Design/methods: Two-stage least squares regression models were performed to address
the endogenous relationships between RN staffing and the outcome-related quality
indicators, and ordinary least squares regression was used for the process-related ones.
This analysis focused on the relationship of RN staffing to nursing home quality indicators,
controlling for organizational characteristics, resources, resident casemix, and market
factors with clustering to control for geographical differences