Abstract
OBJECTIVE: In clinical trials for posttraumatic stress disorder (PTSD), reliable and valid assessment is essential. However, accuracy can be compromised due to site-level variation, assessor unmasking, and participant burden associated with multiple visits to a study site. We therefore created a centralized assessment hub in which raters were trained to administer the Clinician-Administered PTSD Scale for Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (Weathers et al., 2013) remotely via telephone. METHOD: We developed a systematic training protocol for Veteran Affairs (VA) Cooperative Study 591, the largest multisite PTSD psychotherapy trial ever conducted (N = 916, Schnurr et al., 2022). We trained 15 raters, ranging from trainees to licensed clinical psychologists. Our protocol included three steps: (1) didactics, (2) scoring calibration, and (3) mock interviews. Step 3 was repeated until raters achieved 85% reliability with an expert. Randomly selected interviews (n = 200) were used to calculate interrater reliability. RESULTS: Reliability was outstanding for both PTSD diagnosis (κ = .90) and total severity score (intraclass correlation = .98). Assessment time period did not impact PTSD diagnostic agreement (χ² = 5.60; p = .23) but did affect total symptom severity (F = 4.43; p = .002). In contrast, rater educational attainment impacted diagnostic (χ² = 5.00; p = .025) but not total severity score (t = .85; p = .39) agreement. CONCLUSIONS: Our Clinician-Administered PTSD Scale for Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition training protocol was used successfully in the largest PTSD clinical trial to date. The use of a centralized assessment hub ensured highly accurate assessment that likely could not have been achieved using site-based raters across multiple sites. (PsycInfo Database Record (c) 2025 APA, all rights reserved).