Article
Cross-cultural adaptation of the Perceived Risk of HIV Scale in Brazilian Portuguese
Registro en:
TORRES, Thiago S. et al. Cross-cultural adaptation of the Perceived Risk of HIV Scale in Brazilian Portuguese. Health and Quality of Life Outcomes, v. 19, n. 1, p. 1-12, 2021
1477-7525
10.1186/s12955-021-01760-6
Autor
Torres, Thiago S.
Luz, Paula M.
Marins, Luana M. S.
Bezerra, Daniel R. B.
Almeida-Brasil, Celline C.
Veloso, Valdiléa G.
Grinsztejn, Beatriz
Harel, Daphna
Thombs, Brett D.
Resumen
Background: Valid and reliable instruments are needed to measure the multiple dimensions of perceived risk. The Perceived Risk of HIV Scale is an 8-item measure that assesses how people think and feel about their risk of infection. We set out to perform a cross-cultural adaptation of the scale to Brazilian Portuguese among key populations (gay, bisexual and other men who have sex with men and transgender/non-binary) and other populations (cisgender heterosexual men and cisgender women).
Methods: Methodological study with cross-sectional design conducted online during October/2019 (key populations [sample 1] and other populations) and February-March/2020 (key populations not on pre-exposure prophylaxis [sample 2]). Cross-cultural adaptation of the Perceived Risk of HIV Scale followed Beaton et al. 2000 guidelines and included confirmatory factor analysis, differential item functioning (DIF) using the Multiple-Indicator Multiple-Cause model, and concurrent validity to verify if younger individuals, those ever testing for HIV, and engaging in high-risk behaviors had higher scores on the scale.
Results: 4342 participants from key populations (sample 1 = 235; sample 2 = 4107) and 155 participants from other populations completed the measure. We confirmed the single-factor structure of the original measure (fit indices for sample 1 plus other populations: CFI = 0.98, TLI = 0.98, RMSEA = 0.07; sample 2 plus other populations: CFI = 0.97, TLI = 0.95, RMSEA = 0.09). For the comparisons between key populations and other populations, three items (item 2: "I worry about getting infected with HIV", item 4: "I am sure I will not get infected with HIV", and item 8: "Getting HIV is something I have") exhibited statistically significant DIF. Items 2 and 8 were endorsed at higher levels by key populations and item 4 by other populations. However, the effect of DIF on overall scores was negligible (0.10 and 0.02 standard deviations for the models with other populations plus sample 1 and 2, respectively). Those ever testing for HIV scored higher than those who never tested (p < .001); among key populations, those engaging in high-risk behaviors scored higher than those reporting low-risk.
Conclusion: The Perceived Risk of HIV Scale can be used among key populations and other populations from Brazil.