Article Text
Abstract
Objectives HPV associated lesions heavily influence the patients’ psychological health. Merck and coll. developed the HPV Impact profile (HIP) questionnaire to quantify this impact. Previously, while translating this questionnaire to Arabic and validating it in the Lebanese population, we had encountered several issues with its interpretation scheme. This article aims to study the psychometric properties of other proposed schemes found in the literature in order to choose the most adapted one for the Lebanese population.
Methods The Arabic versions of the HIP and HADS questionnaires were administered to 118 Lebanese women presenting for an HPV related consultation. The psychometric properties of the initial domains were studied before and after reverse scoring 8 items carrying a positive connotation (‘adapted domains’) and compared to those of two other item distributions created by Santos et al.
Results Most of the initial domains presented weak alpha Cronbach coefficients and internal consistency. Reverse scoring 8 items considerably improved the coefficients of 6 of the 7 domains. Both of Santos et al.’s distributions had good coefficients. Nevertheless, by modifying and combining these domains, the ‘Lebanese domains’ were created and showed better outcomes. These new domains had better composite reliability (CR) than the adapted domains, but more modest AVEs. Their discriminant validity (HTMT reports) was also satisfying.
Conclusion Merck and coll. ’s item distribution seems flawed; however, reverse scoring the cited items may improve its validity.The adapted domains and the ‘Lebanese domains’ seemed the most suitable for our population. Although the two distributions have their limitations, the ‘Lebanese domains’ were overall superior.