Psychometric evaluation of the Health Care Provider HIV/AIDS Stigma Scale (HPASS) using Rasch analysis.

INTRODUCTION: The role of pharmacists is crucial in the care of individuals with HIV/AIDS. However, stigma in health care settings can be a deterrent to providing appropriate care. This paper assessed psychometric properties and convergent validity of the Health Care Provider HIV/AIDS Stigma Scale (HPASS) among pharmacy students in the United States (US) using Rasch analysis. METHODS: Students enrolled in four US universities were administered the survey (N = 203). Rasch analysis was conducted for each HPASS subscale (Stereotyping, Discrimination, and Prejudice) to assess dimensionality, model data fit, item difficulty, individual stigma, distribution of items and persons across item-person map, and rating scale function. Convergent validity evidence was established by comparing Pearson's correlation coefficients between HPASS subscales and the AIDS Attitude Scale (AAS) Avoidance subscale. RESULTS: Two items in the Prejudice subscale were misfit and therefore removed. The 6-point rating scale did not perform satisfactorily for HPASS subscales. Item difficulty ranges were wide [Stereotyping (-5 to 0.8 logits), Discrimination (-6 to 1 logits), Prejudice (-5 to 0.4 logits)]. Items were biased towards measuring higher levels of stigma. Person separation index was satisfactory (Stereotyping = 2.2; Discrimination = 2.06; Prejudice = 2.17) as was person separation reliability (Stereotyping = 0.83; Discrimination = 0.81; Prejudice = 0.83). Convergent validity was established by showing significant correlations between HPASS subscales and AAS Avoidance (P < .001). CONCLUSIONS: Modifying or removing misfit items of HPASS and exploring alternate rating scales for HPASS subscales will help better assess HIV/AIDS related stigma among pharmacy students.