CMS PBJ Staffing: 35.6% of Nursing Homes Still Below the 3.48 HPRD Floor

Nursing homes report who worked, for how long, and against what census. That is the point of the CMS Payroll-Based Journal (PBJ): a daily, payroll-tied ledger of nurse hours that Care Compare then compresses into hours per resident day (HPRD). When federal regulators published numerical minimums in April 2024 — 3.48 total nurse HPRD, including at least 0.55 registered-nurse (RN) HPRD and 2.45 nurse-aide (NA) HPRD — those numbers became the natural yardstick for reading PBJ prints. The interactive dashboard above scores the latest quarter against that yardstick: how many homes still sit below the floor, which staff type drives the gap, and how ownership, size, and geography shift the picture.

The core answer for the Q4 2024 PBJ vintage is blunt. Of 14,573 homes reporting nurse staffing, 64.4% posted total nurse staff at or above 3.48 HPRD. That leaves 35.6% — roughly 5,190 facilities — still below the published total-nurse floor, even as the national mean total nurse HPRD sits at 3.75. Means above the floor and a large minority below it are not a contradiction; they are a distribution with a long low tail.

What the federal floor actually said

CMS’s Minimum Staffing Standards final rule (CMS-3442-F) set three interlocking numerical requirements for long-term care facilities. Facilities had to provide at least 3.48 HPRD of total direct nursing care. Inside that total, at least 0.55 HPRD had to come from RNs and at least 2.45 HPRD from nurse aides. The remaining 0.48 HPRD could be filled by any mix of RN, licensed practical / vocational nurse (LPN/LVN), or aide hours. Separately, the rule required an RN on site 24 hours a day, 7 days a week, with limited hardship pathways.

Those numbers were never meant as a clinical optimum. CMS and later ASPE analysis treated them as a national baseline — a floor under the lowest-staffed homes — while the enhanced facility-assessment process still required staffing to resident acuity. A 2001 federal time-motion study that many advocates still cite pointed toward roughly 4.1 total care-staff HPRD and 0.75 RN HPRD for adequate care; LTCCC’s Q4 2024 summary shows only 28.3% of providers at or above that higher 4.1 bar. The 3.48 floor is the more policy-relevant benchmark for this article because it was the number CMS wrote into the rule.

Implementation was staggered: non-rural homes faced earlier deadlines for the total HPRD and 24/7 RN pieces; rural homes received longer runways for the RN and NA component floors. In December 2025, CMS repealed the numerical HPRD minimums and the 24/7 RN mandate as enforceable federal standards. That repeal does not erase the measurement question. PBJ still reports hours. Care Compare still publishes HPRD. Analysts, state surveyors, and families still need a shared reference line. This desk keeps CMS-3442-F’s published floors as that reference — a federal benchmark, not a claim that every home is currently cited against it.

The headline share below the total floor

Q4 2024 is the latest clean national PBJ print used here. National mean total nurse staff HPRD was 3.75, up a tick from 3.73 in Q3. The share of providers at or above 3.48 rose from 63.4% to 64.4%. Invert that and the below-floor share is 35.6%, down from 36.6% one quarter earlier. The quarterly path in the dashboard shows a slow grind: estimated below-floor shares near the low-40s in early 2023, then a gradual descent toward the mid-30s by late 2024. Progress is real and glacial.

ASPE’s May 2024 Care Compare brief — a slightly earlier vintage with N ≈ 14,413 after merges — found 59% of homes at or above 3.48, or 41% below. The Q4 2024 PBJ print is a few points better on the total floor. That improvement should not be over-read: definitions of which nurse categories enter “total nurse staff” can differ slightly across summaries, and ASPE’s brief also drops a few percent of homes for missing HPRD or census fields. Directionally, both vintages agree that a large minority of the certified stock still fails the total-nurse line.

MDS daily average census in Q4 2024 was about 1.23 million residents. A 35.6% facility share below the floor does not map one-for-one onto resident exposure — larger homes are more often below the total HPRD line — but it is enough to say that hundreds of thousands of residents live in buildings whose payroll hours still miss the published federal benchmark.

Nurse aides, not RNs, are the widest gap

Breaking the floor into components changes the story. ASPE’s May 2024 cut found:

The nurse-aide standard is the binding constraint for most of the industry. Seventy percent of homes sat below 2.45 NA HPRD in the ASPE snapshot, and the national mean NA HPRD (2.26) itself sits under the floor. RN hours look better at the mean (0.65 versus a 0.55 floor) even though half of homes still miss the RN line — again a distribution story. Total hours can clear 3.48 while the aide and RN pieces fail, or vice versa, because LPN/LVN hours fill the 0.48 residual and because homes mix skill mixes differently.

MedPAC’s March 2026 skilled nursing chapter adds a case-mix-adjusted lens: median facility risk-adjusted RN HPRD in 2024 was about 0.5, similar to recent years and below the unadjusted Care Compare mean. Freestanding and for-profit SNFs sat lower than hospital-based and nonprofit peers. That MedPAC median is not the same series as PBJ’s raw RN HPRD, but it reinforces that “average above the floor” is a weak comfort when the median facility is near or under the RN bar once acuity is held constant.

ASPE’s illustrative 100-bed example shows how close many below-floor homes already are. Among 100-bed facilities below the NA floor, mean NA HPRD was 1.97 — about 2 additional aides per 8-hour shift to reach 2.45. Among those below the total floor, mean total HPRD was 3.07 — about 1.7 nurse staff per shift. The gap is wide in headcount of homes and often narrow in hours for the typical below-floor building. That is why a slow quarterly climb in the at-or-above share is plausible without a sudden labor-market miracle.

Ownership and size bend the distribution

For-profit homes dominate the certified stock and dominate the below-floor counts. In ASPE’s May 2024 cut, 52% of for-profit facilities cleared the total 3.48 floor — meaning 48% did not — versus 80% of nonprofit facilities (20% below). The RN and NA gaps are similar in shape: 42% of for-profits at or above the RN floor versus 74% of nonprofits; 22% versus 51% on the aide floor. Those gaps are on the order of 25–30 percentage points. They line up with a long research literature linking ownership form to staffing intensity, and with MedPAC’s freestanding / for-profit RN medians.

Size works in the opposite direction from what a casual “big building, more staff” intuition suggests for HPRD. Homes with fewer than 50 certified beds were the most likely to clear each HPRD floor in ASPE’s brief (79% at or above total 3.48 among the smallest band in our size table). Larger bands show higher below-floor shares on total HPRD even as they more often clear a crude 24 RN hours per facility-day threshold — because absolute RN hours scale with census. Urban homes more often provide ≥24 RN hours per day (82% versus 68% rural); rural homes more often clear the NA and RN HPRD floors. Geography and ownership interact with local wage markets and Medicaid rate design; the dashboard’s region filter is a starting cut, not a causal model.

States: means above the floor, shares that still miss

State mean total nurse HPRD in Q4 2024 ranged from Alaska at 6.16 and Oregon at 5.04 down to Illinois at 3.27, Missouri at 3.30, and Texas at 3.33. Several large states sit with means near or under 3.48, which almost guarantees high below-floor shares. Desk estimates calibrated to those means put Illinois, Missouri, Texas, Georgia, and New Mexico among the highest below-floor states (roughly 50–58%), while Alaska, Oregon, North Dakota, and Hawaii sit in the single digits to low teens. These state below-floor shares are estimates, not a CMS-published table; treat them as ranked signals consistent with disclosed state means and the national 35.6% rate.

Contract staffing — agency and other non-employee nurse hours as a share of total nurse hours — varies just as sharply. Vermont’s contract share exceeded 26% in the Q4 2024 LTCCC summary; several Gulf and Southern states sat near 1–3%. High contract shares can prop up HPRD in tight labor markets while raising unit cost and continuity risk. The scatter view lets you toggle the Y-axis between estimated below-floor share and contract share against the same mean-HPRD X-axis.

Caveats, repeal, and how to read the floor

Several limits matter. First, PBJ hours are paid hours tied to job codes, not a direct observation of bedside minutes. Admin RN and director-of-nursing hours enter some “total nurse” summaries and are excluded from others; always check the footnote. Second, HPRD divides hours by MDS-derived census; census errors and short-stay churn move the ratio. Third, neither PBJ nor Care Compare timestamps RN presence across a 24-hour clock, so “≥24 RN hours per facility-day” is only an approximation of 24/7 coverage. Fourth, hardship exemptions and state staffing rules mean a home below 3.48 was not automatically out of compliance even while the federal numerical rule was in force. Fifth, the December 2025 repeal removed those numerical federal minimums; enhanced facility assessment remains, and states may keep or tighten their own ratios. Sixth, our state below-floor percentages and some quarterly path points before Q3 2024 are desk constructs — labeled as estimated or carried in the data module.

What survives the caveats is the distributional fact. In the latest full PBJ quarter, about one in three reporting certified nursing homes still posted total nurse hours below the 3.48 HPRD line CMS wrote as a national floor. Half missed the RN piece in the ASPE cut. Seven in ten missed the aide piece. For-profit ownership and larger bed size tilt toward the underside of the floor; small and nonprofit homes clear it more often. The mean can sit above 3.48 while thousands of buildings — and the residents inside them — do not. That is the PBJ story the floor was designed to illuminate, whether or not the floor remains enforceable tomorrow.