Similarity Between Romantic Partners' End-of-Life Treatment Preferences: When Is Surrogate Projection Acceptable?

Purpose: When terminally-ill people are unable to communicate their wishes for medical treatment, surrogates often make such decisions. It is critical for researchers to identify whether couples know each other’s end-of-life treatment preferences because if a dying person is married, his or her spouse is the first priority in most states to be selected as a surrogate. Previous research using hypothetical scenarios finds that the surrogate’s decision does not match the dying person’s preference much of the time. Past research and theory suggests that people assume that others hold preferences similar to their own, and that people project these appraisals inaccurately onto others. Projection, however, is an acceptable strategy as long as couples truly share similar values. The purpose of this study is to learn how similar couples’ end-of-life care preferences are.

Methods: The study conducted dyadic analyses using a nationally representative sample of 1,075 heterosexual American couples aged 18 to 64. Participants were asked to rate the level of treatment that they would like to receive during a serious illness with very little chance of survival (0 = stop all life-prolonging treatment and 10 = continue all treatment). The outcome variable was the difference in treatment level preferences between partners.  Independent variables were marital status, household income, and homogamy on couple characteristics, including age, health status, race, education, and perceived relationship quality. Ordinary least squares regression was used to predict couples’ similarity in the level of treatment preferences.

Results: The average difference between partners’ preference levels was 3.05 (SD= 3.02). Roughly a quarter of the couples reported identical treatment preferences. We found that couples with an age gap of more than three years (B=0.59, p<0.01) and couples with different perceived health status (B=0.66, p<0.05) presented greater dissimilarity in treatment preferences. Cohabiting couples’ preferences were more similar than those of married couples (B=-0.83, p<0.01).

Discussion/Implication: The results suggest that couple characteristics are associated with similarity in their end-of-life treatment preferences. When making decisions as surrogates, it is ideal for romantic partners to have exact knowledge about what their partner would want for end-of-life care. However, if they do not have such knowledge, people often assume and project their own values onto others. Thus, better surrogates would be those who share similar values with their partners. This study found that when couples’ ages are closer and when their health statuses are alike, their treatment preferences are more similar. In addition, the findings show that cohabiting couples share more similar preferences, which indicates their projections may be more accurate than those of married couples. In many states, a married person is automatically appointed as a surrogate if his or her dying spouse does not have one, but cohabiting couples are not given the same legal standing as married couples. Therefore, the study findings offer important implications for legal and medical policies regarding romantic partners’ end-of-life decision making for their loved-ones.