Measuring End-of-Life Care Processes in Nursing Homes

Today, one in three older Americans die in nursing homes (NHs), where end-of-life (EOL) quality of care is woefully inadequate. Our objectives are to: 1) test the reliability and validity of a survey instrument measuring EOL care processes in NHs, and 2) examine factors associated with better EOL care.

A survey instrument, using Likert-scale items, was developed to measure EOL care processes (assessment, delivery, communication among providers, communication with residents/families) and organizational structure in NHs. Also included were items such as turnover, staffing, and facility characteristics. 608 surveys were mailed to all facilities in New York State; 313 were completed (52.4%).

Cronbach's alphas, Pearson's correlations, and regression models were used to assess the reliability and validity of the instrument. Multivariate regressions, with random effects, were used to examine predictors of better EOL processes.

Cronbach's alphas, ranging from 0.63 to 0.82 demonstrate moderate-to-high reliability for process domains. Regression models demonstrate statistically significant relationships between process domains and structural variables. Presence of quality assurance mechanisms and EOL in-service education for nursing staff, are consistently and significantly (p<0.05) associated with better care processes. Management turnover is negatively and significantly associated with assessment (p=0.0001), delivery (p=0.041), and communication with residents/families (p=0.017). Facilities with religious affiliation have higher scores for communication with residents/families (p=0.042).

We developed and tested the reliability and validity of an instrument to measure EOL care processes in NHs. Our findings provide insights for NH leaders striving to improve EOL care processes and ultimately the quality of care for dying residents.