Globally, a substance use disorder (SUD) treatment gap exists, meaning far fewer individuals with SUD receive treatment compared to those who need it [1,2]. This gap is heightened in low- and middle-income countries (LMICs), with only 1%‐4% of individuals in need of treatment receiving minimally adequate treatment [2,3]. There is an urgent need to address the SUD treatment gap for people with HIV, as they are a vulnerable population at high risk for morbidity and mortality as a result of untreated SUD [4,5]. South Africa is home to the highest number of people with HIV worldwide, with an estimated prevalence of 8.15 million [6]. Although antiretroviral therapy (ART) coverage almost doubled in the past decade, with over 70% of people with HIV on ART [2,7], certain groups remain at risk for poor HIV outcomes and ongoing HIV transmission [7-9], including individuals with SUD. Among people with HIV in South Africa, the prevalence of SUD is high and, when untreated, is associated with poor ART adherence and HIV care outcomes [10,11].

Despite the negative impact of untreated SUD on ART adherence, there is little integration of HIV and SUD services globally [12,13]. This care fragmentation creates numerous barriers to accessing SUD care for people with HIV [14,15], contributing to both HIV- and SUD-related morbidity and mortality. A main barrier to delivering evidence-based SUD care is the shortage of providers trained to deliver integrated ART adherence and SUD interventions. In response to workforce shortages, task-sharing models—or training and supervising nonspecialist cadres of health workers—have been implemented to expand access to ART and mental health services [16]. Further, stepped-care models have also seen promising results in LMICs; these models deliver the least resource-intensive intervention first, and only those who continue facing difficulties in care receive the more intensive intervention [17]. A stepped-care design to maximize optimal use of resources is a necessary step toward the rollout and scale-up of accessible intervention programs for people with HIV with comorbid SUD globally [18].

Our team has demonstrated the potential benefits of task sharing with peer interventionists—individuals with relevant lived experiences (eg, SUD and HIV) who are trained and supervised to deliver behavioral interventions [19,20]. Results from our team’s formative work have shown that patients with SUD prefer working with a peer to other health care workers, given their shared lived experience, and that training and supervising peers to improve ART adherence among people with HIV with SUD is highly feasible and acceptable [20]. However, research is needed on how to best implement task-shared, peer-delivered SUD and ART adherence interventions such that they can be feasibly and sustainably integrated into primary health care, particularly using stepped care approaches to enhance the feasibility of integration into primary care.

Our team has developed a peer-delivered behavioral intervention Khanya—meaning “glow, light, direction” in isiXhosa—that includes evidence-based components of behavioral activation, problem-solving, motivational interviewing, and mindfulness-based relapse prevention skills to improve ART adherence and reduce substance use. We have conducted prior pilot work where key stakeholders were interviewed to capture local attitudes toward integration and peer interventionists, and formative work was conducted to gather feedback from these stakeholders to adapt Khanya [21]. A type 1 pilot hybrid randomized effectiveness-implementation trial (n=61) evaluating Khanya demonstrated preliminary feasibility, acceptability, and fidelity for peer delivery, as well as preliminary effectiveness for ART adherence at 3 months measured using Wisepill. Further, results demonstrated reductions in alcohol use among individuals who were also using other drugs [19,20]. Pilot cost-effectiveness data suggested preliminary evidence that Khanya is cost-effective contingent on stakeholders’ willingness to pay for improved ART adherence [22]. Preliminary results from this pilot trial suggested a larger, fully powered randomized trial was warranted. Further, in line with efforts to maximize the efficiency of intervention delivery and allocation of resources, the next steps included testing a stepped-care adapted model of the intervention.

This trial builds upon the pilot study to conduct a fully powered, hybrid type 2 effectiveness-implementation randomized controlled trial (n=160) to evaluate a stepped-care model of Khanya vs enhanced treatment as usual (ETAU) on effectiveness outcomes (HIV and SUD treatment outcomes over 12 months), implementation outcomes (acceptability, feasibility, fidelity, and uptake), and cost-effectiveness of the intervention compared to ETAU. Based on formative work, we hypothesize that a stepped-care approach to delivering the Khanya intervention will improve ART adherence, reduce substance use, and have promising implementation and cost-effectiveness outcomes compared to ETAU. The overall aim of this work is to inform the development and implementation of evidence-based SUD services that can be feasibly task-shared and integrated into low-resource HIV care contexts globally.

This trial is a 2-arm, hybrid type 2 effectiveness-implementation randomized controlled trial comparing the effectiveness of Khanya versus ETAU among people with HIV with SUD at risk of ART nonadherence. The trial design was guided by the RE-AIM (reach, effectiveness, adoption, implementation, and maintenance) framework, which was selected given its dual focus on effectiveness and implementation outcomes and prior applications in Sub-Saharan Africa [23-25]. The RE-AIM framework guided planning and selection of the effectiveness, implementation, and cost-effectiveness outcomes for the trial (see the Measures section).

All participants (n=160) are being recruited from two primary health care clinics located in the periurban suburb of Khayelitsha, situated on the outskirts of Cape Town, South Africa, which has the highest HIV prevalence in South Africa’s Western Cape [26,27]. The population of Khayelitsha is predominantly Black African and isiXhosa speaking. All study materials are translated into isiXhosa and back-translated into English to ensure all materials delivered in isiXhosa are accurate to the original English versions. The primary health care clinics have HIV and tuberculosis services integrated into primary care and are publicly funded. In Khayelitsha, there is a freely available SUD treatment program, Matrix, an evidence-based program run by the City of Cape Town that offers a 16-week rehabilitation focused on cognitive-behavioral strategies to support early recovery and relapse prevention [28-30], which is colocated at one of the primary health care clinics.

Potentially eligible and interested participants demonstrating ART adherence difficulties are referred to the trial by their HIV care team and screened for inclusion by the research team (see Eligibility Criteria). A regular on-site clinic presence by our team also ensures that all potentially eligible participants are aware of the study and can enroll if interested. All study procedures (screening, assessments, and intervention sessions) take place in private rooms located adjacent to the clinic in a built prefabricated unit at one clinic and in a designated study room in the second clinic. Screened participants who meet the study inclusion criteria and express interest in participation are then invited to complete an informed consent process prior to baseline by a trained study research assistant (see Ethical Considerations).

To be eligible for the study, participants must meet the following inclusion criteria: (1) diagnosed with HIV and prescribed ART, confirmed via medical records; (2) ≥18 years of age; (3) at least moderate SUD risk in the past 3 months for one or more nontobacco substances (measured by the WHO ASSIST [World Health Organization’s Alcohol, Smoking, and Substance Involvement Screening Test]: score ≥11 for alcohol, ≥4 for nontobacco substances) [31]; and (4) demonstrated ART nonadherence and/or risk of virologic failure, defined as at least one of the following in the past 12 months: (a) being out of care ≥1 month (confirmed by pharmacy refill data); (b) ≥1 episodes of viral load >400 copies/mL; and/or (c) on second- or third-line ART. Exclusion criteria include (1) severe risk/likely dependence for opiates (WHO ASSIST score >26) because opiate substitution therapy is not widely available; (2) severe alcohol dependence symptoms that warrant medical management of potential withdrawal symptoms/stabilization prior to study participation (ie, due to referral needs to a higher level of care given opioid substitution therapy is not available at the study clinic sites, and severe alcohol dependence symptoms may warrant medical management for withdrawal); (3) inability to provide informed consent or complete study procedures in isiXhosa or English; (4) in the third trimester of pregnancy at enrollment; (5) currently enrolled in another study focused on substance use or ART adherence or another substance use treatment program; or (6) active, untreated, or undertreated severe mental illnesses (eg, untreated psychosis or mania) that may interfere with study procedures. After medical stabilization due to any of the exclusion criteria noted here, individuals can be rescreened for eligibility if they are interested in participating.

After participants provide informed consent, they are invited to complete a baseline assessment, which includes a battery of self-report assessment measures related to psychosocial status, HIV infection history, ART adherence, and current and past SUD, along with biological measure sample collection. At the baseline visit, participants in both conditions are provided with a Wisepill device for real-time, wireless, electronic adherence monitoring [32,33]. The 2-week period between baseline and randomization serves as the baseline Wisepill assessment, and Wisepill adherence monitoring is continued over 1 year.

Randomization to the intervention condition occurs 2 weeks after the baseline assessment (to allow for a 2-week period of Wisepill monitoring as the baseline assessment of ART adherence). Eligible participants are then randomized 1:1 to either ETAU or Khanya in REDCap (Research Electronic Data Capture) by a trained research assistant. The Khanya condition also receives all ETAU services. The randomization module in REDCap contains a random allocation sequence that only members of the data management team at the University of Maryland will have access to [34]. Once “randomize” is selected by the research assistant, the participant’s condition is revealed and cannot be edited. The CONSORT (Consolidated Standards of Reporting Trials) diagram is presented as Figure 1.

Across both conditions, there are 3 major assessment visits after baseline: 3-, 6-, and 12-month follow-ups. All participants also complete a brief assessment at 1-month follow-up (see the CONSORT diagram in Figure 1). These follow-up assessments are conducted by a research assistant who is blinded to participant condition, including for all REDCap assessment procedures. Urine testing is conducted using a 6-panel multitest for commonly used substances (amphetamines, cocaine, methamphetamine, opiates, methaqualone, and delta-9-tetrahydrocannabinol [cannabis]) at all major visits. Participants also undergo a finger prick (an amount of 50 μL) of blood for dried blood spot (DBS) testing and a separate viral load assay if viral load results are not available within 30 days of each study assessment visit via chart review.

A subset of Khanya participants (n=30) will complete a qualitative interview. Participants eligible for the qualitative interview will be identified during weekly supervision meetings with the clinical supervisor, project director, and peer interventionist. During these sessions, the peer interventionist will discuss participants attending Khanya sessions. This will include the new participants who were recently randomized and those currently attending the Khanya intervention sessions. The participants demonstrating notable challenges or ART adherence strategies will be recruited at their 3-month follow-up assessment. This will be after consensus is reached by the respective team members on their eligibility. The criteria will include, but are not limited to, nondisclosure to partners or family, nonacceptance or denial of HIV status or SUD, stigma and discrimination, nonadherence to ART, and continued substance use while on ART. If they consent, a qualitative interview will be scheduled for their 6-month follow-up visit. The qualitative interviews will be conducted by research assistants at 2 study sites using a structured interview guide in isiXhosa (see Multimedia Appendix 1). The interviews will last approximately 30‐45 minutes and will be translated into English after completion and transcribed (see Qualitative Data).

All participants are reimbursed for each major assessment and travel expense using an electronic cashless payment system. Participants do not receive financial incentives to attend intervention sessions beyond reimbursing travel costs incurred to attend intervention visits (ZAR 50 [US $3]). The team also conducts chart extraction at major visits (baseline and 3-, 6-, and 12-month follow-ups) to assess relevant medical details per the study protocol (eg, date of HIV diagnosis, HIV-related laboratory results, and ART medication information) using chart extraction methods previously piloted [19]. At the 1-month follow-up visit, a brief chart extraction is conducted where data are collected to confirm the type of ART the participant is taking, whether they have been diagnosed with any conditions in the past 3 months, and if they have picked up any non-ART medications from the pharmacy. The list of outcome measures and schedule of assessments is depicted in Table 2.

All research assistants regularly complete Good Clinical Practice trainings through the CITI Program (Collaborative Institutional Training Initiative) before study initiation and throughout the trial [73]. They also received in-depth training from project directors on all data collection instruments, the use of REDCap, and the use of Wisepill. Standard operating procedures for informed consent, maintaining confidentiality, data management, adverse events, providing referrals, collecting dried blood spots and urine samples, and conducting medical chart extraction are reviewed by all members of the research team.

Peer interventionist training was delivered over approximately 3 days, with ongoing weekly supervision and booster training over 1 month. Peer training includes basic counseling and peer recovery training, HIV psychoeducation and adherence counseling training, and Khanya intervention–specific skills, including motivational interviewing, problem-solving, behavioral activation, and mindfulness-based relapse prevention. Training included both in-person and virtual components. Initial training in basic counseling skills and intervention-specific skills, such as behavioral activation and problem-solving, was facilitated by a clinical psychologist and psychiatric nurse at UCT. Motivational interviewing training sessions were conducted by a clinical psychologist and peer interventionist at the University of Maryland virtually, with additional support from a peer based in South Africa who delivered the intervention in the pilot. The interventionists also attended an advanced HIV counseling course offered in the HIV Mental Health Research Unit at UCT.

Peer training included sharing of personal lived experience with a peer supervisor, guiding principles of recovery, motivational interviewing, principles in the peer role, stages of change, and setting boundaries in counseling. Interventionists engaged in role-plays of Khanya sessions with the local clinical psychologist, who provided feedback before moving on to subsequent sessions. The interventionists also assisted each other with role-plays as a way of peer tutoring/training.

The readiness of the peer interventionist to deliver the intervention was assessed through the rating of the first 3 sessions using the ENACT to measure therapeutic competency [68]. ENACT findings were discussed by the training team to assess and confirm interventionist readiness for session delivery.

Throughout the study, the two peer interventionists receive weekly supervision from a local bilingual (English and isiXhosa speaking) clinical psychologist in South Africa and support from a US-based peer recovery specialist with relevant lived experience and training in peer recovery in the United States. The interventionists discuss their current caseload and any questions or concerns they have about how to best support each participant. This also provides an opportunity for interventionists to receive support from the clinical psychologist around their own mental health and workload. Supervision includes ongoing training of the intervention material, along with any additional information that might pertain to a specific participant’s case where support is needed. This allows for continued honing of intervention delivery skills and equips peers with additional knowledge to help navigate future scenarios with participants. The clinical psychologist also checks for fidelity and compliance with ethics requirements while also encouraging peer-to-peer support for participant cases and personal well-being. Immediate support for interventionists to debrief with a second clinical psychologist after sessions involving severe risk and distressed participants is also provided.

The main analysis on which the sample size calculation for the effectiveness analysis was based is the effect of the Khanya intervention versus ETAU on ART nonadherence measured using weekly Wisepill data over 12 months. Our pilot data that compared Khanya to ETAU at 3 months revealed a large effect size (d=1.02) on ART adherence in favor of the Khanya intervention. However, given that effect size estimates generated from pilot data using small samples are highly variable and imprecise [74], we expect the true effect size may be smaller or larger than the one observed in our pilot study. Furthermore, we anticipate a slightly smaller effect size in the study because we will assess our primary outcome over 12 months (rather than 3 months in the pilot) and not all patients will receive the full Khanya package. Based on our pilot data, after 1 month, we expect only ~56% of participants would need to be stepped up to the second step of the intervention. Accounting for a conservative anticipated effect size estimate of 0.5 (moderate) and an attrition rate of 20% at 12 months, we will have over 0.8 power to detect a difference in ART adherence between groups at 12 months with our sample size of 160.

This study is registered on ClinicalTrials.gov (NCT05933226; registration date: June 27, 2023), approved by the UCT Human Research Ethics Committee, reference number 077/2022. Informed consent is obtained from all potential participants prior to enrollment. Protocol modifications and amendments will be submitted to the UCT Ethics Committee for review and approval.

During informed consent, participants are informed of all potential risks of study involvement, that participation is voluntary and will not in any way affect their clinical care, and that they may withdraw their participation at any point during the study. The study will adhere to the ethical considerations of informed consent, right to autonomy, privacy, confidentiality, and anonymity. To join the study, the participants will have to fully understand and sign the consent form. A study staff member will review the form with each participant, which will include all study procedures, information about potential risks and benefits of participation, and information regarding whom they can contact for further questions. Participants will also receive information about the biological samples that will be collected at each time point such as the blood and urine samples, storage and disposal of the samples by the laboratory after testing, and how their data will be used. The participant will have as much time as they want to review the informed consent form and ask questions. This will ensure that the participants make an informed decision to participate in this study. As part of the informed consent process, participants provide a separate consent to audio-record intervention sessions (as applicable) and a medical release form to access only relevant medical records for the study. If participants decline to be audio-recorded or to sign the medical release form, they can still participate in the study.

Anonymity will be ensured by the use of screener IDs and participant (study) ID allocation on the day of baseline. The link between names and study ID numbers will be kept separately under lock and key or password protection. The right to autonomy will be upheld by stating to the participants that the participation in this study is voluntary. In addition, it will be stated that the participants can refuse to answer any question, that participants can withdraw from the study at any time, and that study participation will in no way be related to the care they receive at the clinic. Data will be kept confidential, under lock and key or in a password-protected file on a secure server that is only available to authorized study staff. Urinalysis will be conducted by a trained research assistant in a private area of the clinic using 6-panel urinalysis alcohol and drug testing kits. Participants will be compensated with a ZAR 150 (approximately US $11) for each major study assessment. Travel costs will also be reimbursed for all nonmajor visits (ZAR 50 [US $3]).

This study has oversight from a Data Safety and Monitoring Board (DSMB). The DSMB meets annually to review data collection procedures and ensure the scientific and ethical rigor of the study is maintained. DSMB members are independent of the sponsor without competing interests for providing rigorous scientific and ethical oversight. There are no a priori stopping rules for discontinuing or modifying provision of the experimental intervention, although our DSMB has this right and responsibility.

We collect information on adverse events that are experienced by enrolled participants. Adverse events are classified as serious/nonserious, related/unrelated to the study intervention, and expected/unexpected. All serious adverse events are reported to the UCT Human Research Ethics Committee within 24 hours. We report all adverse events to our DSMB committee annually.

This study protocol adheres to the TIDIeR (Template for Intervention Description and Replication) checklist for manuscripts reporting testing of an implementation intervention, and the SPIRIT (Standard Protocol Items: Recommendations for Interventional Trials) checklist for randomized clinical trial protocols (see Multimedia Appendix 1) [81,82].