Validity and reliability of the 36-Item Short Form Health Survey questionnaire version 2 among people living with HIV in Brazil

BACKGROUND: The 36-Item Short Form Health Survey questionnaire (SF-36) is a widely used instrument for evaluating health-related quality-of-life (HRQoL). The psychometric validation of the SF-36 version 2.0 (SF-36v2) in HIV-infected patients had not yet conducted in Brazil. AIM: To test data quality, reliability and validity of the SF-36v2 as a measure of HRQoL among Brazilian individuals living with HIV. METHODS: A cross-sectional study included 393 HIV-infected patients in whom HRQoL was assessed by using the SF-36v2 questionnaire. Demographic characteristics, socioeconomic status and clinical data were also collected. The SF-36 version 1 translated into Brazilian-Portuguese was adapted and introduced version 2 improvements according to the instrument developers. SPSS version 21 was used for data analyses. Confirmatory factor analysis (CFA) and structural equation modelling were performed using AMOS SPSS version 18. T-test for independent samples was used to compare differences between mean levels of HRQoL components in different groups. Linear multiple regression techniques were used to analyze the relationship between Physical Component Summary (PCS), Mental Component Summary (MCS) and independent variables. RESULTS: We performed the CFA and tested the hypothetical measurement model. We included five parameters to improve the adherence of data to the model. All indicators met the requirement suggested by specialized literature (χ2 (gl): 980.7 (527); CFI: 0.949; GFI: 0.873; TLI: 0.943; RMSEA: 0.047; 90% IC: 0.042-0.051). Multiple regression analyses revealed that depression (p = 0.001), family income (p = 0.007), sex (p = 0.001) and age (p< 0.021) were associated with MCS. Comorbidities (p = 0.001), health self-perception (p = 0.001), age (p = 0.001), and sex (p = 0.025) were associated with PCS. CONCLUSIONS: A consistent validation of the SF-36v2 in Brazilian HIV patients were shown. Further studies with SF-36v2 psychometric analyses would be required in other populations to establish Brazilian normative data.