a denial report. The only way to see it is payment variance analysis — comparing the CPT billed to the CPT adjudicated on the 835. If your denial dashboard is clean and your Cigna E/M revenue per visit is falling, this is why.
Humana confirmed on its July 29 earnings call that Medicare Advantage plan exits for 2027 will affect about 600,000 members — roughly 8% of its 7.2 million MA members — concentrated in plans rated 3.5 stars or lower. Federally required non-renewal notices are dated October 2, and Medicare open enrollment runs October 15 through December 7. Humana recaptured just over 40% of members affected by its 2025 exits and expects a similar share this time.
Pull your active Humana MA patients by plan ID now and cross-reference against the exiting plans in your market. For every patient in an exiting plan, January 1 means a new payer, a new eligibility record, possibly a new prior authorization for an in-progress treatment course, and a new network status for your practice. Flag those accounts for re-verification at the first January visit — not at the first denial.
At its Health and Life Sciences Summit in Orlando on September 23, 2026, Oracle Health announced a set of AI capabilities spanning scheduling, financial clearance, clinical documentation, charge capture, billing, and payment. The framing was explicit: move AI upstream, to catch denials, missing information, authorization problems, and inaccurate charges before they become rework. The capabilities are planned for general availability “in the coming months.”
| Capability | Cycle stage | What it does |
|---|---|---|
| Autonomous prior authorization | Front end | Evaluates payer criteria, retrieves documentation rules, pre-populates clinical fields, attaches chart evidence |
| Clinical documentation integrity | Mid-cycle | Analyzes notes in real time against reimbursement criteria and flags gaps |
| Charge capture & integrity | Mid-cycle | Reads procedural narratives and service codes to prevent revenue leakage |
| Autonomous professional fee coding | Mid-cycle | Suggests CPT, HCPCS, and ICD-10 codes inside the coder workflow |
| Automated appeal management | Back end | Evaluates remittance advice, isolates underpayments, compiles appeal packets |
No customer names, performance metrics, or dates were attached. That is not unusual for a conference announcement, but it matters for how you file it: this is a roadmap, not a result. Note also where the list starts — with prior authorization — and where it ends — with underpayment detection rather than denial appeals. That last detail lines up with the Cigna downcoding problem in Section 02: the revenue a practice loses to a paid-but-reduced claim is invisible to a denial-centric workflow.
Black Book’s September 2026 preconference flash survey of 1,184 physician-practice executives, released September 24 ahead of next week’s MGMA Annual Conference, found that only 19% of current AI users report meaningful net work reduction. Buyers have responded by raising the bar: 96% of AI evaluators require integration with existing systems, 89% require defined performance metrics before expanding, and 85% require measurable first-year productivity outcomes. A companion Black Book survey of 1,919 clinicians found 81% require daily workarounds and 65% lose 30 or more minutes a day to avoidable digital work.
Black Book’s own summary puts it precisely: the conversation “has shifted from what a system can do to what work the organization can actually stop doing after implementation.” That is the right procurement question. Before you sign, write down the specific task, the FTE hours it consumes today, and who stops doing it on go-live day. If the vendor cannot help you fill in that last column, the 81% will stay at 81%.
Ambient documentation is the AI category that has produced measurable change — and payers have noticed what it changed. Across large health systems adopting ambient scribes, visits have shifted toward higher-intensity E/M codes; one system saw a 5% increase in Level 5 encounters, worth more than $1,000 per provider per month. One 2026 study projected up to $2.3 billion in added healthcare costs industry-wide. Payers are answering with the automated downcoding described in Section 02. The operational lesson: if an ambient tool raises your level-of-service mix, your documentation has to support the new level on audit, because the payer’s algorithm will test it.
October 1, 2026 falls on a Thursday. That is the effective date for four separate code-file changes, each governed by date of service, not date of submission. Services on Wednesday, September 30 use the old files regardless of when they are billed; services on Thursday use the new ones.
| File | Posted | Effective | Who it hits |
|---|---|---|---|
| FY 2027 ICD-10-CM | Summer 2026 (with 2027 Conversion Table) | October 1, 2026 | Every practice |
| FY 2027 ICD-10-PCS (100+ new codes) | Summer 2026 | October 1, 2026 | Inpatient coders; surgeons whose hospital coding drives co-management |
| October 2026 HCPCS Level II quarterly | Alpha-numeric file updated September 10, 2026 | October 1, 2026 | Drugs, biologicals, supplies, DME |
| NCCI PTP practitioner edits, Q4 | September 1, 2026 | October 1, 2026 | Every practice billing code pairs |
| UHC MA Observation & Discharge policy | UHC reimbursement bulletin | October 1, 2026 | Hospitalists; any practice billing observation care to UHC MA |
Last week’s issue walked through the ICD-10-CM deletions and demotions in detail (I42.0, Z68.1, the S23.42 series). This week the question is narrower: is the new code set actually installed and working? With six days left, there is time for exactly one round of testing.
The October HCPCS quarterly is where drug and biological codes change mid-year. Any practice buying and billing drugs — oncology, rheumatology, ophthalmology, urology, allergy — should reconcile its chargemaster against the October file before the first infusion day in October. A J-code that changed on October 1 and still sits in your fee schedule under the old descriptor is a denial on your highest-dollar claim line.
Cigna’s R49 policy (Section 02) targets 99204, 99205, 99214, 99215, 99244, and 99245 based on the diagnoses on the claim. That turns diagnosis specificity into an E/M payment variable. A level 4 visit coded with a vague or unspecified diagnosis is the claim an algorithm downcodes; the same visit coded with the specific, supported comorbidities the physician actually addressed is the claim that holds. The October 1 ICD-10-CM update adds specificity in several areas — use it.
Speaking at the Jefferies Healthcare Services and Technology Conference in Nashville on September 15, HCA Healthcare CFO Mike Marks described a pattern every practice billing office will recognize: patients are owing more because of benefit design across employer plans, Medicare Advantage, and the exchanges — including movement from silver to bronze plans — but they are not paying more. HCA has seen a slowdown in its ability to collect out-of-pocket amounts, which it attributes to energy costs, premiums, and inflation.
When the largest hospital operator in the country says its self-pay collections are slowing, a physician practice with a fraction of HCA’s patient financial services capacity should assume the same is happening to it — and that it is happening faster. An August 11 MGMA Stat poll found 32% of medical groups reporting higher no-show rates year to date, 58% about the same, and only 10% lower. Higher patient cost-sharing shows up first as a missed appointment and second as an unpaid balance.
A patient who moved from a silver to a bronze exchange plan for 2026 has a larger deductible, and the deductible resets January 1. That makes October through December the months when bronze-plan patients have finally met, or nearly met, their deductibles — and January through March the months when they owe the most per visit. Two practical consequences:
| Metric | Warning zone | Target |
|---|---|---|
| Days in A/R | > 45 | < 35 |
| Patient A/R share of total A/R | > 30% | < 20% |
| Point-of-service collection rate (on known patient responsibility) | < 50% | ≥ 70% |
| Estimates delivered before the visit | < 50% of scheduled visits | ≥ 90% |
| Card on file / payment plan at estimate | Ad hoc | Standard workflow |
| No-show rate | > 10% | < 5% |
| First-pass denial rate | > 10% | < 5% |
| Patient bad debt (% of net patient revenue) | > 5% | < 3% |
Automate the reminder, the estimate, and the payment link as one message, sent 72 hours before the visit. Each of the three on its own is a known workflow; combined, they convert a no-show risk into a scheduled payment. It is the cheapest revenue velocity project available to a practice this quarter.
Black Book Research’s September 24 release names its top-rated medical group technology and services vendors, drawn from 2026 evaluation studies covering 9,018 medical practices, alongside a flash survey of 1,184 practice executives on what they plan to buy in 2027. The spending intentions are clear — and so is the skepticism.
| 2027 technology intention | Share of practices |
|---|---|
| Revenue cycle automation | 70% |
| Cybersecurity / managed IT services | 78% |
| Value-based care technology | 41% |
| Patient access / scheduling technology | 37% |
| AI-enabled administrative or clinical workflows | 32% |
| Analytics / financial intelligence | 19% |
| Outsourced administrative functions | 16% |
| Major EHR / PM replacement | 11% |
The buying timeline is slow and getting slower: 17% expect a selection by Q1 2027, 20% in Q2, and 33% anticipate a purchase in late 2027. 93% favor solutions that integrate with their existing environment, and 70% prioritize a documented implementation and migration approach.
R1’s acquisition of Humata Health, announced in August, is expected to close at the end of the third quarter — which is next week. Humata’s agentic prior authorization technology reports a 96% first-pass approval rate, a 30% reduction in write-offs, and an 83% reduction in rescheduled appointments. It will be folded into R1’s Phare OS. Combined with Oracle’s autonomous prior authorization announcement and Infinx’s Black Book ranking, prior authorization is now the most crowded AI category in the revenue cycle — arriving just as UnitedHealthcare, Aetna, and the CMS-0057-F API mandate shrink the manual version of the work.
Elsewhere in September: Savista acquired ABW Medical to expand ambulatory revenue cycle services; Juno Health partnered with SlicedHealth to embed revenue cycle tools in Juno EHR; and ECLAT Health Solutions launched evaire Recover, an agentic AI platform for A/R and denials. Next week’s MGMA Annual Conference (September 27–30, San Antonio — the association’s 100th year) is where most of these vendors will be pitching. Bring the Black Book 85% standard with you: measurable first-year productivity, in writing.
On September 17, 2026, the HHS Office of Inspector General published compliance audits finding that Medicare Advantage plans operated by Humana and UnitedHealthcare received an estimated $177.8 million in overpayments for 2020 and 2021 after submitting diagnosis codes not supported by patients’ medical records.
| Plan | Unsupported sample | Extrapolated overpayment |
|---|---|---|
| HumanaChoice (H5216) | 178 of 220 (81%) | ~$130.9 million |
| UnitedHealthcare of Wisconsin | 183 of 250 (73%) | ~$47 million |
OIG examined 11 groups of high-risk diagnosis codes, including acute stroke, heart attack, embolism, and sepsis — the conditions most often coded from a historical problem list rather than from a current, treated encounter. Both insurers said the methodology was skewed toward finding overpayments and that they do not plan to refund. Earlier in 2026, OIG issued similar audits of Blue Cross Blue Shield of Alabama, Gateway Health Plan, and Priority Health.
Medicare Advantage risk-adjustment diagnoses are drawn from provider encounter records — the physician claim and the note behind it. The plan bears the refund; the practice bears the consequences that follow: more chart retrieval requests, more retrospective risk-adjustment reviews, and more payer scrutiny of diagnoses that appear on a claim without a matching assessment and plan in the note. The documentation standard is the same one CMS applies in RADV audits: a diagnosis counts if the record shows it was monitored, evaluated, assessed, or treated at that encounter.
The four named conditions are exactly the ones that tend to be mis-carried in ambulatory records: “history of stroke” coded as acute stroke, a months-old MI still coded as acute, a resolved embolism still on the problem list. Run a query this week for acute stroke, acute MI, and embolism codes on office-visit claims. Every hit without a same-encounter assessment and plan is a finding waiting to happen — and, with the ambient-scribe coding intensity trend in Section 03, one more reason payers are pulling charts.
CMS separately barred 11 medical supply companies with more than $3.4 billion in suspected fraudulent billing in 2025 and 2026 from future Medicare Advantage and Part D payments — a reminder that DME referral relationships deserve an annual review.
| Item | Date | Days from today |
|---|---|---|
| FY 2027 ICD-10-CM/PCS, October HCPCS, NCCI Q4 edits | October 1, 2026 | 6 |
| Humana MA non-renewal notices | Dated October 2, 2026 | 7 |
| Medicare open enrollment | October 15 – December 7, 2026 | 20 |
| Preliminary CY 2027 CLFS comment window closes | ~October 21, 2026 | ~26 |
| CY 2027 PFS and CLFS final rates expected | November 2026 | ~40 |
| Continuing resolution expires | December 11, 2026 | 77 |
| CMS-0057-F payer FHIR APIs; CPT 2027 | January 1, 2027 | 98 |
| Medicare telehealth flexibilities expire | December 31, 2027 | 462 |
On price transparency, the Transparency in Coverage proposed rule (CMS-9882-P) would move payer in-network rate and out-of-network allowed-amount files from monthly to quarterly updates and organize them by provider network rather than by plan. Comments closed February 23, 2026; a final rule has not yet been published. When it is, payer MRFs become a more usable benchmark for your own contract rates.
Black Book’s September flash survey is a practice-side survey, and its operational findings read like a portrait of the independent group:
Add 82% citing staffing shortages or labor costs and 79% reporting increased concern about cybersecurity risk. Only 16% are considering outsourcing administrative functions — which means the other 84% intend to solve a staffing problem with the staff they have, plus software.
The lab fee schedule was set without you. Only about 1% of physician office labs reported private-payer data in this cycle, compared with roughly 2% of all Medicare Part B labs and fewer than 20% of hospital outreach labs; hospital outreach labs made up 13.6% of all reporting labs. The rates your in-office lab will be paid in 2027 were derived largely from independent and hospital labs with very different cost structures and payer contracts. A small practice running CBCs, metabolic panels, and urinalysis in-house will absorb up to a 15% cut on rates it had essentially no role in setting. Run the in-office lab P&L against the capped 2027 rates and the 2028 remainder before renewing analyzer leases or reagent contracts.
Downcoding hits hardest where there is no one to catch it. A health system has a payer-contracting team running payment variance reports. A four-physician practice has a biller who posts the 835 and moves on. Cigna’s R49 reductions are paid claims — they do not generate a denial for anyone to work. The small practice is the one least equipped to notice.
Humana’s MA exits cost a small practice proportionally more. A practice with a concentrated Medicare panel in a market where Humana is exiting can see a meaningful share of its patients change payers on January 1. Each one is a re-verification, a new card, and potentially a new authorization for an ongoing treatment course.
The retention finding is the one to act on. When 72% of practices can’t keep revenue cycle staff, every workflow that depends on one experienced biller’s memory is a single point of failure. Documenting payer-specific rules — UHC’s code-and-plan PA list, Aetna’s oncology bundle, Cigna’s R49 codes — in a shared payer matrix is the lowest-cost retention insurance available.
| Metric | Where most independents are | Where you should be |
|---|---|---|
| Days in A/R | 42–50 | < 35 |
| First-pass denial rate | 10–14% | < 8% |
| Payment variance review (billed vs. adjudicated CPT) | Never or ad hoc | Monthly, by payer |
| Point-of-service collection on known patient balances | 30–50% | ≥ 70% |
| Cost to collect | 5–8% of net revenue | < 4% |
| Documented payer rules matrix | In one person’s head | Shared, updated monthly |
| In-office lab margin modeled at 2027 rates | Not modeled | Modeled for 2027 and 2028 |
| AI tools with a named FTE-hour reduction target | None | Every tool, measured at 90 days |
(1) Confirm the October 1 code files are live and test them this week; (2) run a Cigna payment variance report on 99214/99215 for the last six months; (3) pull Humana MA patients in exiting plans and flag them for January re-verification; (4) model your in-office lab at the capped 2027 rates and submit a comment if a high-volume code looks wrong; (5) write down every payer rule your best biller knows.
This week’s specialty picture is driven by three cross-cutting events — the preliminary lab fee schedule, Aetna’s oncology bundle, and Cigna’s E/M downcoding — landing unevenly across service lines.
| Specialty | This week’s update | Data point | Priority action |
|---|---|---|---|
| Primary Care | In-office chemistry and immunology labs face capped CLFS cuts; Cigna R49 targets 99214/99215 | Chemistry −16%, capped at −15% | Model in-office lab margin at 2027 rates; run Cigna E/M payment variance |
| Cardiology | OIG’s high-risk diagnosis groups include heart attack, embolism, and stroke; UHC removes PA on cardiology codes October 1 | 73–81% of sampled MA enrollee-years unsupported | Audit acute MI and embolism codes on office claims for same-visit assessment and plan |
| Orthopedics | UHC PA removal on arthroscopy, fracture care, and injections takes effect; Humana MA exits shift surgical patients’ payers January 1 | 600,000 Humana MA members changing plans | Re-verify authorization status for surgeries scheduled after January 1 on exiting-plan patients |
| Oncology | Aetna bundles chemotherapy, immunotherapy, radiation, and MRI/CT into one authorization; Medicaid now, Medicare and commercial in 1H 2027 | 4 authorizations → 1 | Build the full treatment-plan authorization packet at the planning visit |
| Radiology | Cancer-related MRI and CT fold into Aetna’s oncology bundle; 2027 OPPS proposes PFS-equivalent rates for certain non-contrast imaging in off-campus hospital departments | Site-neutral imaging proposal pending final rule | Confirm imaging is covered under the ordering oncologist’s bundle before requesting a separate auth |
| Pathology / Molecular Labs | Largest uncapped CLFS reductions of any category | Genomic −23%, molecular −22% | Comment by ~October 21; model the 2028 remainder above the 15% cap |
| Gastroenterology | Microbiology panels (including in-house GI pathogen testing) face CLFS reductions; UHC removes colonoscopy and endoscopy PA October 1 | Microbiology −19.3%, capped at −15% | Re-price in-house pathogen panel economics; update scheduling scripts |
| Neurology | Consult codes 99244 and 99245 are within Cigna’s R49 downcoding scope; OIG’s acute stroke finding applies to follow-up visit coding | 6 E/M codes in R49 scope | Code “history of stroke” (Z86.73) rather than acute stroke on follow-up visits |
| Mental / Behavioral Health | Medicare in-person requirement for tele-mental health waived through December 31, 2027; audio-only permitted through the same date | 462 days of waiver remaining | Plan 2028 in-person visit capacity now; patients who start tele-mental health by 12/31/2027 skip the initial six-month in-person visit |
| Otolaryngology | AAO-HNS filed formal opposition to Cigna’s R49 downcoding policy on September 3 | Policy in force since October 1, 2025 | Pull 12 months of Cigna level 4/5 remittances and quantify downcoded dollars |
Oncology is on both sides of this week’s news. Aetna’s bundle is the most meaningful prior authorization simplification any oncology practice has seen this year, and molecular pathology and genomic sequencing face the steepest lab fee reductions. A community oncology practice with an in-house molecular lab should treat the two as one budget conversation: authorization labor goes down, lab margin goes down, and the net depends on its own mix.
Ten moves for the week of September 25, 2026 — each tied directly to something in this issue.