Patient care staffing levels and facility characteristics in U.S. hemodialysis facilities.

BACKGROUND: Higher numbers of registered nurses (RNs) per patient have been associated with improved patient outcomes in acute-care facilities. Variation in and associations of patient care staffing levels and hemodialysis facility characteristics have not been examined previously.

STUDY DESIGN: Cross-sectional study using Poisson regression to examine associations between patient care staffing levels and hemodialysis facility characteristics.

SETTING & PARTICIPANTS: 4,800 US hemodialysis facilities in the 2009 Centers for Medicare & Medicaid (CMS) End-Stage Renal Disease Annual Facility Survey (CMS-2744 form).

PREDICTORS: Facility characteristics, including profit status, freestanding status, chain affiliation, and geographic region, adjusted for facility size, capacity, functional type, and urbanicity.

OUTCOMES: Patient care staffing levels, including ratios of RNs, licensed practical nurses (LPNs), patient care technicians (PCTs), composite staff (RN + LPN + PCT), social workers, and dietitians to in-center hemodialysis patients.

RESULTS: After adjusting for background facility characteristics, ratios of RNs and LPNs to patients were 35% (P < 0.001) and 42% (P < 0.001) lower, respectively, but the PCT to patient ratio was 16% (P < 0.001) higher in for-profit than nonprofit facilities (rate ratios of 0.65 [95% CI, 0.63-0.68], 0.58 [95% CI, 0.51-0.65], and 1.16 [95% CI, 1.12-1.19], respectively). Regionally, compared to the Northeast, the adjusted RN to patient ratio was 14% (P < 0.001) lower in the Midwest, 25% (P < 0.001) lower in the South, and 18% (P < 0.001) lower in the West. Even after additional adjustments, the large for-profit chains had significantly lower RN and LPN to patient ratios than the largest nonprofit chain, but a significantly higher PCT to patient ratio. Overall composite staffing levels also were lower in for-profit and chain-affiliated facilities. The patterns hold when hospital-based units were excluded.

LIMITATIONS: Nursing hours were not available. Two part-time staff were counted as one full-time equivalent, which may not always be accurate.

CONCLUSIONS: The significant variation in patient care staffing levels and its associations with facility characteristics warrants inclusion in future large-scale hemodialysis outcomes studies. End-stage renal disease networks and hemodialysis facilities should attend to quality assurance and performance improvement initiatives that maximize licensed nurse staffing levels in hemodialysis facilities.