A psychometric examination of computerized adaptive measures of posttraumatic stress disorder among military veterans

Brenner et al. (2021) developed and initially tested two computerized adaptive measures of posttraumatic stress disorder (PTSD), one that provides a provisional diagnosis (PTSD computerized adaptive diagnostic screen [CAD-PTSD]) and one that estimates symptom severity (PTSD computerized adaptive severity test [CAT-PTSD]). We expanded on the initial psychometric findings by collecting data regarding test-retest reliability, incremental validity, and respondent burden. A sample of veterans (N = 156, 32% women, 36% Black, 13% identified as Spanish, Hispanic, Latino, Puerto Rican, or Cuban) recruited from three Veterans Affairs medical centers completed the CAT-PTSD and CAD-PTSD, the Clinician-Administered PTSD Scale for DSM-5, the PTSD Checklist for DSM-5 (PCL-5), Primary Care PTSD Screen for DSM-5 (PC-PTSD-5), and a battery of other self-rated scales. Fifty-three participants (34%) completed the measures a second time within 7 days (M_days = 5.41; SD = 1.97) of their first visit. CAT-PTSD scores revealed good convergent validity (r = .78 with Clinician-Administered PTSD Scale for DSM-5 total score), discriminant validity, and test-retest reliability (r = .80). Scores on the PCL-5 and PC-PTSD-5 had similar characteristics. The CAD-PTSD demonstrated poor diagnostic efficiency, κ(.5, 0) = .40, and test-retest reliability (κ = .25), whereas previously established cut-scores for the PCL-5 and PC-PTSD-5 showed fair to good diagnostic utility and adequate to good test-retest reliability. Results suggest that the CAT-PTSD may provide a valid indicator of PTSD symptom severity, but does not offer incremental value beyond the PCL-5 and PC-PTSD-5. The CAD-PTSD was markedly inferior to the use of PCL-5 or PC-PTSD-5 cut scores. (PsycInfo Database Record (c) 2026 APA, all rights reserved).