Assessing whether providing regular, free HIV self-testing kits reduces the time to HIV diagnosis: an internet-based, randomised controlled trial in men who have sex with men.

Background: The risk of onwards HIV transmission is strongly influenced by the interval between HIV infection and its diagnosis. The SELPHI trial examined whether this interval could be reduced by offering free HIV self-testing kits to men-who-have-sex with-men (MSM).

Setting: Internet-based RCT of MSM aged ≥16 years, resident in England/Wales, recruited via sexual and social networking sites.

Methods: The second-stage randomisation of SELPHI was open to participants who used an initial free HIV self-test kit, were HIV-seronegative, and reported recent condomless anal sex. They were randomised to receive a free HIV self-test kit every 3 months (repeat testing[RT] group) versus no such offer (nRT group). The primary outcome was time from randomisation to a confirmed HIV diagnosis, determined from linkage to national HIV surveillance databases. The key secondary outcome was the frequency of HIV testing regardless of test modality.

Results: 2308 eligible participants (1161 RT,1147 nRT) were randomised between April-2017 and June-2018, and followed for 15-27 months. The proportion of participants reporting an HIV test in the previous 3 months was much higher in the RT group (86%) than in the nRT group (39%). Overall, 16 (9 RT,7 nRT) confirmed HIV diagnoses were observed (0.35/100 person-years), with no difference in the time to a confirmed HIV diagnosis (hazard ratio=1.27 [95% CI 0.47-3.41], P=0.63).

Conclusions: Providing regular free self-testing kits to sexually-active MSM was highly acceptable and markedly increased HIV testing. However, in this low incidence cohort it did not result in a demonstrably more rapid diagnosis of incident infections.