Today is the effective date for the April 2026 ICD-10-CM update, and if you were expecting a batch of new diagnosis codes you can relax. There are none. No additions, no deletions, no revisions to the code set itself. The FY 2026 total stays where it was on October 1, 2025. The ICD-10-PCS side of the release adds procedure codes that matter to hospital inpatient coders, but for physician practices the diagnosis side is what matters, and the diagnosis side is all about instructional notes and the index.

That does not make it unimportant. Instructional notes are what your claim scrubber and your coders use to decide whether two codes can appear together and in what order. Sixteen Excludes1 notes have become Excludes2 notes. A "code first" note and a "use additional code" note have both become "code also". A few index entries moved, including the one that tells a coder where Parkinson's disease goes. Each of those changes either lifts an edit that was denying claims or changes what the correct sequence is. Practices that do not update their encoder and scrubber rules today will be applying last year's edits to this year's claims.

Key takeaways

  • No new, deleted or revised ICD-10-CM codes on April 1, 2026. The changes are all in notes and the Alphabetic Index.
  • Sixteen Excludes1 notes became Excludes2, so code pairs that a scrubber rejected in March are allowed in April when documentation supports both.
  • I16.1 and H40.84 now carry "code also" instead of a mandated sequence. Order is decided by the circumstances of the encounter.
  • "Parkinson's disease" now indexes to G20.A1, not G20.C. Problem list mappings built on the old path need a check.

Excludes1 to Excludes2: what the change means

An Excludes1 note means the two conditions cannot be coded together; they are mutually exclusive by definition. An Excludes2 note means the excluded condition is not part of the code, but a patient can have both, so both may be reported. When a note moves from Excludes1 to Excludes2, a combination that was previously an error becomes allowed.

The sixteen conversions are spread across the neoplasm, blood, endocrine, nervous system, circulatory, respiratory, symptoms, injury and Z-code chapters. The categories touched include D18 and D49 in neoplasms, D65, D70 and D72 in blood disorders, E21.2 and E53.8 in endocrine, G36.0 and G43 in the nervous system chapter, I06 in circulatory, J95.82 in respiratory, R12 in symptoms, T43 in poisoning, and Z79.891 among the Z codes. Two examples that primary care, pulmonary and pain practices will meet quickly:

CodePreviouslyNowPractical effect
J95.82 Postprocedural respiratory failureExcludes1: J96.- Respiratory failure, not elsewhere classifiedExcludes2; both may be reported when clinically appropriateScrubber edit that rejected the pair should be removed
Z79.891 Long term (current) use of opiate analgesicExcludes1 against the opioid-related disorder codes in F11.-Excludes2; both may be reportedPatients on prescribed long-term opioids with a documented use disorder can carry both codes
G43.- MigraineExcludes1 against a related headache categoryExcludes2Check the specific note in the tabular before removing the edit
R12 HeartburnExcludes1 against a related symptom codeExcludes2Same: read the note, then update the pair

The Z79.891 change is the one we think will matter most in day-to-day work. Under the old Excludes1 note a patient with documented opioid dependence who was also on long-term prescribed opioid therapy for pain forced coders to choose, and the choice affected risk adjustment and quality measure logic. Now both facts can be recorded. For the other pairs, do not take our word or a summary's word for exactly which code is excluded; open the April 2026 tabular addenda and read the note, because the addenda name the code and a summary usually names only the category.

Sequencing notes that changed

Two instructional notes moved from a mandated order to "code also", which leaves sequencing to the circumstances of the encounter.

CodePreviouslyNowPractical effect
I16.1 Hypertensive emergency"Use additional code" for the underlying hypertension"Code also" the underlying conditionReport both; order by the reason for the encounter
H40.84 Neovascular secondary angle-closure glaucoma"Code first" the underlying condition"Code also" the underlying conditionSequencing no longer mandated; report both
F07.81 Postconcussional syndromeNo sequencing instructionNew "code also" note for sequela of concussion; Excludes1 note expandedReport the concussion sequela code alongside F07.81

A worked example for I16.1: a 58-year-old patient presents to an urgent care with a blood pressure of 224 over 128 and acute end-organ symptoms, on a background of known essential hypertension. Before April 1 the instruction was to report I16.1 with an additional code for the hypertension. After April 1 the instruction is "code also", and either order is permitted. For an emergency visit the hypertensive emergency is the reason for the encounter and would ordinarily be sequenced first; for a follow-up visit about the chronic hypertension where the emergency is documented as resolved, the chronic code leads. The practical change is that a scrubber edit forcing I10 ahead of I16.1, or the reverse, is now wrong.

The Parkinson's disease index change

The Alphabetic Index entry for Parkinson's disease has been restructured. The single main term that covered Parkinson's disease, syndrome and tremor has been replaced with separate entries: "Parkinson's disease" now points to Disease, Parkinson's, and "Parkinson's syndrome or tremor" points to Parkinsonism. The effect is that Parkinson's disease without further specification now indexes to G20.A1 (Parkinson's disease without dyskinesia, without mention of fluctuations) rather than G20.C (Parkinsonism, unspecified).

This is a correction more than a change; G20.C was never a good fit for a patient with a Parkinson's disease diagnosis, and the G20.A and G20.B subcodes that distinguish dyskinesia and fluctuations have existed since October 1, 2023. But if your encoder or your problem list mapping was built on the old index path, patients will continue to be coded to G20.C until someone updates the mapping. For neurology and primary care practices with Medicare Advantage panels, G20.A1 and G20.C sit differently in the risk adjustment model, so this is worth a specific check. It is also a documentation prompt: if the neurologist documents dyskinesia or motor fluctuations, G20.A2, G20.B1 or G20.B2 is the more specific code, and the index change does not choose among those for you.

A neoplasm indexing change worth knowing

An index change also directs the entry for neuroendocrine tumor from D3A.8 (other benign neuroendocrine tumors) to the C7A.- malignant neuroendocrine tumor categories. For practices this mostly affects oncology and gastroenterology. For hospitals it changes severity because C7A codes carry a complication or comorbidity designation. Confirm the documentation actually supports malignant behavior before following the index; the index points, the documentation decides, and a pathology report that says benign is coded as benign regardless of where the index sends a coder who only has the word "neuroendocrine".

What did not change

There are no changes to the Table of Neoplasms, the Table of Drugs and Chemicals, the External Cause Index, or the Official Guidelines for Coding and Reporting. The guidelines you trained on in October still apply. The next full update, with new codes for FY 2027, takes effect October 1, 2026; the proposals were discussed at the spring 2026 ICD-10 Coordination and Maintenance Committee meeting and the final list is usually posted in June.

Questions we hear

Do we have to recode March encounters?

No. Code changes are effective by date of service. A March 28 encounter is coded under the rules in effect on March 28. A claim submitted in April for a March visit uses March rules, and a claim for an April 1 visit uses the new notes.

Our scrubber vendor says it updates quarterly. Is that fast enough?

Not really. If the April edits are not loaded until June, you will get eight weeks of false rejections on the changed pairs. Ask for the date and, in the meantime, have a coder override the specific edits that changed and log each override so the vendor can see the volume.

Where does this fit in coder training?

Mid-year updates are small, which makes them a good teaching moment for newer coders because the reasoning is visible: why is this pair now allowed, and what does that mean for the documentation we ask providers for? Our live coding courses cover each update as it lands, and the Revelrex EHR training environment lets coders practice the changed combinations on realistic charts. If you want an outside read on your scrubber edits, that is part of our RCM audit.

What to do this week

  1. Confirm with your EHR and encoder vendor that the April 1, 2026 files are loaded. Ask for the release note, not a verbal yes.
  2. Export your claim scrubber's diagnosis pair edits and remove or update the sixteen Excludes1 pairs that changed, reading each note in the tabular addenda first.
  3. Remove any scrubber rule that forces a sequence for I16.1 or H40.84.
  4. Search the problem list mappings for Parkinson's disease and confirm they resolve to G20.A- codes, not G20.C, unless the documentation says parkinsonism.
  5. Brief the coders. Fifteen minutes on the two tables above covers most of what they will see.
  6. Watch April remittances for CO-16 and CO-11 denials on the affected codes. A payer whose edits lag the update will deny valid combinations, and those need an appeal with the updated note quoted.