Predictors of Medical Mistrust Among Surrogate Decision-Makers of Patients in the ICU at High Risk of Death: A Pilot Study.

Background: Medical mistrust may worsen communication between ICU surrogate decision-makers and intensivists. The prevalence of and risk factors for medical mistrust among surrogate decision-makers are not known.

Research Question: What are the potential sociodemographic risk factors for high medical mistrust among surrogate decision-makers of critically ill patients at high risk of death?

Study Design And Methods: In this pilot cross-sectional study conducted at a single academic medical center between August 2022 and August 2023, adult patients admitted to the medical ICU and their surrogate decision-makers were enrolled. All patients were incapacitated at enrollment with Sequential Organ Failure Assessment scores of ≥ 7 or required mechanical ventilation with vasopressor infusion. Surrogate decision-maker sociodemographic characteristics were age, race, sex, education, relationship to the patient, employment, prior exposure to a loved one transitioning to hospice or comfort-focused care, and religiousness. The primary outcome was surrogate decision-maker medical mistrust, measured using the Medical Mistrust Multiformat Scale. Multiple linear regression was used to determine sociodemographic characteristics associated with higher medical mistrust.

Results: Thirty-one patients and their surrogate decision-makers were enrolled during the study period, surpassing our goal of 30 pairs and indicating recruitment feasibility. Mean ± SD surrogate age was 53.8 ± 14.5 years, 24 surrogates were female, and mean medical mistrust score was 17.1 ± 5.4. Race was associated with medical mistrust, with Black participants showing higher medical mistrust compared with White participants (β =10.21; 95% CI, 3.40-17.02; = .010). Religiousness was associated with lower medical mistrust (β = -2.94; 95% CI, -4.43 to -1.41; = .003). Prior exposure to hospice or comfort-focused care was associated with higher medical mistrust (β = 7.06; 95% CI, 1.21-12.91; = .025).

Interpretation: We found that recruiting ICU surrogates and measuring medical mistrust within 48 h of ICU admission was feasible. Several surrogate sociodemographic characteristics were associated with changes in medical mistrust. These preliminary findings will inform the design of future studies.