End-of-Life Preferences: A Randomized Trial of Framing Comfort Care as Refusal of Treatment in the Context of COVID-19

Abstract

Background Rates of advance directive (AD) completion in the United Kingdom are lower than in the United States and other western European countries, which is especially concerning in light of the COVID-19 pandemic. UK residents typically complete an advance decision to refuse care (ADRT), whereas US versions of ADs present a more neutral choice between comfort-oriented or life-prolonging care. The purpose of this study is to test whether this framing affects decision making for end-of-life care and if this is affected by exposure to information about the COVID-19 pandemic. Methods In an online experiment, 801 UK-based respondents were randomly allocated to document their preferences for end-of-life care in a 2 (US AD or UK ADRT) by 2 (presence or absence of COVID-19 prime) between-subjects factorial design. Results Most (74.8%) of participants across all conditions chose comfort-oriented care. However, framing comfort care as a refusal of treatment made respondents significantly less likely to choose it (65.4% v. 84.1%, P < 0.001). This effect was exacerbated by priming participants to think about COVID-19: those completing an ADRT were significantly more likely to choose life-prolonging care when exposed to the COVID-19 prime (39.8% v. 29.6%, P = 0.032). Subgroup analyses revealed these effects differed by age, with older participants’ choices influenced more by COVID-19 while younger participants were more affected by the AD framing. Conclusions The UK ADRT significantly reduced the proportion of participants choosing comfort-oriented care, an effect that was heightened in the presence of information about COVID-19. This suggests the current way end-of-life care wishes are documented in the United Kingdom could affect people’s choices in a way that does not align with their preferences, especially in the context of the COVID-19 pandemic. Highlights Participants completing an AD framed as an advance decision to refuse treatment were significantly less likely to choose comfort-oriented care than participants completing an AD with a neutral choice between comfort-oriented and life-prolonging care. Exposure to a COVID-19 prime had an interactive effect on documented preferences in the refusal of treatment condition, with these participants even less likely to choose comfort-oriented care. Policy makers and organizations that design templates for advance care planning, particularly in the time of the COVID-19 pandemic, should be aware how the framing of these forms can influence decisions.