Validation of a Modified Group-Based Medical Mistrust Scale Among Young Latinx Adults in the United States.

ABSTRACT

Medical mistrust is an important barrier to accessing health care among Latinx populations in the United States (US). However, research on the validity and reliability of medical mistrust scales is limited. We examined the validity and reliability of a modified bilingual version of the Group-Based Medical Mistrust scale (mGBMMS) among a sample of Latinx adults. Participants included 308 Latinx adults (ages 18-25), who responded in Spanish (n = 134) or English (n = 174). Following feedback from bilingual/bicultural staff during the English-Spanish translation process, we made three changes to the original GBMMS. Validation testing of our 12-item mGBMMS scale included split-half and internal consistency reliability; discriminant, convergent, and predictive validity; and both exploratory and confirmatory factor analyses. The mGBMMS had good internal consistency (overall sample Cronbach's α = 0.79; Spanish Cronbach's α = 0.73; English Cronbach's α = 0.83). The mGBMMS showed good convergent (moderately correlated with the experiences of discrimination scale, r = 0.46, p < 0.001) and discriminant (weakly correlated with the acculturation scale, r = 0.11, p = 0.06) validity. Split-half reliability was 0.71 (p < 0.001). Exploratory and confirmatory factor analyses found a two-factor solution. The mGBMMS was associated with satisfaction with care (OR = 0.60, 95%CI 0.42-0.87), a sign of good predictive validity. Findings suggest that the mGBMMS is a valid and reliable scale to utilize among bilingual (Spanish/English) populations in the US. Further validation studies should be considered among Latinx respondents of different ages, backgrounds, languages, and US regions.