Staffing HPRD — the metric that predicts almost everything

If you can only look at one metric on a nursing home, look at staffing. Specifically: nursing hours per resident per day, broken out by role and by day-of-week.

Almost every other clinical and quality outcome correlates with staffing — falls, pressure ulcers, antipsychotic use, hospital readmission, family-reported satisfaction. The correlation is not subtle. Buildings in the top quartile of total nurse hours per resident per day post substantially better outcomes than buildings in the bottom quartile, even controlling for case-mix.

This is the metric the industry argues about most because it is the metric that costs the most.

What HPRD measures

Hours per resident per day (HPRD) is exactly what it sounds like: total paid nursing hours in a day, divided by the resident census that day, for a given role.

The roles that matter:

• RN HPRD — registered nurse hours. Most predictive of clinical outcomes. Usually the smallest number.
• LPN HPRD — licensed practical nurse hours. Larger than RN, often the workhorse role.
• CNA HPRD — certified nursing assistant hours. The largest number; the staff residents actually see most of the day.
• Total nurse HPRD — RN + LPN + CNA combined. The headline number.

Therapists (PT, OT, ST) are tracked separately and don't roll into the nursing HPRD.

A useful rough scale for total nurse HPRD on a national basis:

• Below 3.0 — bottom decile. Quality risk is real and showing.
• 3.0–3.5 — common. The middle of the distribution.
• 3.5–4.0 — strong. Top quartile.
• Above 4.0 — exceptional, usually in higher-acuity buildings or in markets with high private-pay mix.

CMS's mandatory minimum standard finalized in 2024 is 3.48 total nurse HPRD with at least 0.55 RN HPRD — a meaningful raise from the historical 0.75 RN minimum, with state implementation timelines staggered through 2029.

Why PBJ changed the conversation

Pre-2016, nursing home staffing was self-reported on a single form filled out at the time of survey. Buildings could (and did) staff up the week before a known survey window and back down afterward.

The Payroll-Based Journal (PBJ) system, mandated by ACA Section 6106 and operationalized in 2016, requires nursing homes to submit actual payroll-derived staffing data quarterly, broken out by role, by day, by shift. PBJ data is what feeds the Five-Star Staffing rating now.

Two implications:

1. The numbers are real. They come from payroll systems, with auditing by CMS.
2. The numbers are detailed. You can see day-by-day, weekday-vs-weekend, and role-by-role.

This second point is where the most useful analysis happens.

The weekend gap

CMS started publishing weekend staffing as a separate metric in 2022, and it has been one of the most informative additions to the dataset in years.

Most nursing homes staff lower on weekends. Some staff substantially lower. The weekend-vs-weekday gap is a strong proxy for operational discipline. Two buildings can have identical total HPRD averages while one runs 3.6 every day and the other runs 4.2 weekdays / 2.8 weekends — and the second is a meaningfully different operation.

CMS now incorporates weekend staffing into the Five-Star Staffing rating. A building can lose stars on weekend underperformance even if its weekday numbers are fine.

When evaluating a building, always check:

• The weekday HPRD.
• The weekend HPRD.
• The gap between them.
• The trend in both over the last 4–8 quarters.

The trend matters more than the level. A building running flat at 3.4 is in steady state. A building that ran 3.8 last year and is now 3.2 is on a trajectory. Trajectories continue.

Case-mix adjustment, briefly

Different residents need different levels of care. A skilled-nursing post-hospital rehabilitation unit averages much higher acuity than a stable long-term-care unit. A direct comparison of raw HPRD across these two would be misleading.

CMS publishes a case-mix adjusted HPRD that normalizes for the resident acuity profile (using the MDS-based RUG/PDPM groupings). For peer comparisons within a market, the case-mix adjusted number is the correct apples-to-apples figure.

For absolute thresholds (like the federal minimum), the raw number is what's measured.

The staffing/quality feedback loop

Here is the loop that operators and analysts both watch:

1. Staffing falls below a sustainable level.
2. Care misses start showing up — falls, missed turn schedules, missed med passes.
3. The next survey cycle catches some of these as F-tags (F689, F684, F758).
4. The F-tags hit the Health Inspection star rating with a lag.
5. Census softens because referrals slow and family confidence erodes.
6. Revenue drops, and staffing comes under further pressure.

The loop runs both directions. Operators who invest into staffing in advance of a survey cycle generally see the improvement reflected in stars 9–18 months later, with QM following last.

How ClearLTC presents staffing

Every facility profile shows:

• Current RN, LPN, CNA, and total nurse HPRD, both raw and case-mix adjusted.
• Weekday vs. weekend split.
• Trend across the last 8 quarters.
• Peer comparison against the county and state distribution.
• The Five-Star Staffing component breakdown.

Operators on the platform can see their own buildings benchmarked against any custom peer set — by county, by chain, by ownership type, or by resident profile.

See the staffing signals in your territory →