Registered nurse staffing intensity, skill mix, and acute-care use among long-stay nursing home residents: a fixed-effects panel analysis.

backgroundNursing homes depend on sufficient staffing and an appropriate skill mix to recognize and manage acute clinical deterioration among long-stay residents. Although nurse staffing has been widely associated with resident outcomes, measures of total nursing hours may obscure clinically important differences in registered nurse (RN) intensity and staffing composition. Longitudinal evidence distinguishing RN skill mix from overall staffing volume, while examining within-facility changes in both hospitalization and outpatient emergency department (ED) use over an extended national panel, remains limited.

methodsWe constructed a nationwide quarterly panel of U.S. nursing homes from 2018Q4 through 2025Q2 using Centers for Medicare & Medicaid Services (CMS) Payroll-Based Journal (PBJ) and Care Compare data. The primary exposures were RN hours per resident day (HPRD) and RN share of total nursing HPRD. Total nursing HPRD was examined as a complementary measure of staffing volume, whereas licensed practical/vocational nurse and certified nursing assistant HPRD were secondary exposures. The primary outcome was the long-stay unplanned hospitalization rate, and the secondary outcome was the long-stay outpatient ED visit rate. Both outcomes reflected rolling 12-month measurement windows refreshed quarterly. We estimated nursing home and quarter fixed-effects models with facility-clustered robust standard errors. Sensitivity analyses used one-quarter-lagged staffing, replaced the Observed Score outcomes with the corresponding CMS "Adjusted Score" (risk-adjusted) fields, and excluded peak-pandemic quarters.

resultsThe sample comprised 14,993 nursing homes and 399,820 facility-quarter observations. Mean total and RN staffing were 3.80 and 0.65 HPRD, respectively. When total nursing HPRD and RN share were entered jointly, total HPRD was not independently associated with either outcome. Holding total HPRD constant, a 10-percentage-point higher RN share was associated with lower hospitalization (β = -0.800; 95% CI, - 1.368 to - 0.232) and outpatient ED visit rates (β = -0.560; 95% CI, - 0.952 to - 0.168). In type-specific models, a 0.1-HPRD increase in RN staffing was associated with lower hospitalization (β = -0.022; 95% CI, - 0.033 to - 0.011) and outpatient ED visit rates (β = -0.016; 95% CI, - 0.025 to - 0.007) per 1,000 resident-days. RN associations remained negative in the lagged, CMS-adjusted-score, and peak-pandemic-exclusion analyses.

conclusionWithin-facility increases in RN staffing intensity and a more RN-intensive skill mix were consistently associated with lower hospitalization and outpatient ED visit rates among long-stay nursing home residents. Although these observational findings do not support causal inference, they suggest that staffing composition should be considered alongside total nursing hours in quality improvement and workforce policy.