Posttraumatic stress disorder in diverse populations: Testing for assessment bias in a nationally representative sample.

Abstract

A growing body of research has emerged to characterize differences in posttraumatic stress disorder (PTSD) symptom presentations in individuals from diverse racial and ethnic groups. However, less research has examined if these observed differences can be attributed to bias within PTSD assessments. Knowledge about potential bias in PTSD assessment is essential for interpreting group differences. If PTSD assessments do not perform similarly across diverse demographic groups, then observed differences may be artificial products of inaccurate measurement, new assessments could be required for individuals from different demographic groups, and we would be unable to accurately detect PTSD treatment effects in patients from diverse groups. We evaluated PTSD assessment bias through tests of measurement invariance for the semistructured, clinician-administered AUDADIS-5 diagnostic assessment of participants in the National Epidemiologic Survey on Alcohol and Related Conditions-III. Participants included those who reported having experienced at least one potentially traumatic event in their lifetime (N = 23,936). Measurement invariance was assessed for participants who identified from several demographic groups (Asian, Native Hawaiian, or Pacific Islander; Hispanic; American Indian/Alaskan Native; and Black) compared to participants who identified as White (non-Hispanic). Overall, PTSD assessment was largely invariant across groups, while small amounts of measurement invariance were detected that can inform future research and clinical adaptations. This work validates prior research that relies on a common conceptualization of PTSD, and it provides several paths for future improvement in research and clinical practice. (PsycInfo Database Record (c) 2023 APA, all rights reserved).