22147: Linking Neighborhood Social Deprivation to Nursing Home Quality: A Structure-Process-Outcome Model (session 25471)

Nursing home (NH) quality often varies across socio-economic gradients. The purpose of this study was to explore the relationship between social deprivation and NH quality. Integrating resource dependence theory with Donabedian's structure-process-outcome model, this study utilized 2019 data from Payroll Based Journal, Care Compare: Five-Star Quality Rating System, and LTCFocus.org. The analytic file included 12,900 free-standing NH observations. The primary variable of interest was severe social deprivation index (SDI) defined as SDI =85/100. Structural equation modeling was used to explore the effects of severe deprivation on NH structural (nursing staff), processes (antipsychotic medications (APMs), and outcomes (falls, pressure ulcers, activities of daily living (ADL) decline, and hospitalizations) quality indicators among long-stay NH residents. NHs in communities with severe SDI showed a decrease of 0.02 registered nurse (RN) hours per resident day (PRD) (p < 0.001) and an increase of 0.03 certified nursing assistant (CNA) hours PRD (p = 0.01). Higher licensed practical nurse (LPN) hours PRD (ß = -1.03; p = 0.003) and higher CNA hours PRD (ß = -0.80; p < 0.001) were significantly associated with reduced APM use. Increased APM use was significantly associated with ADL decline (ß = 0.03; p < 0.001) and hospitalizations (ß = 0.001; p = 0.013). Severe SDI indirectly resulted in higher hospitalizations (ß: 0.01; p: 0.01) and falls (ß: 0.01; p: 0.001) through lower RN staffing but lower pressure ulcers (ß: -0.07; p < 0.001) through higher CNA staffing. Directly, severe SDI increased pressure ulcers (ß: 0.69; p < 0.001) and hospitalizations (ß: 0.09; p < 0.001) and decreased falls (ß: -0.53; p < 0.001). Severe social deprivation affected nursing staff patterns, affecting care quality. These findings highlight the importance of considering environmental factors in the NH quality policymaking process.

view paper (if available)