End-of-life care processes in nursing homes: Association with facility characteristics

We measured end-of-life (EOL) care processes in US nursing homes (NHs), and identified facility factors that characterize good EOL care. A random, national survey of NHs was conducted in 2012. Surveys were obtained from 1,201facilities and linked to Nursing Home Compare database to obtain facility and staffing information. Psychometric assessment of care processes included principal factor and internal consistency reliability analyses. We then fit four OLS regression models for each EOL care process. Psychometric analyses confirmed the dimensionality of 4 EOL process measures: symptom assessment; care delivery; communication-coordination among providers; communication with residents/families. These measures demonstrated good internal reliability (Cronbach's alpha 0.68-0.82) and substantial variation across NHs (mean: 2.67-3.77; SD: 0.56-0.66; where 1=worst and 5=best). Facilities with better EOL staff education had significantly higher score in all four care measures. Having a medical director was significantly associated with better assessment, but not with the other 3 measures. When NH medical director was also a geriatrician, performance on all EOL measures was significantly better. NHs with on-site palliative teams scored better on assessment and communication with residents/families, but NHs contracting for palliative care had worse scores in symptom assessment and communication/coordination among providers. Facilities with higher staff turnover rates had lower scores in assessment and communication/coordination among providers. Associations with facility ownership, religious affiliation, chain-membership, number of residents, nursing hours, and RN/CNA ratios were also examined. The survey provided valid and reliable measures for evaluating EOL care processes. Several modifiable facility-level characteristics were identified as significant predictors of EOL care processes.