The long-stay ADL decline quality measure (N028)

The ADL decline measure counts long-stay residents who need more help with basic daily activities than they did at their last comparable assessment. It is one of the long-stay MDS measures in the Five-Star quality rating, worth up to 150 points, and it is built from just four Section GG items.

The CMS name is “Percent of Residents Whose Need for Help with Activities of Daily Living Has Increased (Long Stay),” CMS ID N028.04. It is not risk-adjusted. State programs use it too: it is scored in the Ohio QIP and the Florida QIP.

The four items, and when they became GG items

The measure reads four “late-loss” activities, all from Section GG:

This wasn’t always so. Before October 1, 2023, the measure was built from Section G. When MDS item set v1.18.11 took effect that day, Section G’s functional items were replaced by Section GG, and the QM User’s Manual v16.0 re-specified the measure on the GG items. CMS froze it on Care Compare from April 2024 and brought it back, with new Five-Star cut points, in January 2025. The current manual, v18.0 (effective January 1, 2026), uses the same four GG items and excludes any comparison where the prior or target assessment is dated before October 1, 2023.

One thing that can confuse: the Five-Star Technical Users’ Guide still describes the late-loss ADLs in the old Section G words, “bed mobility, transfer, eating, and toileting.” The measure itself reads the four GG items above.

Each GG item has more than one column, and the measure picks the column from the kind of assessment. A discharge assessment uses the discharge column (for example GG0170B3). A quarterly, annual or other OBRA assessment that isn’t combined with a PPS assessment uses the OBRA/Interim column (GG0170B5). An OBRA admission or PPS 5-Day uses the admission column (GG0170B1).

What puts a resident in the numerator

The measure compares each item on the target assessment with the same item on the prior assessment:

GG codes run from 06 (independent) down to 01 (dependent), so a lower number is more help. A two-point drop on any one item counts. So does a one-point drop on any two items.

The same table adds the rule that surprises people most:

So a refusal (07), not applicable (09), environmental limitation (10) or not attempted for medical or safety reasons (88) is scored as dependent. A resident who ate with supervision (04) last quarter and is coded as refusing (07) this quarter has dropped three points on Eating. That one item puts them in the numerator.

Who is left out

A resident is excluded from the measure when any of these is true:

• They were already at the floor on the prior assessment. All four items were dependent or not attempted, or three were and the fourth was substantial/maximal assistance (02). There’s nowhere left to decline.
• End of life or coma on the target assessment: comatose (B0100), a prognosis of less than six months (J1400) or hospice care (O0110K1b), including when any of those items is dashed.
• A dash on any of the four items on either assessment, unless the resident already triggers.
• No prior assessment in the window below.
• Either assessment dated before October 1, 2023.

Which two assessments get compared

The target is the resident’s latest qualifying assessment: an OBRA assessment, a PPS 5-Day or an OBRA discharge, dated no more than 120 days before the end of the quarter (or the resident’s discharge, if they left during it). The prior is chosen from the target:

With quarterlies about three months apart, the prior is usually the quarterly before. An assessment 45 days or less before the target can’t be the prior. So a significant change assessment done a month before the quarterly is skipped, and the comparison reaches back to the assessment before it.

When it clears

Every quarter is a fresh comparison. Next quarter’s target is compared with its own prior, which is usually the assessment that showed the decline. If the resident holds steady at the new level, nothing drops and they don’t trigger again. A real decline normally counts in one quarter. Five-Star averages the four most recent quarters, so that quarter stays in the rating until it rolls out.

Coding it accurately

Because the measure is only ever a comparison, the error that costs most is inconsistency: the same resident scored differently from one assessment to the next when nothing really changed. The RAI Manual’s rule for every GG self-care and mobility item:

• Score the right three days. On quarterly, annual and other OBRA assessments after admission, the GG assessment period is the ARD plus the two days before it. The evidence is direct observation plus what the resident, care staff, clinicians and family report and the record documents in those three days.
• Save the “not attempted” codes for activities that didn’t happen. The manual allows them only when neither the resident nor a helper did the activity. If staff did all of the work, that is 01, Dependent.
• Two helpers means 01, even when the second helper only stands by.
• Count facility staff only. A “helper” is facility staff or facility-contracted staff. Hospice staff and nursing or CNA students aren’t counted as helpers.
• Avoid dashes. CMS expects a dash to be rare. A dash removes the resident only if they don’t already trigger on the other items.

How Five-Star scores it

ADL decline is one of the 150-point measures, scored in deciles of the national distribution. Lower is better. Points come from your four-quarter rate, using cut points set from 2023Q4–2024Q3 data:

SuperQM runs this comparison for every long-stay resident and shows the items and the prior behind each one. See how it works.