Two federal changes land in the next four months. Neither is about care. Both decide whether care gets paid for, and one of them decides whether the patient still has coverage at all.
The first is the FY 2027 ICD-10-CM code set, effective for dates of service on or after October 1, 2026: 190 new codes, 30 deleted, four revised, no grace period. The second is CMS guidance letting states sort Medicaid recipients into three tiers when deciding who is too sick to meet a twenty-hour-a-week work or volunteer requirement. Most expansion states begin January 1.
Here is what to fix before each date.
By raw count this is a lighter revision than recent years. That is the trap. A big update gets a project plan and a meeting series. A small one gets skipped, and the changes that did land sit in high-volume everyday coding rather than in the exotic corners nobody bills.
A sprain code that has sat on orthopedic superbills since the 2015 transition stops working. The whole family of codes for organic solvent poisoning is deleted. The code most outpatient practices use for the most common cause of heel pain is replaced by three codes that demand something the old one never did, which is a side.
Laterality is the part that costs money quietly. A code that never asked left or right now asks. The clinician who has documented that diagnosis the same way for eleven years will keep documenting it the same way, because nothing in the encounter tells her otherwise. The claim does not bounce because anyone made a mistake. It bounces because the note no longer contains a field the code requires.
Three mechanics worth putting on a whiteboard:
The Sarah Matt Briefing goes out Tuesday mornings with the operational read on changes like this one, well before the denial reports land. Subscribe here.
CMS guidance, released through a deck on medicaid.gov, lets states sort medical frailty exemptions into three tiers. Tier 1 exempts on diagnosis alone with no further paperwork; the examples given are end-stage renal disease, ALS, and end-stage cancer. Tier 2 may indicate frailty but requires supporting data, such as billing for recent acute care or particular pharmacy codes. Tier 3 goes to case-by-case review.
The example below is CMS's own, and it is the whole argument.
Vision loss from type 2 diabetes is tier 1. Type 2 diabetes managed on several medications, with possible peripheral neuropathy and no recent admissions, is tier 3. Same disease. Different tier.
The tier 3 patient is the ordinary internal medicine panel. I see her on charity care afternoons and nothing about her chart is dramatic. Four medications. Numbness she mentions in passing, at the end of the visit, while putting her coat on. No admissions, because the regimen is working. Under this structure, the fact that her care is working is what puts her in the review queue.
The dates and the numbers:
The AMA called the approach data-driven and said it could reduce documentation burden on beneficiaries and physicians. That is a fair read of the mechanism. A state that can auto-exempt someone from claims data it already holds spares that person a form, and the tier 1 list does real work. Data-driven and correct are not the same claim, though, and disease groups are lobbying right now to move their patients up. Tier placement is going to be a negotiated outcome before it is a clinical one.
On the surface, nothing. One is a code set and one is a coverage rule, and they will be handled by different people in different meetings.
Underneath, both move a decision that used to live in the clinical note into a field on a form, and both put the consequence on the patient rather than on the person filling out the field.
The heel-pain code will get fixed, because a denial is loud and somebody gets a report about it. The tier 3 patient generates no report at all. She just stops coming.
Written for practice leaders and health system executives who have to turn a federal calendar into a staffing decision. The Briefing ships every Tuesday: subscribe here.
Surgery-trained, currently practicing internal medicine (charity care). Advisory practice focused on clinical AI governance, vendor evaluation, and implementation strategy for health systems and health-tech companies. If you are on a board, an operating team, or a clinical-AI committee trying to draw that line, that is the work I do.
Weekly clinical reality on AI, governance, and what actually works inside health systems.