47 Mirror of the mind: How personal preferences shape perceptions of partners end-of-life wishes

<h3>Objectives</h3> Spouses frequently serve as surrogate decision-makers at the end of life, yet evidence suggests they frequently misinterpret each other’s preferences. This study examines the extent to which individuals’ own end-of-life preferences influence their predictions about their partner’s preferences and identifies key factors associated with prediction accuracy within older couples. <h3>Methods</h3> We analyzed data from 654 partnered adults aged 58 and older who participated in the Swiss sample of Wave 8 (2019/2020) of the Survey of Health, Ageing and Retirement in Europe (SHARE). Participants reported their own and their partner’s preferences across 14 end-of-life care domains. Prediction accuracy was defined as concordance between a respondent’s prediction and the partner’s self-reported preference. Using probit regression models, we assessed whether sharing the same preferences increased prediction accuracy, controlling for sex, age, education, language region, self-rated health, and other socio-demographic factors. <h3>Results</h3> The proportion of respondents who reported the same preferences as their partners ranged from 36.5% (for choosing where to die) to 90.4% (for life-prolonging treatment), highlighting substantial variation across different end-of-life and medical treatment preferences. Respondents were substantially more likely to correctly predict their partner’s preferences when the two shared the same views. This pattern was consistent across all 14 domains, with average marginal effects ranging from 0.38 to 0.61 (p &lt; 0.001). In contrast, socio-demographic characteristics showed no systematic association with prediction accuracy. <h3>Discussion</h3> Our findings reveal a strong projection bias in surrogate decision-making: individuals tend to rely on their own preferences to infer those of their partner. While shared preferences can enhance prediction accuracy, they may also obscure the absence of real dialogue. These results highlight the importance of advance care planning approaches that actively engage both members of a couple to facilitate mutual understanding and reduce the risk of misinterpretation.