Psychometric Properties of the Posttraumatic Stress Disorder Checklist with Veterans Residing in Hawai‘i

Abstract

Posttraumatic stress disorder (PTSD) is a prevalent mental health condition that has affected combat veterans across war eras. The PTSD Checklist (PCL) has been developed as a self-report questionnaire to assess the presence and severity of PTSD symptoms. There is currently a lack of research investigating the use of the PCL for the measurement of PTSD among veterans across diverse racial/ethnic backgrounds. The primary goal of this study was to investigate the sensitivity, specificity, predictive power, cut score, factor structure, internal consistency, and convergent validity of the PCL in a sample of veterans residing in Hawai‘i, including veterans of Native Hawaiian/Pacific Islander and Asian racial/ethnic descent. Participants were treatment-seeking male veterans (N = 282) who were referred to one of two randomized clinical trials of telemental health interventions for veterans with combat-related PTSD and completed the baseline assessment, during which the measures included in the data analyses for the current study were administered, to determine study eligibility. The PCL, Clinician-Administered PTSD Scale (CAPS), and additional self-report questionnaires that measured associated psychological symptoms were administered during pre-treatment baseline assessment sessions. Psychometric evidence was collected from the full sample of veterans and across the White, Native Hawaiian/Pacific Islander, and Asian racial/ethnic groups. A cutoff score of 47 best distinguished between veterans with and without PTSD (area under the receiver operating characteristic curve = 0.76, sensitivity = 0.89, specificity = 0.59, positive predictive power = 0.95, negative predictive power = 0.40). The 3-factor and 4-factor structure model for measuring symptoms of PTSD with the PCL both evidenced adequate model fit, with the 4-factor emotional numbing model demonstrating better model fit. The PCL items were found to have internal consistency of  = 0.92. The PCL subscales were significantly correlated with the corresponding scales of the CAPS and other measures of mental health symptomatology. Differences were found in the convergent validity of the PCL across racial/ethnic groups. Future research investigating the PTSD symptom structure in DSM-5 and psychometric properties of the PCL-5 are warranted. This study provides support for the cross-cultural reliability and validity of the PCL for the assessment of PTSD in veterans.