Decision-Making by Medical Surrogates for End-of-Life Patients

This figure is a grouped vertical bar chart titled 'Patient Preferences and Surrogate Decision Making about Elderly Patients'. The y-axis, labeled 'Share of respondents indicating the surrogate should choose life-extending care for the hypothetical patient', ranges from 0% to 100% in increments of 20. The x-axis represents two patient types: 'Physical Impairment' and 'Dementia'. For each type, there are 3 preference scenarios: 'Patient preferences unknown', 'Patient prefers comfort care', and 'Patient prefers life-extending care'. For physically impaired patients: 72% support life-extending care when preferences are unknown. 23% when comfort care is preferred. 92% when life-extending care is preferred. For patients with dementia: 24% support life-extending care when preferences are unknown. 11% when comfort care is preferred. 68% when life-extending care is preferred. The note on the figure reads, The survey asked if a spousal surrogate should choose life-extending care for a hypothetical incapacitated 85-year-old. Patient's preferences are potentially expressed in an advance directive. The source line reads, Source: Researchers' calculations using data from their experimental survey.


As the population ages, the need for surrogate decision-makers for patients near the end of their lives is rising. When hospitalized older adults are unable to actively participate in decisions about their care, surrogates must make choices, often with limited information. Advance care planning with written directives may improve surrogate decision-making, but directives have limitations: preferences may change after completion, directions may not apply to the ultimate situation, and there can be communication challenges between the surrogates and care teams.

In How Do Surrogates Make Treatment Decisions for Patients with Dementia? An Experimental Survey Study (NBER Working Paper 32116), researchers Lauren Hersch NicholasKenneth M. LangaScott D. Halpern, and Mario Macis examined surrogates’ decisions involving elderly patients with and without dementia. They tested whether choices are influenced by the patient's advance directive and by the surrogate's preferences. 

Study participants were more likely to recommend that a surrogate override a dementia patient’s preference for life-extending treatment in favor of comfort care but override a non-dementia patient’s preference for comfort care in favor of life-extending treatment.

Their experimental survey asked 1,466 respondents if a spousal surrogate should choose life-extending treatment or comfort care for an incapacitated 85-year-old person hospitalized due to a severe illness such as a heart attack. The vignettes randomized the patient’s dementia status, advance planning, and gender. All hypothetical patients had underlying health impairments — either physical or cognitive — that would remain after life-extending care.

Respondents were more likely to recommend life-extending care for patients with physical impairment but not cognitive impairment (65 percent) than for patients with dementia but not physical impairment (32 percent). Recommendations for life extension for patients with both physical and cognitive impairments (35 percent) were similar to those for patients with dementia only. The researchers did not find any gender difference in recommended treatments. Respondents were less likely to suggest that surrogates follow advance directives that indicate a preference for life-extending care for patients with dementia only (68 percent) than for patients with physical impairment only (92 percent). When the advance directive specified a preference for comfort care, 89 percent recommended following this preference when the patient had dementia, compared to 77 percent for non-dementia patients. With no directive indicating the patient’s preferred treatment, over 70 percent of respondents recommended life-extending care for the physically impaired but only one-quarter did so for the dementia patients.

Personal experience influenced surrogates: those previously exposed to individuals with dementia were less likely to recommend following life extension wishes for dementia patients (47 percent) than surrogates without exposure (74 percent). Respondents’ own preferences also played a major role. For patients with no documented preference, respondents were more likely to recommend treatment that aligned with their own preferences. For dementia patients, life-extending care was recommended by 11 percent of those preferring comfort care for themselves and 51 percent of those preferring life extension despite disability. For non-dementia patients, the respective percentages recommending life-extension based on respondents' own preferences were 49 percent and 92 percent.

The results suggest that older adults choosing surrogates whose preferences align with their own increases the likelihood of preference-concordant end-of-life care. Ongoing communication, as preferences evolve over time or in response to changes in the patient's health, can also help surrogates make decisions that are consistent with the patient's true preferences.

— Leonardo Vasquez

The researchers acknowledge financial support from the National Institute on Aging under grant R01AG059205.