Pandemic Effects on Stability of End-of-Life Preferences and Patient-Surrogate Dyad Congruence

Whether a largescale disaster alters people's previous decisions about their end-of-life care is unknown.

We examined the effects of a disaster, the COVID-19 pandemic, on stability of end-of-life care preferences among dialysis patients and on patient-surrogate goals-of-care congruence.

We used a natural experimental design to examine goals-of-care preferences pre- and postexposure to the pandemic during a pragmatic trial testing SPIRIT (sharing patient's illness representations to increase trust), an evidence-based advance care planning (ACP) intervention. There were 151 patient-surrogate dyads who prior to the pandemic lockdown had completed baseline (T1) and postintervention assessments (T2) regarding their goals-of-care preferences in two end-of-life scenarios. Of those 151 dyads, 59 intervention, and 51 usual care dyads consented to be in the present study and completed the goals-of-care tool two additional times, at enrollment (T3) and six months later (T4), along with the COVID stress scale (CSS). Dyad congruence was ascertained by comparing patient and surrogate responses to the goals-of-care tool.

There were no changes over time in the proportions of patients who chose comfort-care-only in the goals-of-care tool. The proportion of patients who chose comfort-care-only and dyad congruence were higher in SPIRIT compared to usual care, but there was no interaction between that treatment effect and exposure to the pandemic. CSS was associated with neither patients' preferences nor dyad congruence.

The pandemic alone did not appear to influence patients' goals-of-care preferences or dyad congruence. This finding supports the stability of value-based end-of-life preferences in general, even during a disaster.