BACKGROUND: HIV self-testing (HIVST) could promote HIV tests; however, knowledge about the mechanisms by which providing HIVST kits impacts the uptake of HIV testing is limited. This study aimed to examine how self-efficacy mediates the association between providing HIVST kits and HIV testing frequency.
METHODS: In this randomised controlled trial, HIV-negative men who have sex with men (MSM) were recruited in China and randomly assigned 1:1 into intervention and control groups. MSM in the control group could access site-based HIV testing services (SBHT). MSM in the intervention group could access SBHTs plus free HIVST kits. HIV testing self-efficacy, number of SBHTs, number of HIVSTs, and the total number of HIV tests were assessed every 3 months for 1 year.
RESULTS: Data from 216 MSM (intervention group: 110; control group: 106) were included in analysis. Results from Pearson’s correlations and point-biserial correlations showed that participants with higher self-efficacy scores underwent a higher number of HIV tests (r = 0.241, P < 0.001), more HIVSTs (r = 0.162, P < 0.001), and more SBHTs (r = 0.138, P < 0.001). Bootstrap tests using PROCESS indicated self-efficacy partially mediated the effect of providing HIVST on the number of HIVSTs (indirect effect: 0.018, 95% bias-corrected confidence interval [BC CI]: 0.003–0.035; direct effect: 0.440, 95% BC CI: 0.366–0.513) and on total HIV tests (indirect effect 0.053 (0.030–0.787); direct effect 0.452 (0.365–0.539)), and completely mediated the effect of providing HIVST on the number of SBHTs (indirect effect 0.035 (0.019–0.056); direct effect 0.012 (-0.059 to 0.083)).
CONCLUSIONS: Our findings indicated self-efficacy mediated the effect of HIVST provision on HIV testing frequency, suggesting improving self-efficacy might be an effective way to promote HIV testing among Chinese MSM.
Self-efficacy mediates the effect of providing HIV self-testing kits on HIV testing frequency among Chinese men who have sex with men: Results from a randomised controlled trial
Sexual Health, 20 (2), 118-125. doi: 10.1071/SH22110.