Efficacy of a Decision Aid Consisting of a Video and Booklet on Advance Care Planning for Advanced Cancer Patients: Randomized Controlled Trial

Context: Few randomized controlled trials of advance care planning (ACP) with a decision aid (DA) show an effect on patient preferences for end-of-life (EOL) care over time, especially in racial/ethnic settings outside the U.S. Objectives: The objective of this study was to examine the effect of a decision aid consisting of a video and an ACP booklet for EOL care preferences among patients with advanced cancer. Methods: Using a computer-generated sequence, we randomly assigned (1:1) patients with advanced cancer to a group that received a video and workbook that both discussed either ACP (intervention group) or cancer pain control (control group). At baseline, immediately after intervention, and at 7 weeks, we evaluated the subjects' preferences. The primary outcome was preference for EOL care (active treatment, life-prolonging treatment, or hospice care) on the assumption of a fatal disease diagnosis and the expectation of death 1) within 1 year, 2) within several months, and 3) within a few weeks. We used Bonferroni correction methods for multiple comparisons with an adjusted P level of 0.005. Results: From August 2017 to February 2018, we screened 287 eligible patients, of whom 204 were enrolled to the intervention (104 patients) or the control (100 patients). At postintervention, the intervention group showed a significant increase in preference for active treatment, life-prolonging treatment, and hospice care on the assumption of a fatal disease diagnosis and the expectation of death within 1 year (P < 0.005). Assuming a life expectancy of several months, the change in preferences was significant for active treatment and hospice care (P < 0.005) but not for life-prolonging treatment. The intervention group showed a significant increase in preference for active treatment, life-prolonging treatment, and hospice care on the assumption of a fatal disease diagnosis and the expectation of death within a few weeks (P < 0.005). From baseline to 7 weeks, the decrease in preference in the intervention group was not significant for active treatment, life-prolonging treatment, and hospice care in the intervention group in the subset expecting to die within 1 year, compared with the control group. Assuming a life expectancy of several months and a few weeks, the change in preferences was not significant for active treatment and for life-prolonging treatment but was significantly greater for hospice care in the intervention group (P < 0.005). Conclusion: ACP interventions that included a video and an accompanying book improved preferences for EOL care.