The impact of caregiver’s role preference on decisional conflicts and psychiatric distresses in decision making to help caregiver’s disclosure of terminal disease status.

10106 Background: A decision aid (DA) increases knowledge, decreases decisional conflicts and regrets and improves post-decision satisfaction, emotional distress. However, few DA trials have revealed whether decisional role preferences have an impact on patient-reported outcomes by decision making. The objective of this study was to investigate the impact of caregiver’s decisional role preference on decisional conflicts and psychiatric distresses in decision making. Methods: 406 of 444 caregivers of terminally ill cancer patients enrolled onto a previous trial determining the efficacy of the decision aid about disclosure of terminal disease status were included in this analysis. The analysis outcomes were change score of decisional conflicts using the Decision Conflict Scale (DCS) and depression and anxiety using the Hospital Anxiety and Depression Scale (HADS) at 1 and 3 months from baseline. Participants were divided into 4 groups: active caregiver who received DA (active-DA), active caregiver in control group (active-control), passive caregiver who received DA (passive-DA), and passive caregiver in control group (passive-control). Linear mixed model was conducted to find out the impact of caregiver’s decisional role preference on the DCS and the HADS. Results: Among 406 caregivers, 137 (33.7%) showed active role preference, and 269 (66.3%) showed passive role preference. In post-hoc analysis of adjusted differences of change scores between passive-DA and active-DA groups, non-significant differences were observed in DCS. However, at 3 months, change scores of HADS depression subscale increased as 4.43 (95% confidence interval (CI), 0.78-8.07; P< 0.007; effect size (ES) 0.71) and those of HADS anxiety subscales increased as 4.14 (95% CI, 0.37-7.91; P= 0.021; ES 0.61) in passive-DA group than in active-DA group, showing moderate to large difference. Conclusions: These findings suggest that information about decision making might be provided with tailored format for how much individual wish to involve in decision making.