Assessing the factorial structure and measurement invariance of PTSD by gender and ethnic groups in Sri Lanka: An analysis of the modified Harvard Trauma Questionnaire (HTQ)

Abstract

The Harvard Trauma Questionnaire (HTQ) remains the most widely used screening measure for post-traumatic stress disorder (PTSD) in the refugee and post-conflict field. The present study is the first to test the factorial structure and measurement invariance of the HTQ according to DSM-5 criteria across gender and ethnic groups in the ethnically diverse society of post-conflict Sri Lanka. The survey sample included 5136 participants (86% response rate) followed up 12 months after a baseline nationally representative survey undertaken in Sri Lanka in 2014. Exposure to conflict-related traumatic experiences (TEs) generating a trauma count (TC), and symptoms of PTSD were assessed using a modified version of the HTQ adapted to the local context. The final analytic sample included 4260 participants after excluding records with missing data on key variables. We conducted Multigroup Confirmatory Factor Analysis (MG-CFA) to test the four-factor (DSM-5 consistent) and three-factor (DSM-IV-TR) models of PTSD, then assessing measurement invariance of the four factor model by gender and ethnic groups. The three-factor and four-factor DSM-5 model each produced a good fit across the sample as a whole. In addition, there was configural, metric, and scalar invariance for the four-factor model both by gender and ethnicity. The trauma count was directly associated with each of the symptom domains of the four factor model. Our findings provide support for the capacity of the modified HTQ to measure the DSM5 construct of PTSD across gender and key ethnic groupings in Sri Lanka. Confirmation of our findings in other cultures will be important.