County-Level Nursing-Home RN Staffing Deserts

One state dominates the lowest end of the ranking. Louisiana parishes occupy a striking share of the lowest fifty counties, and they hold the very bottom: Bienville Parish reports just 0.156 registered-nurse hours per resident-day, roughly 28% of the federal floor and the equivalent of fewer than ten registered-nurse minutes per resident per day. It is followed closely by a tight cluster of south-Louisiana parishes — the Acadiana belt around Lafayette, Iberia, and St. Mary, and the industrial corridor through Calcasieu and the Baton Rouge region. These are not isolated weak facilities; they are whole parishes where the pooled registered- nurse coverage sits near a fifth of the standard.

The Louisiana pattern is instructive because it combines the forces that produce a staffing desert. Louisiana pairs one of the lower Medicaid nursing-home reimbursement environments in the country with a heavily for-profit ownership base and a thin rural registered-nurse labor pool, especially outside the New Orleans and Baton Rouge metros. Where Medicaid covers most nursing-home resident-days and the reimbursement rate is tight, facilities lean on lower-cost licensed practical nurses and certified nurse aides and minimize the more expensive registered-nurse line. The pooled county number is what that strategy looks like when it is summed across every facility and every day in a quarter.

Oklahoma, east Texas, and rural Missouri form the next band. The recurrence of the same states across dozens of adjacent counties is the signature of a regional, structural shortfall rather than a handful of troubled homes — and it is exactly the pattern a facility-by-facility lookup is least equipped to reveal.

The geographic story matters because nursing-home staffing is set by local labor markets, state Medicaid reimbursement, and regional ownership patterns — not by a single national dial. A registered nurse in a rural parish competes for the same scarce clinical labor as the local hospital, and in low-reimbursement states the nursing home usually loses that competition. When dozens of adjacent counties all sit below the floor, the result is a regional care desert: a resident cannot simply choose a better-staffed facility one town over, because the next town reports the same shortfall.

The market-size view sharpens the access story. Smaller markets — counties with only a handful of nursing homes — tend to carry a higher share of below-floor counties, which is exactly what the regional-desert hypothesis predicts: thin rural labor markets and lower Medicaid reimbursement push registered-nurse coverage down, and there are few alternative facilities to absorb the gap. Larger metropolitan counties are not immune, but they post a higher pooled RN-HPRD on average and more internal variation, because a single well-staffed academic-affiliated facility can lift the county pool even when others lag.

Excellent facility-level tools already exist. The Long Term Care Community Coalition (LTCCC) publishes quarterly PBJ staffing snapshots and per-facility tables, and NursingHome411 lets families look up a specific home and read its staffing, ownership, and citation history. Those are search interfaces: you arrive knowing a facility or a ZIP code, and they answer “how is this home doing?”

This study answers a different question — “where is the system failing, and how badly?” — and it answers it as a ranked, county-level analysis rather than a lookup. Three things make it distinct:

• Unit of analysis. The county, not the facility. By pooling every facility-day in a county and weighting by resident census, the study surfaces structural access deserts that a one-facility lookup cannot reveal — a county where every home is understaffed reads very differently from one weak facility among strong neighbors.
• Ranking against a statutory floor. Every county is expressed as a share of the 0.55 federal RN standard, producing a single national leaderboard rather than per-home report cards.
• Signed, reproducible provenance. The full ranked dataset ships as JSON and CSV, each row carrying a 14-field provenance record tying it to the CMS source, the snapshot quarter, the methodology version, and the chained integrity attestation. The aggregation SQL is published so any analyst can reproduce the table.

In short: LTCCC and NursingHome411 tell a family about a home; this study tells a journalist, regulator, or researcher about a region. The two are complementary — this analysis is the map that tells you which counties are worth a facility-level look.

For most of the program's history, federal law required only that a nursing home provide “sufficient” nursing staff and a registered nurse for eight consecutive hours a day, seven days a week — a standard so elastic that staffing varied enormously from state to state. The 2024 CMS Minimum Staffing Rule changed that for the first time, attaching a hard, hours-based number to the obligation: 0.55 RN-HPRD and 3.48 total nurse-hours per resident-day, alongside a 24/7 on-site registered-nurse requirement.

That numeric minimum did not survive. The federal staffing mandate was rescinded on December 3, 2025, effective February 2, 2026, before its phase-in took hold — which means there is now no in-force federal floor on how few registered-nurse hours a nursing home may staff. With the mandate gone, per-facility and per-county transparency becomes the primary remaining mechanism for staffing accountability. That is precisely the gap this study fills.

This study takes no position on whether the mandate should return. It uses the 0.55 figure as what it now is — a published federal benchmark, a fixed yardstick — and measures, county by county, how far reported staffing sits beneath it. A parish at 28% of that benchmark is not a tuning problem; reaching it would require multiplying registered- nurse hours several times over, against the same rural labor and reimbursement constraints that produced the shortfall. Whether or not a federal floor ever returns, the distance documented here is the same.

Journalists can lead with a defensible county ranking and drill into the named parishes and counties at the bottom, then cross-reference the deficiency and ownership records that facility-level tools surface. Researchers and policy analysts can download the full 1,362-county dataset, join it to Medicaid reimbursement, ownership, or rural-urban classifications, and reproduce the roll-up from the published SQL. Regulators and ombudsman programs can use the regional view to target oversight where shortfalls are systemic rather than isolated. And families comparing options in a low-ranked county gain the context that the problem may be regional — a signal to widen the search radius or weigh the staffing question more heavily.

Source. CMS Payroll-Based Journal (PBJ) Daily Nurse Staffing public-use file, calendar quarter 2025Q2. PBJ is a mandatory quarterly submission under §6106 of the Affordable Care Act; facilities report daily, payroll-tied staffing hours by job category along with daily resident census derived from the Minimum Data Set (MDS). The CMS file is U.S.-Government-Works (public domain). This snapshot comprises 1,322,867 facility-day records across 14,537 facilities in 52 states and territories.

Registered-nurse hours.Total RN hours sum three PBJ job codes: RN Director of Nursing (code 5), RN Administrative (code 6), and Direct-Care RN (code 7) — the same composition CMS uses for the reported RN staffing measure. The study also carries direct-care RN hours (code 7 only) separately, so readers can see how much of a county's registered-nurse coverage is bedside versus administrative.

Resident-days.The denominator is the sum of daily MDS census across the quarter. Hours-per-resident-day is therefore a census-weighted average: a facility's busy days count proportionally more than its quiet days, and large facilities influence the county pool in proportion to the residents they serve.

County roll-up.Each facility is mapped to its county via the PBJ state and county FIPS fields, joined to a state-FIPS table to form the 5-digit county code. County RN-HPRD = Σ RN hours ÷ Σ resident-days across all facilities in the county. This pooled construction is deliberately resistant to a single facility's reporting quirk.

Minimum-facility threshold. Counties with fewer than 3reporting facilities are excluded. Very small counties produce volatile ratios — one facility's payroll month can swing the whole county — so the threshold trades a small amount of coverage for materially more stable rankings. After the threshold, 1,362 counties qualify, covering 12,447 of 14,537 facilities.

Reproducibility. The exact aggregation is published as scripts/research/county-staffing-deserts.sql. The resulting snapshot is committed verbatim and served at the JSON and CSV endpoints below. Methodology version county-staffing-deserts/v1 is pinned; a future PBJ quarter will be published as a new dated snapshot rather than overwriting this one.