Health Professional vs Layperson Values and Preferences on Scarce Resource Allocation

Abstract

Importance: COVID-19 prompted rapid development of scarce resource allocation (SRA) policies to be implemented if demand eclipsed health systems' ability to provide critical care. While SRA policies follow general ethical frameworks, understanding priorities of those affected by policies and/or tasked with implementing them is critical.

Objective: To evaluate whether community members and health care profesionals (HCP) agree with SRA protocols at the University of California (UC).

Design, setting, and participants: This survey study used social media and community-partnered engagement to recruit participants to a web-based survey open to all participants aged older than 18 years who wished to enroll. This study was fielded between May and September 2020 and queried participants' values and preferences on draft SRA policy tenets. Participants were also encouraged to forward the survey to their networks for snowball sampling. Data were analyzed from July 2020 to January 2024.

Main outcomes and measures: Survey items assessed values and preferences, graded on Likert scales. Agreement was tabulated as difference in Likert points between expressed opinion and policy tenets. Descriptive statistics were tested for significance by HCP status. Free text responses were analyzed using applied rapid qualitative analysis.

Results: A total of 1545 participants aged older than 18 years (mean [SD] age 49 [16] years; 1149 female participants [74%], 478 health care practitioners [30%]) provided data on SRA values and preferences. Agreement with UC SRA policy as drafted was moderately high among respondents, ranging from 67% to 83% across domains. Higher agreement with the interim policy was observed for laypersons across all domains except health-related factors. HCPs agreed more strongly on average that resources should not be allocated to those less likely to survive (HCP mean, 3.70; 95% CI, 3.16-3.59; vs layperson mean, 3.38; 95% CI, 3.17-3.59; P = .002), and were more in favor of reallocating life support from patients less likely to those more likely to survive (HCP mean, 6.41; 95% CI, 6.15-6.67; vs layperson mean, 5.40; 95% CI, 5.23-5.58; P < .001). Transparency and trust building themes were common in free text responses and highly rated on scaled items.

Conclusions and relevance: This survey of SRA policy values found moderate agreement with fundamental principles of such policies. Engagement with communities affected by SRA policy should continue in iterative refinement in preparation for future crises.

Figure 1.. Responses to Scarce Resource Allocation (SRA) Questions by Health Care Worker Status

Plots demonstrate distribution of responses. The blue strips represent self-identified health care workers while the orange strips represent laypersons, with the triangles between strips showing the mean response for each question for respondent type. ^aIndicates P < .05. ^bIndicates P < .001. ^cIndicates P < .01.

Figure 2.. Proportion of Respondents Expressing Agreement With University of California Scarce Resource Allocation Policy Tenets as Written by Domain

For each domain, participants were included if they responded to at least 2 items from the domain. For overall agreement, participants were included if they responded to at least 1 item from each domain. HCP indicates health care personnel. ^aIndicates P < .001 between groups.

Figure 3.. Differences in Agreement With Exemptions to Scarce Resource Allocation Policy by Respondent Characteristics, Stratified by Health Care Professional Status

Fractional model additionally controlling for age, marital status, employment, education, households with children, and essential worker status.