The investigators will implement a 24-month fractional factorial design study (Aim 1). The investigators will recruit 510 patients initiating antiretroviral therapy (ART) at three City of Cape Town ART clinics. Each patient will have adherence monitored using the Wisepill® electronic adherence monitoring device (EAM). After eligibility has been confirmed, each participant will be randomized to one of 16 experimental conditions. Each condition includes a unique combination of five adherence intervention components. Three of these components focus on identifying individuals with poor adherence, with increasing degrees of sophistication, with immediate linkage to adherence support. Two components focus on supporting good adherence. They both supplement the existing adherence support program delivered at the study clinics operated by City of Cape Town (standard of care component). Based on Self-Determination Theory, the investigators postulate these intervention components will: 1) enhance feelings of autonomy support, social support, and knowledge; 2) improve motivation and self-competence; and 3) increase ART retention, adherence, and viral suppression. A subset of the participants, as well as clinic staff, will be invited to in-depth interviews to explore mediating factors (Aim 1) and the implementation process (Aim 2); and the data collected in Aims 1 and 2 will be used to explore cost effectiveness (Aim 3).

The study's primary research goal is to identify the optimal combination of evidence-based and scalable HIV interventions for low-resource, high-burden settings. The investigators propose to 1) test the relative contribution of five promising intervention components; 2) collect cost and other implementation data; and 3) create a multi-component intervention package to optimize cost-effectiveness and implementation success. Of the five components, three are methods of non-adherence detection plus patient outreach; two are adherence support methods that can be integrated into Cape Town healthcare systems. These will not overcome all challenges that ART patients experience (e.g., structural barriers such as food insecurity) but they represent scalable, feasible, acceptable, and effective options. Notably, they are all behavioral approaches grounded in the experience and priorities of local health officials with whom the investigators have worked to identify scaleable interventions. While the study will be in Cape Town, it is broadly adaptable to other resource-limited settings.

The gold standard for testing interventions is the randomized controlled trial (RCT), which minimizes bias when testing cause and effect of a new exposure. When testing an intervention with more than one element, however, untangling the effect of individual elements is impossible. Indeed, data on the performance of individual components and their interactions-critical for developing and refining the components of a packaged intervention-is lost in an RCT. Notably, clinical care typically relies on packages of services, not single interventions, and packaged interventions are recommended for ART support. An effective way to test a multi-component intervention is to use the novel Multiphase Optimization STrategy (MOST), an engineering-inspired method for identifying the most efficacious combination of components in a packaged intervention, thus allowing researchers to drop inactive or weakly-performing components and construct an optimized package based on effect, cost, and other features. Once the optimized multi-component intervention is chosen, an RCT or quasi-experiment can follow to determine whether the optimized package yields superior outcomes compared to existing standards. MOST encompasses three phases: 1) preparation; 2) optimization; and 3) evaluation, often in an RCT. In this project, we have completed preparation, including a pilot study in Cape Town. SUSTAIN will comprise the middle optimization phase. The evaluation phase will be the focus of a future study.

The specific aims are:

Aim 1. Employ a highly efficient fractional factorial design to determine the effects of five intervention components on the primary outcome (HIV viral suppression) and secondary outcomes (ART adherence measured by EAM, ART retention per clinic records, days of unsuppressed virus, time to nonadherence detection, and time to linkage to support). The investigators will explore effect mechanisms quantitatively and qualitatively.

Aim 2. Evaluate the intervention components to address implementation, service, and client outcomes according to the Proctor framework. Data collection will involve tracking of intervention component use, time and motion studies, and quantitative surveys and qualitative interviews with participants and staff.

Aim 3. Use the effectiveness data collected in Aim 1 and the implementation and client outcomes in Aim 2 to model the multi-component intervention optimized for cost-effectiveness and implementation success.

Study Summary This study is designed to advance the translation of evidence-based interventions into clinical settings to benefit patients. There is ample evidence on what works to support ART adherence and retention-much of it from our own research. The investigators partnered with local officials and clinical staff in Cape Town to review the evidence and to conduct formative research to identify the most effective, acceptable, and feasible intervention options for patients and providers. The proposed study represents the next critical step: the investigators will test the intervention components that emerged from the formative work, encompassing elements to both rapidly identify nonadherent patients and to strengthen the support they receive once identified, to provide the data needed to construct the most cost-effective and sustainable multi-component intervention. The choice of intervention components will allow a critical test of advanced monitoring technology compared to simpler tools to identify nonadherence. By using an innovative MOST design to guide collection and analysis of efficacy, cost, and other implementation data, the study aligns with NIH's goals of using novel scientific methods to advance implementation science.

Inclusion Criteria - Main study:

• Adults (≥18 years) and adolescents 16-17 years.
• HIV-positive and attending a local City of Cape Town (COCT) clinic to commence ART.
• Able to provide full informed consent, with a written signature. For those who are illiterate, a witness will be present throughout the process and will sign the form, while the participant will add their right thumb print. For those who are aged 16-17 years, informed written assent will be obtained, and the adolescent must have a parent or guardian who can provide full informed consent (see **below for how parent/guardian is defined for this purpose).
• Access to a working cellphone and willingness to receive study-related messaging on that phone.
• Willingness to comply with study procedures, including providing regular updates of contact details /locator information, and use a EAM device for the duration of participation.

Other Inclusion Criteria:

Aim 1: In-depth interviews (IDIs) with subset of trial subjects at baseline and months 12 and 24.

• Participation in the main trial.
• Self-reported prior experience with substance use, depression, gender inequity, stigma, or transport/clinic issues.

Aim 2: Questionnaires and IDIs with staff members at study clinics (three total clinics).

• Adults (≥18 years)
• Staff at study clinics, providing HIV care and/or treatment.

Aim 2: Focus group discussion (FGD) with City of Cape Town officials.

• Adults (≥18 years)
• Staff at City of Cape Town.

Exclusion Criteria - Main study:

• Clinical conditions as assessed by the COCT clinic clinicians at first visit e.g. renal disease, which preclude the use of a single tablet regimen (with the exception of those on tuberculosis (TB) treatment who are required to take an extra dose of dolutegravir daily).
• Planning to leave Cape Town permanently within the next 24 months.
• Being perinatally infected with HIV. Being infected from birth typically means a set of experiences and complications at a young age that require unique and special attention.
• If an adolescent, taking their ART medication as a syrup, as they are required to use the electronic adherence monitor (Wisepill device), which is only suitable for tablets.

Other Inclusion Criteria:

Aim 1: IDIs with trial subjects.

• None.

Aim 2: Questionnaires and IDIs with staff members at clinics. • None.

Aim 2: FGD with City of Cape Town officials.

• None.

We have not yet made a detailed plan, but tentatively, we will deposit de-identified IPD in a repository at Boston University (BU) School of Public Health and the Desmond Tutu Health Foundation that can be made available to other researchers once our planned analyses are complete.