On October 1, 2026 the CMS Home Health PDGM grouper moves from v07.1.26 to v07.2.26 and picks up the FY2027 ICD-10-CM code set. It adds 190 codes, deletes 30, and does not change the clinical group, comorbidity subgroup or principal-diagnosis flags of any code already in the table.
That last sentence is the one I would want a coding team to hear first. The risk in this update is not a code quietly moving to a different group. The risk is billing a code that no longer exists, or reaching for a new unspecified code that CMS assigns to no group at all.
Source: CMS, Oct 2026 HH PPS Grouper Software (HH PDGM v07.2.26), posted August 20, 2026, tables effective October 1, 2026. Every count below was computed by comparing Diagnosis_Codes.txt and the eight supporting tables in that package against the April 2026 package (v07.1.26), then reconciled against the HHGS Summary of Data Changes v07.2.26 that CMS ships inside the same ZIP. Successor codes come from the CDC NCHS ICD-10-CM Conversion Table FY2027. Nothing here is inferred from an ICD-10-CM chapter.
Tip
The grouper chooses its table by the claim's from date. A 30-day period that starts on September 30 groups under the FY2026 table even though it ends in October. Only periods with a from date of October 1, 2026 or later use the FY2027 table. The primary diagnosis eligibility checker shows which table it is running on in the version stamp at the top of the page.
What changes in the PDGM grouper on October 1, 2026?
CMS describes v07.2.26 as a data-only release with no logic or interface changes. I read the Java source in both packages to check, and that is accurate: the v07.2.26 grouper class extends the v07.1.26 class without overriding anything. What changed is the data.
| Change | Count | Where it is stated |
|---|---|---|
| Diagnosis codes added | 190 | Summary of Data Changes, "Added ICD-10-CM Diagnosis Codes" |
| Diagnosis codes deleted | 30 | Summary of Data Changes, "Deleted ICD-10-CM Diagnosis Codes" |
| Code titles revised | 4 | Summary of Data Changes, "Modified ICD-10-CM Diagnosis Code Descriptions" |
| Existing codes whose clinical group changed | 0 | Diagnosis_Codes.txt, v07.1.26 vs v07.2.26 |
| Existing codes whose comorbidity subgroup changed | 0 | Diagnosis_Codes.txt, v07.1.26 vs v07.2.26 |
| Existing codes whose unacceptable or unspecified flag changed | 0 | Diagnosis_Codes.txt, v07.1.26 vs v07.2.26 |
| New codes flagged unacceptable as principal | 9 | Summary of Data Changes, "Unacceptable Principal Diagnosis Additions" |
| New codes flagged unspecified as principal | 11 | Summary of Data Changes, "Unspecified Principal Diagnosis Additions" |
| Codes in the table | 74,719 to 74,879 | Diagnosis_Codes.txt row counts |
The CDC's FY2027 conversion table lists exactly 190 new codes effective October 1, 2026, and they are the same 190 CMS added to the grouper. The two sources agree one for one.
Of the 190 new codes, 77 land in MMTA Other, 66 get no clinical group, 22 go to Musculoskeletal Rehabilitation, 8 to MMTA Cardiac and Circulatory, 6 to MMTA Infectious Disease, Neoplasms and Blood-Forming Diseases, 5 to MMTA Gastrointestinal and Genitourinary, 2 each to MMTA Endocrine and Wound, and 1 each to Behavioral Health and MMTA Surgical Aftercare. Most of the MMTA Other additions are toxic-effect codes (T52, T59, T65) that home health will rarely see as a principal.
Did any existing code change its clinical group?
No. Zero of the 74,689 codes that appear in both tables changed clinical group, comorbidity subgroup, unacceptable-principal flag, unspecified-principal flag, manifestation flag, external-cause flag or primary-awarding flag.
Three smaller things did change for existing codes, and none of them moves a claim.
Three subchapter labels widened. Chronic lower respiratory diseases is now J40-J4B (was J40-J4A) to make room for J4B Pulmonary mycetoma. Diseases of peritoneum and retroperitoneum is now K65-K6A (was K65-K68) for the new pelvic abscess codes. Genetic disorders, not elsewhere classified is now QA0-QA1 (was QA0-QA0). Seventy codes carry the new label. The subchapter ID numbers CMS uses to decide whether a secondary diagnosis is eligible for the comorbidity adjustment did not change, so nothing about comorbidity eligibility moves.
354 thoracic fracture codes gained a code-first note. Every S22 code (fractures of ribs, sternum and thoracic spine) now carries the convention "if applicable, spinal cord injury (S24.0-, S24.1-)". The CDC's FY2027 tabular addenda shows the same edit: "Code also" became "Code first" under S22. In the grouper this is validity flag 05, a caution that the chart may need re-sequencing. CMS still groups the claim. CMS's summary does not list these 354 because it only itemizes new codes; the change is in Code_First_Conditions.txt and Diagnosis_Codes.txt.
Three code-first conventions were reworded. The convention on the R29.7- NIHSS score codes, "code first the type of cerebral infarction", now reads I60-I63 instead of I63 alone, so a hemorrhagic stroke satisfies it. The convention on J99 adds bronchomycosis (B49). The convention on the F02.8- dementia codes corrected one code reference in its wording. Inserting the S22 convention as a new row also renumbered the rows after it, which is why 1,118 codes show a different convention number but the same convention text.
Which deleted codes need a replacement on October 1?
These are the deletions a home health agency is most likely to have on an active chart. The clinical group in the last column is read from the v07.2.26 table, and the successor codes are from the CDC conversion table.
| Deleted code | Was | Successor codes (FY2027) | Successor group |
|---|---|---|---|
| I42.0 Dilated cardiomyopathy | H, MMTA Cardiac | I42.01 Familial-genetic dilated cardiomyopathy; I42.09 Other dilated cardiomyopathy | H |
| I42.00 Dilated cardiomyopathy, unspecified | No group | ||
| I42.8 Other cardiomyopathies | H | I42.81 Arrhythmogenic cardiomyopathy; I42.89 Other cardiomyopathies not elsewhere classified | H |
| I49.8 Other specified cardiac arrhythmias | H, Heart_10 | I49.81 Brugada syndrome; I49.82 Ventricular bigeminy; I49.89 Other specified cardiac arrhythmias NEC | H, Heart_10 |
| D69.1 Qualitative platelet defects | K | D69.11 Glanzmann thrombasthenia; D69.19 Other qualitative platelet defects | K |
| M72.2 Plantar fascial fibromatosis | E, MS Rehab | M72.21 right foot; M72.22 left foot | E |
| M72.20 unspecified foot | No group | ||
| M72.2 (plantar fasciitis was indexed here) | E | M67.A01 Plantar fasciitis, right foot; M67.A02 left foot | E |
| M67.A09 Plantar fasciitis, unspecified foot | No group | ||
| M86.8X1 to M86.8X7 Other osteomyelitis by site | E | Right and left codes, e.g. M86.8X11 right shoulder, M86.8X12 left shoulder | E |
| Unspecified-side codes M86.8X19, X29, X39, X49, X59, X69, X79 | No group, flagged unspecified | ||
| M86.8X8 Other osteomyelitis, other site | E | M86.8X80 skull; M86.8X81 face and sinuses; M86.8X89 other site | E |
| S23.420 Sprain of sternoclavicular joint (A, D, S) | E | None listed in the CDC conversion table | |
| T52.8X- Toxic effect of other organic solvents | A | T52.81- alkenes; T52.82- cycloparaffins; T52.89- other organic solvents | A |
| Z68.1 BMI 19.9 or less, adult | No group, unacceptable | Z68.18 BMI 18.4 or less; Z68.19 BMI 18.5 to 19.9 | No group, unacceptable |
| Z87.890 Personal history of sex reassignment | No group | Z87.8901, Z87.8902, Z87.8903, Z87.8904, Z87.8909 (five codes; there is no .8905 to .8908) | No group, unacceptable |
The pattern to notice: when CMS splits a code, the new "unspecified" child usually gets no clinical group even when the parent had one. I42.0 was an accepted MMTA Cardiac principal for years. From October 1, I42.00 is not. The documentation has to support familial-genetic or other dilated cardiomyopathy, or the chart needs a different principal.
Which new codes have no clinical group?
66 of the 190. A code with no clinical group cannot lead a claim; the grouper returns "Principal diagnosis not assigned to a clinical group". The ones a home health coder could plausibly reach for:
| New code | CMS description | Clinical group |
|---|---|---|
| I42.00 | Dilated cardiomyopathy, unspecified | No group |
| M72.20 | Plantar fascial fibromatosis, unspecified foot | No group |
| M67.A09 | Plantar fasciitis, unspecified foot | No group |
| M86.8X19, X29, X39, X49, X59, X69, X79 | Other osteomyelitis, unspecified side | No group, flagged unspecified |
| J34.839 | Odontogenic sinusitis, unspecified | No group |
| K74.0A | Hepatic fibrosis, moderate fibrosis | No group (but see comorbidity below) |
| QA1.71, QA1.790, QA1.791, QA1.792, QA1.798 | Lynch syndrome and other inherited cancer predisposition syndromes | No group |
| R78.72 | Abnormal gadolinium level in blood | No group |
| Z77.32, Z77.33, Z77.40 to Z77.49 | Exposure to burn pits, Agent Orange, blast overpressure | No group |
The table above holds 24 of the 66. The other 42 are 28 vanishing-twin pregnancy codes (O31.4-), four unspecified ectopic pregnancy codes (O00.129, O00.139, O00.519, O00.529), the two BMI codes (Z68.18, Z68.19), the gadolinium exposure code Z77.013, and the seven personal-history codes (Z86.17, Z87.8901, Z87.8902, Z87.8903, Z87.8904, Z87.8909, Z87.893). Home health would not lead with any of those.
Every new sided code in the osteomyelitis and plantar families does group. The rule for October is simple: document the side.
Which new codes carry a comorbidity subgroup?
Nine of the 190 new codes were placed in an existing comorbidity subgroup. Whether a subgroup can raise the comorbidity adjustment depends on two CMS tables that did not change: Comorbidity_Groups.txt says whether a subgroup can award the low tier on its own, and Comorbidity_Interactions.txt lists the 98 subgroup pairs that award the high tier.
| New code | CMS description | Clinical group | Comorbidity subgroup | What the subgroup can do |
|---|---|---|---|---|
| C78.31 | Secondary malignant neoplasm of larynx | K | Neoplasm_17, Secondary Neoplasm of Respiratory and GI Systems | Low tier on its own |
| C78.32 | Secondary malignant neoplasm of pharynx | K | Neoplasm_17 | Low tier on its own |
| C79.83 | Secondary malignant neoplasm of oral cavity | K | Neoplasm_17 | Low tier on its own |
| I49.81 | Brugada syndrome | H | Heart_10, Dysrhythmias, includes Atrial fibrillation and Atrial flutter | Low tier on its own; high tier paired with Endocrine_4 |
| I49.82 | Ventricular bigeminy | H | Heart_10 | Same as above |
| I49.89 | Other specified cardiac arrhythmias NEC | H | Heart_10 | Same as above |
| E89.830 | Post bariatric hypoglycemia | I | Endocrine_5, Obesity, and Disorders of Metabolism and Fluid Balance | High tier only, paired with Circulatory_10 |
| K74.0A | Hepatic fibrosis, moderate fibrosis | No group | Gastrointestinal_4, Alcoholic Liver Disease, Chronic Hepatitis, Fibrosis and Cirrhosis of the Liver | Neither: not a low-tier subgroup and in no interaction pair |
| K76.83 | Intestinal failure-associated liver disease | J | Gastrointestinal_6, Other Disorders of the Liver | Neither: not a low-tier subgroup and in no interaction pair |
I49.8 was already Heart_10, so the three I49.8x successors keep exactly what the parent had. Nothing here is a new opportunity. It is continuity, and it only holds if the successor code is on the claim.
Which new codes cannot lead a claim at all?
CMS flagged nine new codes as unacceptable in the principal position: Z68.18, Z68.19, Z86.17 (personal history of Clostridioides difficile infection), Z87.8901, Z87.8902, Z87.8903, Z87.8904, Z87.8909 and Z87.893. That takes the unacceptable-principal list from 1,603 codes to 1,611. The one code that left the list, Z68.1, left because it was deleted.
CMS flagged eleven new codes as too unspecified for the principal position: the seven unspecified-side osteomyelitis codes and four unspecified ectopic pregnancy codes (O00.129, O00.139, O00.519, O00.529).
One reconciliation note, because I would rather say it than have someone find it. CMS's summary lists 22 new codes under "Code First Diagnosis Additions". The table itself carries a code-first convention on 17 of them. The five QA1.7- inherited cancer syndrome codes carry none, and the CDC tabular gives QA1 a "code also" note rather than "code first". The tool follows the table, because the table is what the grouper runs.
What should a home health coder do before October 1?
- Pull every active patient carrying one of the 30 deleted codes. I42.0, I42.8, I49.8, D69.1, M72.2, M86.8X1 to M86.8X8, S23.420, T52.8X-, Z68.1 and Z87.890. Any 30-day period with a from date of October 1 or later needs the successor.
- Do not default to the unspecified child. I42.00, M72.20, M67.A09 and the unspecified-side osteomyelitis codes group to nothing. If the record supports the specific code, use it. If it does not, query the physician now rather than in November.
- Check periods that straddle the date. The grouper uses the from date. A period starting September 25 groups under the FY2026 table for its whole 30 days. The next period does not.
- Update the laterality habit for osteomyelitis and plantar conditions. Right, left or no group.
- Re-run anything you had memorized. Codes in the J40, K65 and QA0 ranges have new subchapter labels. The subchapter pages and the list of codes CMS will not accept as principal are rebuilt from the FY2027 table.
How does the primary diagnosis tool handle the change?
The PDGM primary diagnosis eligibility checker and the PDGM coding console read the CMS grouper table directly. Through September 30 they answer from v07.1.26. From October 1, 2026 they answer from v07.2.26 with all 74,879 FY2027 codes. The version is stamped on the page, and every clinical group, comorbidity subgroup and flag the tool shows is CMS's own row, never a guess from the chapter.
If you want the wider context, the CY 2027 home health proposed rule covers the payment side of the same year, and the compliance calendar tracks this and every other CMS date.
Frequently asked questions
Does the October 1, 2026 grouper change the PDGM clinical group of any existing code?
No. Comparing Diagnosis_Codes.txt in CMS grouper v07.1.26 with v07.2.26 row by row, no code that exists in both tables changed its clinical group, its comorbidity subgroup, or its principal-diagnosis flags. The changes are 190 new codes, 30 deleted codes, four revised code titles, and edits to code-first conventions.
Which grouper version applies to a 30-day period that starts in September and ends in October?
The v07.1.26 table. The CMS grouper software selects the table version by the claim's from date, and Version_Range.txt in the package runs v07.1.26 through September 30, 2026 and v07.2.26 from October 1, 2026. A period with a from date of September 30 groups under the FY2026 table even though it ends in October.
Is I42.0 still a valid primary diagnosis after October 1, 2026?
No. I42.0 Dilated cardiomyopathy is deleted for FY2027. Its replacements are I42.00, I42.01 and I42.09. I42.01 and I42.09 keep the MMTA Cardiac and Circulatory group. I42.00 Dilated cardiomyopathy, unspecified is assigned no clinical group and cannot lead a claim.
What happens if a period starting on or after October 1 is billed with a deleted code?
The v07.2.26 table has no row for the code, so the grouper cannot assign a clinical group and the claim comes back for correction. Periods with a from date on or before September 30 are unaffected because they group under the v07.1.26 table.
How many ICD-10-CM codes does the FY2027 grouper table hold?
74,879, up from 74,719 in v07.1.26. Of those, 30,966 (41.4 percent) are assigned no clinical group and 1,611 are flagged unacceptable as a principal diagnosis.
Did the comorbidity subgroups or the interaction pairs change?
No. Comorbidity_Groups.txt (120 subgroups) and Comorbidity_Interactions.txt (98 pairs) are byte-identical between v07.1.26 and v07.2.26. Nine of the 190 new codes were placed into existing subgroups. No existing code moved.
Where do these numbers come from?
From the tables inside the CMS grouper package itself, computed row by row against the April 2026 package, then reconciled against the HHGS Summary of Data Changes v07.2.26 that CMS ships in the same ZIP. The 190 additions also match the CDC FY2027 conversion table one for one.
Sources
- CMS, Home Health PPS Grouper Software: Oct 2026 HH PPS Grouper Software (HH PDGM v07.2.26), posted August 20, 2026. Tables read: Diagnosis_Codes.txt, Clinical_Groups.txt, Comorbidity_Groups.txt, Comorbidity_Interactions.txt, Code_First_Conditions.txt, Diagnosis_Subchapters.txt, Validity_Flags.txt, Return_Codes.txt, FI_Responses.txt and Version_Range.txt, compared with the Apr 2026 package (HH PDGM v07.1.26, posted February 10, 2026).
- CMS, HHGS Summary of Data Changes v07.2.26 (Final), shipped in the documentation folder of the same package.
- CDC NCHS, ICD-10-CM FY2027 files: ICD-10-CM-CONVERSION-TABLE-FY2027.xlsx, icd10cm-addenda-2027.zip (tabular addenda), and the ICD-10-CM Official Guidelines for Coding and Reporting FY2027 (October 1, 2026).
- CMS, CY 2026 HH PPS Final Rule (CMS-1828-F), for the twelve clinical groups and the comorbidity adjustment design the tables implement.
Reza Djangi, OTR/L, is an occupational therapist in home health and the founder of Logicly. He reads the CMS tables so the tools on this site never have to guess.
Keep the whole chart in view
Logicly shows the diagnosis list, the certification period and the visit plan on one screen, so a code that changes on October 1 does not become a returned claim in November.