Abstract
To assess whether national reductions in antipsychotic prescribing in U.S. nursing homes were shared equitably or masked widening gaps between top and bottom performers, we examined distributional trends across facilities from 2013 to 2023 using national facility-level data (CMS Provider Data Catalog; LTCFocus). We modeled long-stay antipsychotic measure performance using linear quantile mixed regression, assessed mobility across performance quartiles with transition matrices, and evaluated facility characteristics associated with trajectories. From 2013 to 2019, prescribing fell across all quantiles, with the steepest reductions among the worst performers, narrowing gaps with the best. After 2020, progress diverged: lower-performing facilities experienced sharp setbacks while top performers maintained gains. Mobility analyses showed entrenched disadvantage, with bottom-quartile facilities disproportionately likely to remain there. Higher Medicaid share predicted decline, whereas higher star ratings predicted improvement. The pandemic reversed earlier equity gains, amplifying structural disparities and exposing unequal capacity to sustain improvement.