Psychometric properties of the fifth version of the Posttraumatic-Stress-Checklist in a clinical sample of refugees

Abstract

Background The validity and reliability of the fifth version of the Posttraumatic Stress Checklist (PCL-5) have been demonstrated in numerous studies in Western cultures, various languages, and increasingly in non-Western cultures. However, its applicability in a heterogeneous sample of refugees is controversial because the PCL-5 is based on the Western understanding of post- traumatic stress disorder (PTSD). This study aimed to measure the psychometric properties of the PCL-5 in a sample of refugees and to assess the suitability of this questionnaire for screening refugees living in a Western country for PTSD. Method An existing data set of 1202 refugees was analysed utilising Cronbach’s alpha, inter-item and inter-subscale correlations, convergent and discriminant validity, and Receiver Operating Characteristic (ROC) analysis. For validity assessment, the PCL-5 was compared to the Clinician-Administered PTSD Scale for Diagnostic and Statistical Manual of Mental Disorders (CAPS-5), and the Brief Symptom Inventory (BSI). Results Clusters generally measured the underlying constructs of PTSD well, however, the expected pattern between subscales was not confirmed, and a ceiling effect was observed. ROC analysis yielded a cut-off score of 44 and inconsistencies were found concerning convergent and discriminant validities. The PCL-5 does not appear to discriminately capture the severity of trauma experienced by refugees. It might be that the previously assumed cut-off score is too low for our treatment-seeking sample as the PCL-5 does not adequately capture the sequential, prolonged, and ongoing traumatisation of refugees. This should be considered when developing and deciding on a screening instrument to be used for refugees. Keywords: PCL-5, PTSD, DSM-5, assessment, refugees, psychometric properties, trauma