Patient preferences for medical care at the end of life and surrogate concordance: A post-hoc analysis of a prospective, randomized controlled trial

Background: For patients with advanced-stage, chronic illness, end-of-life care preferences are poorly understood. We sought to characterize patients’ end-of-life care preferences, and hypothesized that patients exposed to... [ view full abstract ]

Background: For patients with advanced-stage, chronic illness, end-of-life care preferences are poorly understood. We sought to characterize patients’ end-of-life care preferences, and hypothesized that patients exposed to an advance care planning decision aid would choose less aggressive versus more aggressive medical care. We also hypothesized that surrogate concordance would be lower when patients chose less aggressive care.

Methods: After randomization and completion of either a Standard advance directive form or the Making Your Wishes Known (MYWK) interactive decision aid program, patients with advanced-stage, chronic illnesses and their designated surrogates were asked to indicate whether they would ‘want’ or ‘not want’ a series of interventions by responding to questions about 6 clinical vignettes (n=28 questions). Answers were classified as desiring either more or less aggressive medical treatment. Descriptive statistics and Fisher’s exact test were used to compare patient preferences between the MYWK and standard ACP groups and to assess surrogate concordance based on whether patients chose more versus less aggressive care.

Results: There were 267 patient-caregiver dyads enrolled in the study and 265 were included because they completed all study questions. The average age of patients was 64 and 46% were female. Patients in the MYWK group were more likely (67%) to choose aggressive care compared to the Standard ACP group (61%) (p<0.0001). Surrogate concordance was significantly lower (52%) when patients chose less aggressive care compared to  more aggressive care (87%) (p<0.0001).

Conclusion: Surprisingly, patients who completed a comprehensive decision aid that provided information on end-of-life medical interventions, clarified values and goals, and assisted patients in documenting their end-of-life preferences was associated with patients choosing more aggressive care. These findings challenge the notion that ACP reduces patients’ preferences for aggressive end of life care and raise questions about how ACP can help surrogate decisions better align with patient preferences.