Prevalence and Predictors of Medical Mistrust Among Socioeconomically and Racially Diverse Cancer Patients in Philadelphia

Background/Objectives: Medical mistrust (MM) is associated with adverse health outcomes, but few studies have assessed MM in cancer patients. MM is frequently measured using the Medical Mistrust Inventory (MMI), measuring institutional MM (e.g., government), or the Group-Based Medical Mistrust Scale (GBMMS), measuring race-based MM. We sought to assess the prevalence of MM among cancer patients diverse by age, sex, race/ethnicity, and socioeconomic status (SES), recruited from an urban safety net hospital and a suburban comprehensive cancer center. Methods: Patients completed a one-time survey. The primary outcome was MM as measured by the GBMMS and MMI tools. Covariates included demographics, treatment campus (urban vs. suburban), and psychosocial measures relevant to MM. Results: Purposeful sampling recruitment resulted in 200 participants (survey completion: 74.6%). The median age was 60 years, with 62% female, 45% African-American, 15% Hispanic, 47.5% education <= HS diploma, and 51.5% income <= USD 50,000/yr. Elevated MMI and GBMMS scores (moderate-to-high) were seen, respectively, in Hispanic (20.7% and 33.4%) and African-American (AA) patients (31.8% and 48.9%), compared with White patients (14.3% and 9.9%). The MMI and GBMMS tools captured complimentary aspects of MM in cancer patients (Spearman's 0.531, p < 0.0001). MMI was associated with lower education (0.034) and race (p = 0.04), while GBMMS was strongly associated with race (p < 0.001), urban campus (p = 0.035), and mistrust of government/health organization information (both p < 0.05). Higher MMI/GBMMS scores were both associated with research mistrust and mistrust of information from physicians. Conclusions: Institutional and race-based MM are prevalent among cancer patients diverse by age, sex, race/ethnicity, and SES. Lower education was associated with institutional MM but not race-based MM.