Study Hypothesis: Social mobilisation will attract and engage young people into decentralised sexual reproductive health (SRH) services where HIV prevention is tailored to need. Decentralised risk informed (differentiated) biosocial HIV prevention will reduce the overall prevalence of sexually transmissible HIV amongst young people aged 15-30.

Study aims:

1. Measure the impact of social mobilisation into decentralised SRH services that provide tailored HIV prevention on the prevalence of transmissible HIV.
2. Evaluate the acceptability, practicability, and reach of social mobilisation and decentralised SRH with tailored HIV prevention to deliver differentiated biosocial HIV prevention.

Study design: A step-wedge cluster randomised control trial comparing the effect of the Intervention (social mobilisation by peer navigators into mobile nurse-led SRH clinics) with Standard of Care (SoC) at Primary Health Care clinics, in reducing the proportion of 15-30 year olds with sexually transmissible HIV and increased uptake of risk-informed (differentiated) HIV prevention Intervention: Peer navigator social mobilisation intervention includes safe spaces, structured psychosocial and health needs assessment; Peer-mentorship with tailored health promotion, psychosocial support and lay-counselling; provision of condoms, HIV self-tests and pregnancy tests. The mobile SRH clinics provide sexually transmitted infection (STI) care, contraception, HIV testing and antiretroviral therapy (ART) if positive and Pre-Exposure Prophylaxis (PrEP) for those eligible (based on national guidelines) and negative.

SoC at Primary Health care clinics includes, contraception, HIV testing and ART if positive and PrEP for those eligible and negative.

Population: The eligible population are ~26,000 15-30-year-old residing in 40 administrative areas (clusters) of the uMkhanyakude district in rural KwaZulu Natal. The 40 clusters will undergo a stratified randomisation to early versus delayed intervention roll out.

Data collection: Research data will be collected from four sources: i) three random representative surveys of n= 2000, 15-30-year-olds; ii) programme, process, and clinical data; iii) qualitative data collected during the process evaluation iv) cost data using bottom-up ingredient-based costing and top-down costing using the budgets and expenditure reports.

Study Duration: 36 months.

This implementation trial aims to measure the effect, of 'social mobilisation into decentralised SRH services that provide tailored HIV prevention' on the prevalence of transmissible HIV amongst young people aged 15-30, as well as cost-effectiveness, practicability and reach.

Trial design: investigators will use a cRCT and realist process evaluation with costings. The stepped wedge design consists of 40 clusters (administrative areas) randomised to two intervention steps (early and delayed), 15 months apart. The investigators compare prevalence of transmissible HIV and uptake of risk-informed HIV prevention in intervention areas with contemporaneous controls. The investigators will conduct a cost effectiveness analysis and use a realist process evaluation to interrogate the extent to which the intervention components deliver according to the theory of change and support demand, uptake and retention in risk-differentiated biomedical HIV prevention amongst young people. Investigators will use the process evaluation data to inform any modifications within the intervention arm and to inform stop/go decisions at the end of step1.

Study setting: The study is taking place in the uMkhanyakude district in rural KZN, which is mostly rural with over 85% youth unemployment and a high burden of HIV.

Study population and eligibility criteria: About 26,000 young people aged 15-30 residing in 40 administrative areas (clusters) of the uMkhanyakude district in rural KwaZulu Natal are eligible to receive the intervention and control services. Investigators use the AHRI health and demographic surveillance as a sampling frame to randomly select three separate cross-sectional samples of n=3600 (90 per cluster) 15-30-year-olds stratified by gender at baseline, midpoint (before the second step of scale up) and end-line to measure trial outcome. Based on previous studies in this setting, investigators anticipate that ~2800 will be contactable and eligible and that n=2000 (~50 per cluster) men and women aged 15-30, will be willing and able to provide consent to be included in the study.

Study interventions, eligibility criteria and outcomes are described below

Sample size calculations: Based on current data 8% of the 15-30-year old's have a transmissible HIV viral load and ~35% are aware of their HIV status and either on PrEP or on ART and undetectable. With 50 per cluster (n=2000 15-30-year-olds total) in each of the three survey waves, assuming an ICC within waves of 0.1 for transmissible HIV and 0.4 for intervention uptake and a decay (autocorrelation) for both outcomes between waves of 0.9, our two-step stepped wedge trial design provides 90% power to detect an increase from 35% to 47% in uptake of intervention and 80% power to show a reduction of proportion with transmissible HIV from 8% to 4% between the intervention and standard of care. The estimates of ICC are based on the coefficient of variation (k) between clusters estimated in the range 0.7-1.0 for intervention uptake seen in a trial of peer navigators in the same area. In these calculations investigators assume that the intervention effect is fully realised within 12 months (gap from baseline to midline survey) and does not decline. Investigators will also report the intervention effect after 27 months (gap from baseline to end line).

Data collection:

(i) Following informed consent, the participants will complete a 20-30-minute interviewer administered RedCap survey in IsiZulu. Sensitive questions in this survey are self-filled. Questions will include awareness of HIV status, awareness and uptake of PrEP/VMMC, ART, exposure to youth groups, peer navigators, and mobile clinics, sociodemographic, sexual risk (e.g., number of partners, condom use, and transactional sex), reproductive health (e.g., contraception, pregnancy, fatherhood and VMMC); and mental health (PHQ9, alcohol and drug use). Interviewers will offer point of care HIV testing and collect Dry Blood Spots (DBS) or pediatric blood collection tubes (<2ml) for HIV ELISA and viral load testing on those that test positive.

(ii) Investigators will measure linkage to ART/PrEP through identifying eligible young people who link for PrEP/ prevention counselling or ART initiation within mobile services. Investigators will use the bar-coded vouchers and other identification within RedCap to confirm linkage. Within the 11 PHC investigators will use the ClinLink software which is effective in identifying individuals in the surveillance area and collecting reasons for attending when they attend PHC. Investigators will look at reasons for attending the 11 PHC in all 15-30-year-olds through CliniLink as well as ART and PrEP data in tiernet linked to the HDSS. Investigators will use an algorithm to identify which cluster the individual came from, including the bar code on the coupon they bring, their name, age, phone number, their HDSS number, and area of residence.

(iii) Investigators will collect data on uptake and retention in all components of the intervention in all 26,000 15-30-year-olds in the study area. a) Investigators will collect the programme data records from the peer navigator tools. This is a RedCap participant support management tool on the peer navigator tablets. They record the unique ID of the young people they meet and the GPS co-ordinates, demographic, and contact details of those who accept health promotion. They then complete a psychosocial and health needs assessment and record the health promotion, service and/or referral provided. All referrals are made using barcoded referral slips that are scanned into the tablet. They will finally generate an action plan including follow-up plans for those who need ongoing support e.g., on PrEP or ART. b) Youth groups attendance will be recorded (by scanning in barcodes). c) Investigators will collect clinical data from the RedCap clinical management tool. Investigators will use the clinical data to record HIV testing, ART uptake if positive, PrEP eligibility screening and PrEP uptake if negative, and other services received. Investigators will also measure retention, adherence, and reasons for stopping and/or restarting PrEP. By using this programme data that links the community outreach with the clinical services investigators will understand the reach and coverage of the programme. This includes coverage of treatment and reductions in HIV acquisition as a proportion of all young people in the area and as a proportion of the young people reached by the programme. Investigators will interrogate any systematic differences in uptake by gender, age and socio-economic characteristics and explore what works for whom and how through the process evaluation.

(iv) Qualitative data will include IDIs will be conducted peer navigators (n=10), nurses/clinical research assistants (n=6) stationed in clinics in the participating communities, and a purposive sample of young people aged 15-30 years (n=50). The interviews will be conducted by ARHI trained fieldworkers including a senior social scientist in English and isiZulu. This will enable the researchers to understand, contextualise and explore some of the issues around the intervention. To limit disturbances and ensure privacy, the IDI will be conducted in a private space suitable for the participant, and audio recorded with interviewees' consents. Researchers will conduct natural group discussions with community group and intervention delivery staff (n=7; 1 with clinic staff, 2 peer navigators' groups and 4 community groups (older men, older women and younger men and younger women). The group discussion will contain 6-8 people and will last approximately 120 minutes.

(v) The study team will work closely with HE2RO to establish costs. Investigators will adapt the data collection tools that were used in the STAR trial to measure the costs of the peer navigator intervention to collect bottom-up ingredient-based costs. This will be complemented with a top-down costing approach using the AHRI study budgets and expenditure reports.

Analysis: An intention-to-treat analysis will be conducted for each primary outcome. A significance level of 0.05 will be used unless otherwise stated. A detailed statistical analysis plan will be completed prior to the end of data collection.

To quantify the effect of the intervention on sexually transmissible HIV, the trial statistician will fit a logistic regression model to estimate the odds ratio (OR) and 95% confidence interval (CI), adjusting for design factors and survey wave, acknowledging the clustering of the data through generalized estimating equations. Since randomisation may result in an imbalance in the distribution of HIV risk factors across the intervention steps, they will also adjust for pre-specified predictors known to be associated with HIV transmission risk in this population at cluster and individual level.

To measure the uptake of universal risk informed HIV prevention intervention, the study statistician will calculate the proportion and 95% CI of participants who consent to participate in the intervention and are aware of HIV status; on treatment if positive; and have undergone risk assessment by peer navigators or SRH clinic if HIV negative (and taken up PrEP if HIV negative and eligible).

To examine the cost of transmissible HIV averted, investigators will measure the costs in intervention and control arms and compare the two arms in their cost-effectiveness in achieving endpoints, i.e., the cost per transmissible HIV case averted and the cost per case linked to risk differentiated biosocial HIV prevention. To establish costs, the investigators will use both a bottom-up ingredient-based costing approach and a top-down costing approach using the study budgets and expenditure reports.

Interim analyses: To inform modifications (6months): In the first quarter of the intervention arm, investigators will look at the uptake of the three different components of the intervention. Based on the pilot data investigators expect n=13,000 (50%) of all 15-30-year-olds receive at least one structured assessment from peer navigators; 8,000 (30%) of all 15-30-year-olds are seen in sexual health services; and 1000 (7%) will have engaged with effective use of PrEP. To be on track to reach these targets, by the first quarter of intervention role out in all 20 early clusters investigators would expect peers to have conducted needs assessment on n>=2000; n>=1000 would have been seen in the mobile clinic and n>= 75 would have started PrEP/ART. If these parameters aren't met in the first quarter the process data will be used to enhance the intervention. To inform a stop/continue decision: By the end of step 1 if less 40% (n<5000) of 15-30-year-olds undergo a peer-led needs assessment and/or less than n< 300 - i.e., less than one third of the 900 we expect to be eligible for risk informed PrEP - attend the SRH clinic investigators would consider stopping the intervention.

In-depth Interview Interviews will be based on a topic guide, and it is estimated that they will take up to an hour. They will be recorded and transcribed verbatim, with the permission of participants. Data from the IDIs will be analysed using NVIVO software. The software will be used for categorisation and coding of identified themes from the interview transcripts. Identified themes (including participants' quotes) and interview transcripts will be reviewed and compared by the research team for inconsistencies and adequate representation of participants' comments. A thematic analysis of all interview data will be conducted using framework approach.

Inclusion Criteria:

• Resident in the cluster
• Any gender
• Aged 15-30
• If aged 18-30 and able and willing to consent to participate
• If aged 15-17, able and willing to assent and whose guardian or parent are able and willing to consent to their participation.

Exclusion Criteria:

• Unable or unwilling to provide consent
• Aged 14 or less
• aged 31 or more

Study Protocol, SAP and ICF's are available now and Analytic code as soon as it is finalised for the primary analysis.

Everyone. It will be available at time of analysis of primary outcome through AHRI data repository.

Restricted randomisation was used to randomly allocate administrative areas in equal measure (to early and delayed roll-out of the intervention steps (20:20). Restricted randomisation was used to ensure that the intervention steps are reasonably balanced with respect to key characteristics (obtained from the HDSS) specifically, adolescent and young adult population size and density, new versus old areas, and geographic location along the N2 road. Interventions were assigned during a public ceremony to ensure transparency and fairness in the randomisation and engagement of the community. Randomisation and allocation were kept separate throughout. The statistician and investigators did not participate in the public randomisation and will remain blinded until the analyses of primary and randomised secondary outcomes have been finalised. Investigators remain blinded to allocation throughout. The participants and intervention delivery teams are not blinded.