Skip to content

QM User’s Manual v18.0: what changed, and the new antipsychotic QM

v18.0 (Jan 2026) re-specified the long-stay antipsychotic measure (N047.01) with claims and encounter data, and changed its exclusions. Also the DFS adjuster change and the March 2026 errata.

September 29, 2026 · 5 min read · SuperLTC

Version 18.0 of CMS’s MDS 3.0 Quality Measures User’s Manual took effect January 1, 2026. For most measures, nothing about the calculation changed. The exception is the one that matters most to many buildings: the long-stay antipsychotic measure no longer comes from the MDS alone. Here is what changed from v17.0, and what it means for your rate and your stars.

What v18.0 changed

CMS lists two notable changes, plus a small item-number update carried over from the October 2025 MDS:

MeasureChange in v18.0
Antipsychotic medication, long stayRe-specified to add Medicare and Medicaid claims and encounter data to the MDS. The CMS ID changed from N031.04 to N047.01.
Discharge Function Score (S042.03)The “no physical or occupational therapy” risk adjuster now comes from the discharge item O0425B and O0425C, because the O0400B and O0400C therapy minutes were retired from the MDS.
Ability to walk independently worsened (N035.06)Its sex risk adjuster now reads A0810, Sex, which replaced A0800, Gender, on October 1, 2025.

CMS says the MDS transition did not affect any other nursing home quality measure.

The antipsychotic measure now counts claims

“The Percent of Residents Who Received an Antipsychotic Medication (Long Stay) (CMS ID: N047.01) has been re-specified to include Medicare and Medicaid claims and encounter data, in addition to Minimum Data Set 3.0 (MDS) data.”
MDS 3.0 QM User’s Manual v18.0, Notable Changes, page iv

Under v17.0, a long-stay resident was in the numerator only if N0415A1, Antipsychotic, was checked on the target assessment. That item looks back seven days. Under v18.0, a resident is also in the numerator if a claim or encounter record shows an antipsychotic during the target period while they were in the building:

“The resident has a claim or encounter record for antipsychotic medication during the target period while the resident is in the facility.”
MDS 3.0 QM User’s Manual v18.0, Chapter 2, Table 2-31, page 48

Two kinds of records count: a Medicare Part D or Medicaid pharmacy record, timed by its fill date, and a physician-administered antipsychotic (such as a long-acting injection) on a Medicare outpatient or physician claim or a Medicaid claim, timed by its service date. A fill or injection while the resident was discharged from the building doesn’t count.

CMS explained why in its memo announcing the change: the MDS may not capture every resident prescribed an antipsychotic, and claims catch prescribing that falls within the stay but outside the seven-day MDS look-back. CMS put the national rate at 14.64% under the old measure and 16.98% under the new one.

The exclusions changed too

v17.0 (N031.04)v18.0 (N047.01)
Schizophrenia, Tourette’s, Huntington’sExcluded if coded on the MDS (Tourette’s could come from the prior assessment).Must be coded on the target or prior assessment and appear as a diagnosis on a Medicare or Medicaid claim or encounter within a one-year look-back window.
HospiceNot an exclusion.Excluded if the resident had Medicare Part A or Medicaid hospice, or was enrolled in hospice, in any month from the start of the target period to the end of the episode.
EnrollmentNot a condition.Excluded unless continuously enrolled, month by month, in Medicare (fee-for-service or Medicare Advantage) or in Medicaid only, over set windows. From the start of the target period, Medicare enrollment must include Part D.
N0415A1 not answered (dash)Excluded if not otherwise in the numerator.No longer listed as an exclusion.

The schizophrenia change is the one to act on. A diagnosis coded on the MDS no longer excludes a resident by itself; there has to be a matching diagnosis on the claims. CMS’s memo says the aim is to stop over-reported schizophrenia diagnoses from excluding residents. The hospice exclusion works the other way and takes out residents who used to count.

See your building’s long-stay antipsychotic rate, free.

When it reached Care Compare and Five-Star

  • January 2026. The re-specified measure replaced the old one on Care Compare and in the Five-Star quality measure rating. Its Five-Star cut points were set to split providers into ten equal deciles.
  • April 2026. CMS found that some residents with a history of schizophrenia had not been excluded in the January data and fixed the calculation. It says the fix did not significantly change scores, but it reset the measure’s cut points using data from 2024Q4 through 2025Q3.
  • iQIES. The re-specified measure is on the facility- and resident-level QM reports back to the second quarter of 2024.

In Five-Star the measure is worth 15 to 150 points. Under the cut points in the July 2026 Technical Users’ Guide, a rate of 4.26% or less earns 150 points, and a rate above 28.36% earns 15.

The March 2026 errata

On March 30, 2026, CMS posted an errata to v18.0. The first printing said public reporting would start in January 2026 using third-quarter 2025 data. The errata took the public-reporting language out of the manual, pointing readers to the Five-Star Technical Users’ Guide instead, and said the measure is on iQIES reports from Q2 2024. It also replaced the supplemental code lists (the NDC and HCPCS codes that identify antipsychotics in claims, and the hospice and exclusion-diagnosis codes), because the files first posted with v18.0 were not up to date.

What to do with it

  • Code N0415A1 exactly as the RAI manual says. The MDS still counts, and claims are added on top of it, not instead of it.
  • An antipsychotic filled or given outside the seven-day look-back can now show up through claims. Plan the GDR conversation around the whole quarter, not the ARD.
  • For a resident excluded for schizophrenia, confirm the diagnosis is on the claims, not only on the MDS. If it isn’t, the resident is in the measure.
  • Your own records don’t include Medicare and Medicaid claims, so a count built from the MDS or your orders can run lower than CMS’s. The iQIES resident-level report shows CMS’s answer for each resident.

For the measure itself, item by item, see the long-stay antipsychotic measure (N047.01). SuperQM reconciles its own count against CMS’s resident-level report so the difference is visible, resident by resident.

See where your building stands.

Your stars, your measures and your state program. Free, from public data.

Rather talk first? Book 20 minutes