Smartphone game reduces high-risk sexual debut in Kenyan adolescents in 45-month RCT

BACKGROUND: Low-cost and scalable behavioural interventions to prevent HIV among young people in Africa are needed. As access to smartphones grows, digital interventions will become increasingly viable. We assessed the efficacy of an interactive narrative-based smartphone game (Tumaini) designed to increase age and condom use at first sex. METHODS: In this two-arm, non-blinded, individually randomised trial, participants aged 12-14 years were recruited at community level in Kisumu, Kenya. Sex-stratified block randomisation (1:1; block size of 10) was undertaken by study staff with no contact with participants. For three 5-7-week gameplay periods during school holidays in the first 3 years, participants in the intervention group received Tumaini and participants in the control group received Brainilis, a maths game, on study-provided low-cost smartphones. Participants were followed-up for 45 months, completing a baseline survey and 12 follow-up surveys. Assignment was known to trial staff but masked to data analysts. The primary outcome was a binary measure of high-risk sexual debut (no condom use) versus low-risk (condom use) or no sexual debut, assessed at endline in the intention-to-treat population. This trial is registered with ClinicalTrials.gov, NCT04437667, where it is suspended pending extension. FINDINGS: Between Oct 1 and Dec 3, 2020, 996 adolescents (mean age 14·0 years [SD 0·6]; 499 in the intervention group and 497 in the control group) were enrolled. At endline, 974 (485 in the intervention group and 489 in the control group, and 484 female and 490 male) completed the study and were included in the intention-to-treat analysis. Excluding those reporting sexual debut before baseline, 10% (41 of 409) of participants in the control group reported a high-risk sexual debut compared with 6% (23 of 407) of participants in the intervention group (risk ratio [RR] 0·56 [95% CI 0·34-0·92]). For female participants, 10% (21 of 207) in the control group and 3% (seven of 206) in the intervention group reported a high-risk sexual debut (RR 0·34 [95% CI 0·15-0·77]). For male participants, this outcome was not significantly different, with 10% (20 of 202) of participants in the control group versus 8% (16 of 201) of participants in the intervention group reporting a high-risk sexual debut (RR 0·80 [95% CI 0·43-1·51]). There was no difference in experience of or age at sexual debut. No serious adverse events resulted from trial participation. INTERPRETATION: The results of the study provide evidence to support the potential of digital behavioural HIV prevention, serious games for sexual health, and pre-sexual risk intervention among adolescents in sub-Saharan Africa. FUNDING: US National Institute of Mental Health and Emory Center for AIDS Research.