Masao Ichikawa1, Shinji Nakahara, Susumu Wakai. 1. Dept. of Community Health, School of International Health Graduate School of Medicine, University of Tokyo, 7-3-1 Hongo, Bukyo-ku Tokyo, Japan. masao@m.u-tokyo.ac.jp
Abstract
BACKGROUND: Cross-cultural use of the cutoff points determined in Indo-Chinese refugees of the Hopkins Symptom Checklist-25 (HSCL-25) and the Harvard Trauma Questionnaire (HTQ) is common in refugee mental health research but it might have caused misclassifications. METHODS: We assessed the validity of the predetermined cutoff points in identifying depression and post-traumatic stress disorder (PTSD) cases among Afghan refugees, with the algorithm method as a gold standard. We estimated the prevalence of depression and PTSD, using the HSCL-25 and the HTQ with the cutoff points and the algorithm method, and calculated sensitivity, specificity, positive and negative predictive values of the cutoff points. RESULTS: Nine in ten PTSD cases/non-cases were correctly identified as cases/non-cases, and nine in ten respondents classified as PTSD cases/non-cases were actually cases/non-cases. In contrast, only one in three non-depression cases was correctly identified as a non-case, and three in five respondents who were classified as depression cases was actually a case. The prevalence of depression was estimated to be 86% with the cutoff points and 53% with the algorithm method, and that of PTSD 46% and 44%, respectively. CONCLUSIONS: The cutoff point approach might have low specificity and positive predictive value particularly for depression in our study. Cross-cultural use of the scale cutoff points determined in one cultural group to another needs to be reconsidered.
BACKGROUND: Cross-cultural use of the cutoff points determined in Indo-Chinese refugees of the Hopkins Symptom Checklist-25 (HSCL-25) and the Harvard Trauma Questionnaire (HTQ) is common in refugee mental health research but it might have caused misclassifications. METHODS: We assessed the validity of the predetermined cutoff points in identifying depression and post-traumatic stress disorder (PTSD) cases among Afghan refugees, with the algorithm method as a gold standard. We estimated the prevalence of depression and PTSD, using the HSCL-25 and the HTQ with the cutoff points and the algorithm method, and calculated sensitivity, specificity, positive and negative predictive values of the cutoff points. RESULTS: Nine in ten PTSD cases/non-cases were correctly identified as cases/non-cases, and nine in ten respondents classified as PTSD cases/non-cases were actually cases/non-cases. In contrast, only one in three non-depression cases was correctly identified as a non-case, and three in five respondents who were classified as depression cases was actually a case. The prevalence of depression was estimated to be 86% with the cutoff points and 53% with the algorithm method, and that of PTSD 46% and 44%, respectively. CONCLUSIONS: The cutoff point approach might have low specificity and positive predictive value particularly for depression in our study. Cross-cultural use of the scale cutoff points determined in one cultural group to another needs to be reconsidered.
Authors: Michael Hollifield; Teddy D Warner; Nityamo Lian; Barry Krakow; Janis H Jenkins; James Kesler; Jayne Stevenson; Joseph Westermeyer Journal: JAMA Date: 2002-08-07 Impact factor: 56.272
Authors: Eboni M Taylor; Emad A Yanni; Clelia Pezzi; Michael Guterbock; Erin Rothney; Elizabeth Harton; Jessica Montour; Collin Elias; Heather Burke Journal: J Immigr Minor Health Date: 2014-12
Authors: Susan M Meffert; Karen Musalo; Akram Osman Abdo; Omayma Ahmed Abd Alla; Yasir Omer Mustafa Elmakki; Afrah Abdelrahim Omer; Sahar Yousif; Thomas J Metzler; Charles R Marmar Journal: Med Confl Surviv Date: 2010 Apr-Jun
Authors: Dabney Evans; David C Buxton; Andrey Borisov; Amita K Manatunga; Dawa Ngodup; Charles L Raison Journal: Soc Psychiatry Psychiatr Epidemiol Date: 2008-04-09 Impact factor: 4.328
Authors: Bayard Roberts; John Browne; Kaducu Felix Ocaka; Thomas Oyok; Egbert Sondorp Journal: Health Qual Life Outcomes Date: 2008-12-02 Impact factor: 3.186