Family members' experience improves with care preference documentation in home based primary care.

Abstract

The Veterans Health Administration (VA) implemented the comprehensive life-sustaining treatment (LST) Decisions Initiative to provide training and standardize documentation of goals of care and LST preferences for seriously ill Veterans to improve end-of-life (EOL) outcomes. LST documentation is expected for all Home-Based Primary Care (HBPC) Veterans because they are at high risk of hospitalization and mortality. A retrospective, cross-sectional analysis compared associations between Bereaved Family Survey (BFS) EOL care ratings and LST documentation. Participants were Veterans who died August 1, 2018 through September 30, 2019 in one of 55 VA HBPC programs. Regression modeling generated odds for key BFS outcomes. LST template completion rate was plotted by month to understand the interaction between time, LST completion rate, and EOL care family ratings. LST preferences were documented for 39% of HBPC Veterans. Family members rated overall EOL care as excellent for 53% of Veterans but significant divergence in BFS ratings occurred during the last 7months of the study with 60% of family members of LST completers rating care as excellent compared with 48% for Veterans lacking LST documentation (p=0.003). The adjusted odds of rating overall care in the final month of life as excellent was higher among those with a completed LST template (1.64 95% CI 1.19, 2.26). Higher rates of LST documentation were associated with more favorable ratings of EOL but not in initial months following implementation of the comprehensive initiative; however, LST documentation rates were lower than expected among HBPC Veterans. Following an initial period of implementation of a comprehensive national initiative to promote Veteran choice about care during serious illness, documented LST preferences were associated with better family ratings of EOL care. HBPC clinicians may improve the bereaved family experience by using LSDTI tools and training to elicit and document preferences.