ASSESSMENT OF THE CROSS-CULTURAL VALIDITY OF AN HIV SYMPTOM DISTRESS MODULE IN AN INTERNATIONAL HIV CLINICAL TRIAL
Author(s)
Antoine Regnault, PhD, Associate Statistician1, Shalaka Marfatia, MS, MPH, Manager Global Outcomes Research2, Michael Louie, MD, MPH, Director2, Muriel Viala, MSc, Senior Project Manager11Mapi Values France, Lyon, France; 2 Pfizer Inc, New York, NY, USA
Objective: To assess whether the Symptom Distress Module (SDM) developed by the AIDS Clinical Trial Group showed cross-cultural validity in an international HIV clinical trial. Methods: The SDM was included in a Phase IIb/III trial to assess symptoms perceived by HIV-1-infected treatment-naïve patients. The cross-cultural validity of 25 language versions of the SDM was assessed using baseline data of 759 patients from 3 treatment arms, each having received zidovudine and lamivudine in addition to one of the following: maraviroc 300mg QD, maraviroc 300mg BID (approved dose for treatment-experienced patients) and efavirenz 600mg QD. Given the large number of versions, creating relatively homogeneous cultural groups was necessary for analysis. Seven cultural groups were defined according to language and geographical considerations: European Germanic, Polish, European Romance, Occidental English, American Spanish, Bantu and African Indo-European. The cross-cultural validity of the SDM was assessed by applying ordinal logistic regressions to detect Differential Item Functioning (DIF), and using the STATIS approach, which explores distances between item correlation matrices. Results: Most items did not function differentially between cultural groups: only four symptoms showed DIF (fatigue, fevers, feeling anxious and headache) and the greatest cultural differences were observed for fatigue. The African Indo-European versions of the ASDM presented the highest number of differences from the other versions. With the STATIS approach, the Bantu and European Germanic groups were the furthest from the Occidental English but no clear meaningful difference was found in the expressed symptom pattern across cultural groups. Conclusion: Considering the substantial heterogeneity of cultures included in the study, these statistical findings, together with the rigorous methodology applied for the linguistic validation of the questionnaire, support the cross-cultural validity of the SDM. These findings also indicate that culture has a limited impact on the symptoms expressed by HIV-1-positive individuals starting antiretroviral therapy.
Conference/Value in Health Info
2008-05, ISPOR 2008, Toronto, Ontario, Canada
Value in Health, Vol. 11, No. 3 (May/June 2008)
Code
PIN40
Topic
Methodological & Statistical Research
Topic Subcategory
PRO & Related Methods
Disease
Infectious Disease (non-vaccine)