Comparing the effectiveness of secondary distribution of HIV self-testing to testing card referral in promoting HIV testing among men who have sex with men in China: A quasi-experimental study

BackgroundSocial network-based HIV self-testing (HIVST) is useful to promote HIV testing. Secondary distribution is one social network-based method whereby individuals (indexes) access multiple HIVST kits and distribute them to their social networks (alters). This quasi-experimental study compared the effectiveness and cost of two social network-based HIV testing strategies (HIVST secondary distribution and HIV testing card referral) in promoting HIV testing among Chinese men who have sex with men (MSM). MethodsMSM aged 18 years or older were recruited in Guangzhou, Guangdong Province. From May to September 2019, indexes recruited during that period could distribute HIVST kits to people within their social network. Indexes recruited from October 2019 to January 2020 could refer HIV testing cards to people within their social network for free facility-based tests. Participants could access 1-5 HIVST kits or testing referral cards for distribution. Alters were encouraged to upload a picture of their test results and complete an online survey. Indexes and alters received an incentive to report test results. ResultsTwo hundred thirty-nine potential participants were assessed for eligibility and 208/245 (84.9%) were eligible. Among those who completed baseline assessment, 154/208 (74.0%) completed one month of follow-up. Overall,106 indexes were recruited in the HIVST arm and 102 in the testing card arm. The two arms had similar socio-demographic characteristics. At the one-month follow-up, 92 indexes in the HIVST arm self-reported having distributed self-test kits to 179 unique alters, and 62 in the testing card arm self-reported having distributed testing referral cards to 26 unique alters. Additionally, 69/92 (75%) in the HIVST arm distributed any test to friends or sexual partners compared to 18/62 (29%) in the testing card arm, with a risk difference of 46% (95% CI 31%, 61%). Indexes in the HIVST arm distributed an average of 1.95 (SD=1.90) tests, compared to 0.42 (SD=0.78) in the testing card arm, with a risk difference of 1.53 (95% CI 1.09, 1.96). Subgroup analysis suggested that indexes in the HIVST arm who self-identified as gay (p = 0.007) or were previously tested for HIV (p = 0.02) were more likely to distribute. The HIVST arm had a higher total cost and higher testing coverage compared to the testing card referral arm. The ICER per alter tested was $52.78. ConclusionsSecondary distribution of HIVST engaged more MSM to distribute tests to their social network and reached more MSM for test. MSM who self-identify as gay or who have previously tested for HIV were more effective in distributing tests. Future testing approaches should include HIVST kits in voluntary counselling and testing settings and incorporate digital strategies for secondary distribution.