Assessment of an instrument scale measuring the knowledge of antiretroviral therapy among HIV cases

Abstract Background: Antiretroviral therapy (ART) is currently the most effective measure to treat people were living with human immunodeficiency virus (PLHs) and reduce HIV transmission. While there are many factors that reduce adherence to ART, PLHs’ knowledge about ART may determine the level of adherence. It is necessary to design and assess an instrument scale that measures the knowledge of antiretroviral therapy among HIV cases. Method: A cross-sectional study was conducted among HIV cases in Honghe Hani and Yi Autonomous Prefecture, China. Both exploratory and confirmatory factor analyses were used to examine the latent factors of antiretroviral therapy knowledge scale. Internal consistency was assessed separately for the scale and its dimensions by estimating Cronbach’s alphas, split-half reliability and Spearman's correlation coefficient. ANOVAs were used to compare the scores of different dimensions with sociodemographic characteristics. Results: Exploratory and confirmatory factor analyses were performed in two independent samples: a development sample (N = 205) and a validation sample (N = 205). According to factor loadings, four factors were extracted. They had high internal consistency reliability (Cronbach’s alpha: 0.70 - 0.95) and good construct validity (standardized factor loading range: 0.46–0.86) in the scale. Goodness of fit indices indicated that a four-factor solution fit the data at an accepted level (χ2/degree ratio=1.980, RMSEA=0.069, GFI=0.909, CFI=0.957, NFI=0.917, TLI=0.944). ANOVAs indicated that the score was higher among HIV cases who were Han, had spouses/partners, were non-famers or migrant workers, initiated ART, and had a high school or above education. Conclusion: The psychometric assessment indicated that this ART knowledge scale had accepted internal consistency and discriminant construct validity. It can be used to assess the knowledge of antiretroviral therapy for HIV cases.