Impact of a youth and caregiver intervention on HIV self-management among youth living with HIV: a protocol for the Family Connections cluster RCT in Zambia

Abstract

Background

Families play a critical role in adolescent management of chronic illnesses; however, HIV care and treatment programs rarely engage older adolescents and young adults (AYA) with their family caregivers. Innovative approaches are needed though, as AYA consistently have worse HIV-related outcomes than adults and lag on all the UNAIDS 95-95-95 targets. To address this gap, a family-based group intervention called Family Connections, was pilot tested for feasibility and acceptability in Zambia from April to November 2016. Family Connections consists of 10 group sessions every other week, for AYAs and their caregivers, to improve HIV disease management, decrease HIV stigma, and to promote health behaviors. Building on the successfully pilot, this study evaluates the intervention’s impact on AYA’s HIV self-management.

Methods

This study is a parallel-group, two-arm, pair-matched cluster randomized controlled trial. The clusters are the health facilities where AYAs receive their HIV care. A total of 800 participants, comprising 400 AYA and 400 caregivers from 20 clinics (10 intervention, 10 control), will be enrolled. AYA-caregiver pairs will be included if the AYA has a detectable viral load or high self-stigma at baseline. Clinics in a matched-pair will be randomized, after completion of enrollment at both sites, to either the 6-month intervention or the standard of care comparison arm. The primary outcome, assessed immediately post-intervention, is AYA’s HIV self-management, defined as a composite score of undetectable viral load and low self-stigma.

Discussion

Family Connections aims to equip AYAs and their caregivers with skills that enhance AYA’s HIV self-management and lessen caregiver burden. If successful, this intervention will provide a valuable approach for AYA, a group for whom there are limited proven effective interventions.

Trial registration

National Clinical Trial (NCT) Identifier Number: NCT05358795. Registered 2022-05-02.

Administrative information

Note: the numbers in curly brackets in this protocol refer to SPIRIT checklist item numbers. The order of the items has been modified to group similar items (see http://www.equator-network.org/reporting-guidelines/spirit-2013-statement-defining-standard-protocol-items-for-clinical-trials/).

Title {1}	A protocol for the Family Connections cluster RCT in Zambia: Impact of a youth and caregiver intervention on HIV self-management among youth living with HIV (ages 15–21)
Trial registration {2a and 2b}	National Clinical Trial (NCT) Identified Number: NCT05358795
Protocol version {3}	Protocol Version 7: 2024–08–20
Funding {4}	Grant Number: 1R01MH121157 Funded by: NIMH
Author details {5a}	1 Johns Hopkins Bloomberg School of Public Health, Department of International Health, Baltimore, USA 2 National Health Research & Training Institute, Ndola, Zambia 3 Independent consultant, TLP Consulting 4 Independent consultant, formerly FHI360
Name and contact information for the trial sponsor {5b}	Julie Denison, PhD, MHS Johns Hopkins Bloomberg School of Public Health 615 N. Wolfe Street Baltimore, Maryland 21,205
Role of sponsor {5c}	The principal investigator (PI) is responsible for study design; collection, management, analysis, and interpretation of data; writing of the report; and the decision to submit the report for publication. The PI is the ultimate authority over these activities and works in collaboration with the study team and the funder, NIMH. There is no conflict of interest.

Introduction

Background and rationale {6a}

This study will evaluate the impact of Family Connections, a family-based group intervention, on achieving undetectable viral load (VL) and lower self-stigma scores among older adolescents and young adults (AYA) ages 15 to 21 years. Families play a critical role in adolescent management of chronic illnesses; however, HIV care and treatment programs rarely engage older adolescents and young adults with their family caregivers. Innovative approaches are needed though, as AYA consistently experience worse HIV-related outcomes than adults and lag on all the UNAIDS 95-95-95 targets [1]. This challenge is particularly present in the World Health Organization (WHO) African Region, where an estimated 90% of the world’s 1.7 million adolescents living with HIV reside [2] and most HIV clinics are adult-oriented and lack programs that focus on youth.

Despite this reality, there have been no systematic efforts to implement or assess the impact of family caregiver-based group interventions to improve HIV self-management among older AYA living with HIV. There is a growing body of literature from East and Southern Africa, including Zambia, calling for interventions to involve families in supporting AYAs living with HIV to self-manage their treatment [3–8] and to promote family involvement as the AYA transitions into adulthood and adult HIV services [9–12]. Data also supports peer group interventions in reducing HIV-related stigma among adults [13] and in supporting adolescent ART adherence [14]. Given that HIV is a highly stigmatized chronic illness, peer groups provide a safe environment in which to share experiences and learn from others who understand the day-to-day reality of living with HIV [15]. Despite these data, few family-based interventions exist that specifically address the needs of older AYAs aged 15‒21 years.

In response to this gap, we pilot tested the feasibility and acceptability of a youth/caregiver intervention for AYA living with HIV called Family Connections. The pilot test consisted of a small randomized controlled trial with 50 adolescent/caregiver pairs (n = 100) randomly assigned to the intervention or comparison group in two HIV clinics in Ndola, Zambia (1R34MH 10526401). The intervention consisted of 10 in-person group meetings, held separately for adolescents and caregivers, over 6 months. We measured adolescents’ adherence to antiretroviral therapy (ART) using medical event monitoring systems caps and assessed adolescent viral load at baseline and endline. We also measured stigma-by-association and caregiver burden among the enrolled caregivers.

We found that Family Connections was highly feasible and acceptable [16]. Of the 24 intervention pairs, 88% attended eight or more of the 10 intervention sessions, and most adolescents (96%), and all caregivers would recommend the intervention to their peers. Although not powered to find differences in VL failure or ART adherence, the pilot data showed a signal for efficacy in reducing adolescents’ HIV-related feelings of worthlessness (intervention: 54 to 22% vs. control: 38 to 35%, p = 0.06) and shame (intervention: 58 to 30% vs. control: 54 to 58%, p = 0.07), and reducing caregiver burden (mean scores: intervention: 0.16 to − 0.25 vs. control: − 0.15 to − 0.25, p = 0.08). This protocol builds upon the successes and lessons learned during the pilot study to evaluate the impact of Family Connections as a group-based intervention that may shift care practices from a focus on individuals to one that incorporates family caregivers to support healthy HIV treatment and prevention behaviors among AYA living with HIV.

Objectives {7}

The overall objective of the cluster RCT is to evaluate the impact of Family Connections, a family-based group intervention, on achieving HIV self-management defined using a composite variable, as having an undetectable viral load (VL < 20 copies/mL) and low self-stigma (score of 1 or less on a three-item scale) among AYA ages 15 to 21 years. By increasing social and family support and decreasing self-stigma, we expect that youth will improve their ART adherence resulting in an undetectable VL. Specific aims include:

• AIM 1: Assess the impact of Family Connections on achieving HIV self-management defined as having an undetectable viral load (VL <20 copies/mL) and low self-stigma (score of 1 or less) among AYA. A cluster randomized controlled trial will be conducted comparing 200 pairs of AYA and caregivers at 10 intervention health facilities versus 200 pairs at 10 control health facilities (n=800 total: 400 AYA, 400 caregivers).

• AIM 2: Assess the impact of Family Connections on caregiver burden (e.g., How often do you feel stressed between caring for your adolescent and trying to meet other responsibilities for your family or work?) and social support among the 200 AYA caregiver participants in the intervention arm versus the 200 AYA caregiver participants in the control arm (n=400).

• AIM 3: Examine if the impact of Family Connections on AYAs’ HIV self-management is moderated by developmental differences among youth assessed through measures of cognitive functioning, executive functioning, impulse control, and emotional regulation.

Trial design {8}

The Family Connections study is a parallel group, two-arm, pair-matched cluster randomized superiority control trial with 1:1 allocation between intervention and control arms. Clusters are the clinics where AYAs receive their HIV care. A cluster design was chosen after a pilot study revealed evidence of contamination between intervention and control groups despite measures taken to reduce it [17]. Participants consist of dyads of AYAs living with HIV and their parents or caregivers. The clusters are pair matched to improve balance between the two study arms. The superiority framework was selected to compare the standard of care at each clinic to the Family Connections intervention on an AYA’s ability to self-manage their HIV. The standard of care at each clinic has core components based on the Zambia Ministry of Health consolidated guidelines [18], with additional components that vary and may include adolescent-friendly spaces, youth group programs, and youth outreach programs. These clinic features were not fixed when clinic pairs were matched. The various programs a clinic offers are documented during baseline data collection and will be assessed as potential confounding effects. The allocation of the intervention is masked to all during recruitment and enrollment at a matched-pair of clinics. Once enrollment is completed within a matched-pair, the allocation is unmasked (Fig. 1).

Fig. 1.

Participant timeline from study enrollment to study end

Methods: participants, interventions, and outcomes

Study setting {9}

The study is recruiting participants from 20 health facilities providing ART services in the Copperbelt Province of Zambia, which has one of the highest prevalence of HIV in the country at 11.9% (CI 8.6–15.2) [19]. Recruitment is occurring in urban health facilities in five districts, Ndola, Kitwe, Chingola, Mufulira, and Luanshya.

Eligibility criteria {10}

Health facilities were included in the sampling frame if they were public facilities in the Copperbelt Province of Zambia, within approximately 2 h of the city of Ndola, offered ART services, and had at least 60 AYAs active on ART as of August 2023. Health facilities include all levels of facilities, from primary care clinic through referral hospitals. Specialty clinics and hospitals, for example eye clinics and pediatric hospitals, were excluded.

Eligibility for participants includes criteria for both the AYA and their caregiver. Eligibility is determined through a review of the AYA’s medical records first. To be eligible, AYAs must be between 15 and 21 years of age, HIV positive, and aware of their status according to the health facility staff, on ART for at least 6 months, and on first-line ART. After the initial medical record review for eligibility, additional criterion is checked during recruitment and include that the AYA should live within 30 min of the clinic by personal transportation, able to speak English or Bemba, available to attend 10 sessions on Saturdays over a period of 6 months and being available for the next 14 months for study visits at midline (~ 6 months after enrollment) and endline (~ 12 months after enrollment).

Eligible AYA are asked to bring a caregiver on enrollment day who is at least 22 years of age or older. Additional eligibility criteria for the caregiver are checked before enrollment and include caring for the AYA who is enrolling in the study, living within 30 min of the clinic by personal transportation, able to speak English or Bemba, available to attend 10 sessions on Saturdays over a period of 6 months and being available for the next 14 months for study visits at midline and endline. Caregivers are excluded if they work at the study clinic. Being a biological parent or family member is not an eligibility criterion. Both the AYA and caregiver must be eligible to continue in the study. All enrolled participants who have completed the Family Connections intervention and midline data collection are then eligible for participation in the in-depth interviews (IDI) of whom a subset will be selected.

Who will take informed consent? {26a}

Trained study staff seek written informed consent from each participant, both AYA and caregiver, separately. Consent forms are paper based and written at a Flesch-Kincaid grade level of eighth grade or lower and then translated into Bemba, the main language spoken in the Copperbelt Province. All consent forms are back translated from Bemba to English by an independent translator and consistency is checked and corrections made as needed before use during enrollment.

The first step of the informed consent process is the study staff asks the participant’s preferred language, Bemba or English, and then reads the informed consent forms out loud with the potential participant and ask questions about risks and benefits to ensure comprehension. A copy of the informed consent form in the preferred language of the participant is offered to each participant to follow along. Informed consent forms explain that only AYA/caregiver pairs whose AYA has a detectable viral load or screens positive for self-stigma at baseline data collection move into the next phase of the study. Paper copies of signed informed consent forms are kept in a locked cabinet separate from any enrollment or data collection materials.

In the event a potential participant is illiterate, the study team will engage an impartial witness during the consent process and obtains a thumb print from the participant. The impartial witness is a health facility staff member who is not directly involved in the study and is literate. The witness is asked to watch the consent process and sign the consent form to indicate that informed consent was conducted appropriately.

For youth younger than 18, the study team obtains written individual parental/guardian permission from the parent or legal guardian of the AYA as required by the Zambian Health Services Research Act of 2013. After parental/guardian permission is obtained, written assent is sought from minor AYAs. The assent forms are read to the participant and the interviewer asks questions to ensure comprehension.

Caregivers who engage in the research as a participant must provide their own informed consent. In many cases, the person who provides parental/guardian permission is the same caregiver the youth invites to participate in the study. In some cases, the invited caregiver is different. In either situation, the caregiver who will participate in the group meetings will undergo informed written consent processes.

In cases when a consented caregiver is no longer able to attend intervention sessions due to prolonged health issues, a summarized consent form will be used for proxy caregivers who attend the intervention. The proxy caregiver is allowed to join the intervention sessions if they fulfill the above eligibility criteria and provide informed consent. Proxy caregivers may only participate in the group sessions and post-session surveys, they do not participate in midline and endline data collection points. The consent form for a proxy caregiver details that the caregiver will join a family connections intervention group without participating in baseline/midline or endline surveys.

Additional consent provisions for collection and use of participant data and biological specimens {26b}

The last section of the parental/guardian permission, AYA informed consent, and AYA assent forms is a series of questions regarding specific procedures of the study. There are eight questions that the participant (and their parent/guardian if under 18 years of age) must agree with to be enrolled in the study. These include specific questions about survey participation, blood draws, urine collections, study contact, data collection from medical record, viral load results to be placed in medical record, follow-up data collection points, and attendance at group sessions if assigned to the intervention arm. There are also three questions that are not required for participation but are included as optional. These include permission to contact the caregiver or adult if study staff are unable to reach the AYA, participation in an in-depth interview in the study, and permission for the research team to contact them regarding future research opportunities. The last section of the caregiver consent contains all the above questions for AYAs excluding the questions about biological sampling (blood and urine collection and testing).

Interventions

Explanation for the choice of comparators {6b}

The study will compare the intervention arm, consisting of 10 youth and caregiver group meetings, to the control arm which consists of the standard of care for AYALHIV. Standard of care in this settings is based on the Zambia Ministry of Health consolidated guidelines (2023) that includes the following core services: routine clinical care for HIV treatment including laboratory testing (CD4, VL tests); adherence counseling, with enhanced adherence counseling for clients with unsuppressed viral loads; multi-month dispensing of ART including differentiated service delivery for stable clients; screening and treatment of STIs; prevention and screening for opportunistic infections; and clinical monitoring of kidney and liver function [18]. Some clinics also offer monthly youth group meetings and youth friendly strategies for young people living with HIV.

Intervention description {11a}

The intervention consists of ten group sessions for AYAs and their caregivers. Caregivers and AYA are separate for the start of sessions and then brought together at the end of the sessions, to share information and skills learned. The ten group sessions are held, approximately twice per month, over 5 to 6 months (Fig 1). Sessions are held on Saturday to increase caregiver participation and to avoid conflicting with AYA’s school commitments. Trained facilitators deliver the intervention at the health facilities. Health facility staff are available as technical experts to answer clinical questions. Group session topics range from basic information about HIV to positive prevention and handling stigma or discrimination (see Table 1). Participants who attend intervention sessions receive an attendance certificate.

Table 1.

Family Connections Intervention sessions

Session number	Topic
1	Family Connections for success
2	Understanding HIV
3	Treatment and adherence
4	Communication skills
5	Growing and changing
6	Exploring your feelings
7	Nutrition and self-care
8	Sexual and reproductive health
9	Stigma, discrimination, and disclosure
10	Stepping into the future

The intervention group sessions aim to (1) improve understanding of HIV among AYA and caregivers; (2) help AYA develop strategies for healthy living including adherence to ART; (3) build the capacity of AYA to make informed decisions about their sexual and reproductive health; (4) build the capacity of caregivers to support AYA; and (5) help AYA develop life skills to communicate their HIV diagnosis effectively and to plan for their futures. The theoretical frameworks guiding the Family Connections intervention and evaluation are the positive youth development (PYD) framework and social cognitive theory (SCT) [20–25]. The PYD framework and SCT complement each other and have guided the refinement and prior pilot testing of the Family Connections intervention, making them fitting tools to explain and predict changes targeted in this impact study.

Criteria for discontinuing or modifying allocated interventions {11b}

AYA and caregivers in the intervention health facilities may voluntarily withdraw from the Family Connections group sessions at any point in time while continuing their participation in the study. They may also withdraw from the study entirely. If a participant discontinues from the intervention but not from the study, the remaining study procedures will be completed as planned and reasons for withdrawal from the group sessions will be recorded.

Strategies to improve adherence to interventions {11c}

Reminder calls to intervention participants are completed every week before and intervention session. Intervention adherence is assessed with facilitator logs and attendance logs, which are administered and completed at each intervention session. If a participant fails to return to the health facility for a scheduled intervention, the study team will attempt to contact the participant. The study team makes every effort to regain contact with the participant (where possible, three telephone calls and, if necessary, a home visit). Contact attempts will be documented in the study files. Should the participant continue to be unreachable for intervention session and data collection visits, they are considered lost to follow-up.

Relevant concomitant care permitted or prohibited during the trial {11d}

All study participants, in both study arms, receive standard services currently available to AYA living with HIV in these public health care facilities. Participants may also be a part of other programs at the clinic, many targeted towards adolescents living with HIV. The surveys for baseline, midline, and endline include questions about other services received at the clinic.

Provisions for post-trial care {30}

Harm from trial participation is expected to be rare but potentially severe (e.g., unintended disclosure) and will be addressed following guidance from the IRB in Zambia.

Outcomes {12}

Primary outcome

The primary outcome will assess the change in AYAs from baseline to midline (post intervention) on a composite measure of HIV self-management consisting of an AYA’s viral load and their reported self-stigma [26, 27] (Fig. 1; see Table 2). The self-stigma scale includes three items and asks because an AYA is living with HIV do they feel (1) ashamed, (2) guilty, and (3) sometimes worthless. Each question has dichotomous answer choices of agree or disagree. The overall self-stigma scale is the sum of the three individual items and dichotomized as having low self-stigma (< 2 on the 3-point scale) or high self-stigma (2 or more on the 3-point scale). The HIV management outcome is a dichotomous outcome of achieving an undetectable viral load (< 20 copies/mL) and a low level (< 2) of self-stigma [26, 27]. Failing to achieve HIV management is defined as having a detectable viral load (≥ 20 copies/mL) and/or having high self-stigma (2 or more). The comparison is between AYAs in the Family Connections intervention and the control condition on the impact on this outcome.

Table 2.

Outcome measures for primary and secondary outcomes and analyses

Outcome	Measure	Description	Validation context	Internal consistency	Examples of use in Sub-Saharan Africa
AYA primary composite outcome: HIV Self-Management
Undetectable viral load (VL)	Molecular VL test, < 20 copies/mL	Viral load test result	NA	NA	Standard of care
Internalized self-stigma	Adapted from Internalized AIDS Related Stigma Scale (IA-RSS) [26, 27]	Adapted scale, 3-item, agree/disagree responses. A high score of 2 or more will be considered a screen positive	Adults in US, South Africa, Eswatini	Cronbach’s α = 0.74; α = 0.73 [17]	Study team’s prior research with adolescents and young adults in Zambia [17, 28]
Caregiver secondary outcomes
Caregiver burden	Zarit Burden Interview [29]	Adapted 12-item [30]; 5-point Likert-scales; freq. of challenges in caregiving	Adults in South Africa [30]; Adults in England [31]	Cronbach’s α = 0.82; [30] α = 0.85–0.89 [31]	Adults in South Africa [32]; NIH-funded pilot study in Ndola, Zambia
Social support	Medical Outcomes Study Social Support Survey (MOS-SSS) [33]	20-item (4 sub-scales), 5-point Likert scales, frequency of someone providing different types of support	U.S. adults with hypertension, diabetes, coronary heart disease, & depression [33]	Cronbach’s α = 0.91–0.96 for sub-scales, 0.97 overall	HIV + Kenyan youth, 15–19 years [34]; adults living with and without HIV Kenya [35]); Ghana/South Africa adults [36–42]
Stigma by association	Adapted from published studies [27, 43, 44]	Adapted scale, 4-dichotomous items; stigma caring for an HIV + AYA	––-	––-	NIH-funded pilot study in Ndola, Zambia
AYA secondary outcomes: developmental moderating factors
Cognitive function	PROMIS Pediatric Cognitive Function [45]	Abbreviated 7-item, 5-point Likert scales; frequency of difficulties in cognitive abilities/everyday tasks, past 4 weeks	U.S. adults [46]	Cronbach’s α = 0.94	––
Emotion regulation	Difficulties in Emotion Regulation Scale (DERS-SF) [47, 48]	Abbrev.18-item (6-subscales); 5-point Likert scales; difficulties identifying/managing emotions	U.S. adolescents, 13–17 yrs. [49]	Cronbach’s α = 0.93	Adults in South Africa [50]
Executive function	Flanker Task	Interactive assessment, participant selects button corresponding to displayed arrow	U.S. adults [51]	NA	HIV + youth in Zambia [52]
Impulsivity	Balloon Analogue Risk Task	Interactive assessment, participant inflates virtual balloon to receive virtual prize	U.S. young adults, 18–25 years [53] U.S. inner-city youth from 9th–12th grade [54, 55]	NA	Adolescent girls and young women in South Africa [56]

Secondary outcomes

Secondary outcomes will assess the change from baseline to midline in caregivers’ burden, social support, and stigma by association among the intervention arm with the control arm. Further, secondary analyses will assess the moderating influence of AYA developmental factors on the impact of Family Connections on the primary composite outcome of achieving low self-stigma and undetectable viral load.

Sample size {14}

The outcome is measured at the individual level and randomization occurs at the cluster level. The number of clusters, 20, was set in advance. The trial’s primary endpoint is measured at midline and will be based on the proportion of AYA achieving HIV self-management (composite outcome) comparing the intervention and control participants. All AYAs that continue in the cluster RCT portion of the trial past enrollment and baseline data collection will have a detectable viral load and/or high-self-stigma. During a pilot study of the intervention, a 24% drop in detectable viral load was observed among the AYAs that attended the session. With these results and 20 clinics (cluster), we will have 91% power to detect a 15% difference between the intervention and control group at midline in the composite outcome (see Table 3). This estimated power is based on a two-sided comparison, an alpha = 5%, and an intra-class correlation (ICC) = 5% to account for the clustered design. The sample size was determined with a non-matching study design, and clinic matching may improve power if effective. Variability in ICC is accounted for in the sample size, for example, the same change in the composite outcome from 95 to 80% having a detectable VL and/or high self-stigma, can be detected with 86% power with an ICC of 7%. If there is greater variability in the control group, this sample size has 82% power to detect a 15% difference from 90 to 75% in the control and intervention group with the other assumptions maintained (including the ICC of 5%). The ICC will be estimated from the study data and any loss in power that occurs will be acknowledged.

Table 3.

Power estimates for different assumptions of proportions of participants with detectable VL and/or high self-stigma at midline and ICC values given 10 clinics per arm

Detectable VL & high self-stigma		ICC
Control	Intervention	5%	6%	7%
95%	80%	92%	89%	86%
90%	75%	82%	79%	75%
80%	62%	83%	79%	75%

Recruitment {15}

Recruitment begins with health facility staff generating a computerized list of age-eligible patients who are active on ART from the Zambian SmartCare electronic medical record system. Study teams add in any paper-based forms that are not yet in the electronic system. The list is then randomized to recruit individuals in a random order. Health facility staff retain the original randomized list that contains the identifiers while study staff keep a randomized list that only contains ART numbers.

Recruitment begins with the participants on the list whose last routine viral load test result was detectable or missing. If there is an insufficient number of participants recruited from this group, recruitment process continues with individuals on the list whose most recent routine viral load test was undetectable. This process allows youth most vulnerable to negative health and wellbeing outcomes to be enrolled in the study. Individuals on the recruitment list are screened against specific eligibility criteria for the study, including the AYA knowledge of their HIV status based on information provided by health facility staff.

Once identified as a potential participant, the health facility staff make the first contact with an AYA to introduce the study using a standardized script. Health facility staff make up to three contact attempts before counting that youth as not reachable. The contact attempts are spread throughout 1 week and include at least one home visit. Zambian law defines youth who are 17 and younger as minors when participating in research. For all minor AYAs, contact begins with their next of kin listed in their medical record. Health facility staff verify that the AYA knows their HIV status and request permission to contact the AYA from the next of kin. If awareness of the AYA’s status is not confirmed by the next of kin, then the health facility staff do not attempt to contact the minor.

After contacting the youth, health facility staff introduce them to the study and ask if they would like to learn more from a study team member. If yes, they are connected to a member of the study team who then provides more information about the study, confirms AYA eligibility, and invites the youth to come to the health facility with a caregiver of their choice on designated enrollment days. The AYA’s name and contact information is collected to ensure that only potential participants identified through the sampling process attend enrollment day. Caregiver selection is done by the youth participant, and caregivers are not generally contacted by the study team unless the AYA explicitly asks for assistance with introducing the study to their caregiver. The study team has had success in maintaining confidentiality while recruiting AYA and their caregivers using similar methods in previous research in Zambia, including our previous pilot study [16]. AYA who have a positive composite score (a detectable VL > 20 copies/mL, and/or high stigma score = + 2) continue into the cRCT portion of the research. Those who did not meet the composite score requirements for continuation were informed of their VL and that their participation in the study had ended.

Assignment of interventions: allocation

Sequence generation {16a}

Health facilities are the clusters and the unit of randomization. Those health facilities meeting the eligibility criteria were pair-matched to increase power to detect an effect of the intervention. First, an objective and replicable approach — coarsened exact matching — was used to create strata where similar facilities could be matched into pairs [57]. Facilities were stratified by number of AYA 15–21 years of age active on ART and the type of health facility (either primary health clinic or hospital). The sampling frame had an even number of hospitals and an odd number of clinics; the smallest two clinics and largest were dropped prior to matching to reach the clinic sample size. The result was three strata for both clinics and hospitals. A random number generator was used to pair-match facilities within each stratum. Once pair-matched, clinic pairs were randomized to generate the enrollment sequence of clinics into the study. The random allocation of the facilities in a pair to the two arms, intervention and control, was done using the ralloc command in Stata using set seed for reproducibility.

Concealment mechanism {16b}

The statistician that generated the intervention allocation assignments is not involved in the day-to-day operations of the study and is the only person with access to the allocation among the study team. They are based in Baltimore, USA.

Implementation {16c}

The biostatistician generates and maintains the intervention allocation for all study clinics. The study team based in Ndola, Zambia, is responsible for enrolling participants and is not aware of intervention allocation during enrollment. Once enrollment for a pair of clinics is complete, the study team in Zambia requests the allocation for the completed pair from the statistician.

Assignment of interventions: blinding

Who will be blinded {17a}

Everyone involved in recruitment and enrollment will be blinded to the intervention allocation until recruitment is complete at one pair of clinics. After recruitment ends, the biostatistician reveals the intervention allocation to the study team. The study team members then inform the clinics which study arm they have been assigned to and call all participants that are continuing in the trial to inform them if their clinic is assigned to the intervention or control arm. The in-person, group-based nature of the intervention and its delivery at the health facility precludes our ability to blind study staff, participants, and health facility staff to intervention assignment after recruitment is complete. The data will be analyzed without the analyst being blinded to the allocation.

Procedure for unblinding if needed {17b}

There is no provision for unblinding as all parties are informed of the allocation after enrollment into the study and before the intervention takes place.

Data collection and management

Plans for assessment and collection of outcomes {18a}

Data collected at study visits include questionnaires for both AYAs and caregivers as well as blood and urine collection for AYAs. Questionnaires are administered in a private location at the clinic by a trained member of the study staff. All data collector study staff underwent Good Clinical Practice training and training on the specific data collection instruments used in the study. Each questionnaire was reviewed question by question in both English and Bemba to ensure that all study staff understood the question premise and would be able to help explain the question to participants if they were not clear what was being asked. All data collector study staff participated in multiple rounds of pilot testing of the questionnaires to gain familiarity and resolve any questions about the instrument.

At baseline, data collection questionnaires are administered to all AYA and caregivers after enrollment at the study health facility. The questionnaires take approximately 1 h to complete. The questionnaires for the AYAs cover the following topics: basic demographic information including sex, age, and education level; family relationships; developmental factors; HIV history and stigma; relationship status and sexual and reproductive health history; trauma, adverse childhood experiences, and resilience; and other factors known to affect HIV self-management including food insecurity, alcohol use, anxiety, and depression. The questionnaire for caregivers covers basic demographic information, family relationships, perceptions of the AYA’s developmental status, HIV resilience, caregiver burden, social support, and stigma by association.

As a part of the AYA questionnaire, participants complete two interactive assessments to quantify key aspects of cognitive development: executive function and impulsivity. The interactive assessments are conducted using Presentation® software (Version 23.0, Neurobehavioral Systems, Inc., Berkeley, CA). Partway through the questionnaire, participants complete the assessments using the tablet’s touch screen. Executive function is assessed using the Flanker Task [51], which measures executive function by assessing inhibitory control, or the ability to ignore irrelevant stimuli and focus on the relevant stimulus. The Flanker task is a widely used assessment of executive function that is included in the NIH’s battery of cognitive tests and is currently being used in research with youth living with HIV in Zambia [52]. Impulsivity is assessed using the Balloon Analog Risk Task, or BART [53] which correlates with real-world impulsive behavior among U.S. adolescents [58] and with substance use among adults [55, 59]. This is a widely used assessment of impulsivity and has been used with adolescent girls and young women (17–26 years) in South Africa [56].

After baseline, eligible AYA and their caregivers continue into the cluster RCT portion of the research. Participants who are not eligible to continue are informed about their VL result and thanked for their time. Midline and endline data collection occurs only for the AYA/caregiver pairs whose AYA had at least one of the composite outcome characteristics (detectable viral load and/or high self-stigma) at baseline and continued in the study. Midline and endline questionnaires for AYA include family relationships; developmental factors; emotional regulation, HIV history and stigma; alcohol and drug use; depression and anxiety; trauma and resilience. Midline and endline questionnaires for caregivers include family relationships, perceptions of the AYA’s developmental status, resilience, caregiver burden, social support, stigma by association, and depression. At intervention sites, AYA and caregivers are also asked about intervention acceptability and changes in HIV-related knowledge and behavior at midline and endline.

Plans to promote participant retention and complete follow-up {18b}

As part of the enrollment data collection process, the study team collects contact information, including home addresses and phone numbers, from the AYAs and caregivers. The team also records the preferred methods of contact including days and hours. All identifying information is kept in locked cabinets at the study site with access limited to the core study team only. Study staff contact all participants who do not attend scheduled visits, at first by phone followed by house visits as necessary, using the routine care procedures that health facility staff follow when conducting follow-up and home visits for AYAs. The study team also contacts participants before their scheduled data collection day and group meetings to remind them of the scheduled time and date. For each study visit, whether for data collection or intervention participation, participants received 200.00 Kwacha in cash as compensation for time and travel expenses. Participants also receive a bottle of water during enrollment and a snack and refreshment during intervention visits.

Data management {19}

All study data is collected and maintained on REDCap electronic data capture tools hosted on servers at JHU [60, 61]. During the recruitment phase, eligibility data are collected on paper and entered in the REDCap database with double data entry by independent data entrants. Double entries are compared, and any discrepancies are resolved by looking at the source document where the data was originally recorded.

The questionnaire data collected during enrollment, baseline, and endline is captured using the REDCap mobile application on study tablets and synced to the servers on a weekly basis. Questionnaires were built into the REDCap platform and include several data quality checks embedded in the data collection form itself. The data entry forms contain range checks on numeric fields, skip patterns, warnings when dates are out of range, and alerts for missing data.

Data from the interactive assessments are also collected on study tablets using Presentation® software (Version 23.0, Neurobehavioral Systems, Inc., Berkeley, CA, www.neurobs.com) and uploaded to REDCap servers. Results of the urine test are available on the same day as data collection and recorded in the REDCap mobile application along with a photo of the test cassette for verifications of results. Viral load results are provided by the Zambian MOH standard laboratory system. The individual file of each VL result is saved to a secure server by the laboratory technician, study staff then enter the VL results into individual records in the REDCap database.

Data quality reports are routinely run through the REDCap web interface and queries are generated and resolved for any anomalies that are found.

Confidentiality {27}

Procedures are put in place to maintain confidentiality, and every effort is made to protect the privacy and confidentiality of the study participants. Only study staff trained on research ethics, good clinical practice in social and behavioral research, and the importance of informed consent and confidentiality interact with study participants and collect data. The training is conducted by investigators with support from the principal investigator and other study team members. Participant privacy is maintained by conducting informed consent procedures and collecting data in a private room or separate area. Information AYA share with the study team is not shared with the caregiver and vice versa. These procedures maintain privacy and confidentiality for both the AYA and caregivers in the study.

Once enrolled, all participants are assigned a unique participant identification number (PID). All identifying information is kept securely in locked files and on password-protected computers with access limited to the study team on a need-to-know basis. The linking document between the PID and personally identifying information are separate paper-based forms and are securely stored accessible by senior study staff only. The VL blood samples are collected and tracked using standard MOH forms that include personal identifiers, such as name, medical record number and date of birth, as needed by the laboratory to return the results to the participant’s medical record. The samples and the forms are sent directly to the laboratory for processing and do not contain any linkage to their study PID. The laboratory provides the study team with a copy of the VL results that have personal identifiers, through a secure server. The study team matches the result to a participant and enters the data into the study database with the PID. Hard copies of any paper forms and data collection equipment such as tablets are stored in locked spaces when not in use. For electronic data collection forms, all data collection devices (tablets and computers) are encrypted and password protected, and data is saved on a secure server. Any downloaded data is stored on password-protected computers only accessible to study staff. To ensure confidentiality, data with personally identifiable information are stored in separate datasets with more limited study staff access. Data is erased from tablets used for data collection after data have been successfully transferred to the secure server, data has been checked for quality and completeness, and any questions or quality issues have been resolved.

Plans for collection, laboratory evaluation, and storage of biological specimens for genetic or molecular analysis in this trial/future use {33}

Whole blood and a urine sample are collected at every data collection point for AYAs.

Blood is collected from the trained phlebotomists at each health facility in accordance with MOH guidelines for VL testing. Blood samples are transported and securely stored at NHRTI after having been centrifuged and extracting the plasma from the whole blood. Viral load testing will use the Roche HIV Test platforms (or equivalent alternative) with a limit of detection of 20 copies/mL. All VL samples with greater than 1000 copies/mL are stored for future ART resistance testing. VL testing will be performed according to SOPs and appropriate quality control measures as practiced by NHRTI. Laboratory source data will be retained by the laboratory according to local requirements. All local bio-hazardous materials regulations are followed to ensure blood samples are collected, transported, and stored in an appropriate repository and are destroyed after the viral load tests are completed and verified. Study staff are trained on safe biological specimen handling. All VL test results will be placed in patient medical files in accordance with Zambian MOH guidelines [18].

A point-of-care (POC) test using urine to measure treatment adherence is used on all AYA whose treatment regimen contains Tenofovir at each data collection study visit. The POC test developed by ABBOTT® has high specificity (95–100%) and sensitivity (96–99%) for detecting Tenofovir 24 h after taken when compared to the gold standard, urine LC–MS/MS TFV test [62–64]. Participants are asked supply a urine sample in a cup and then the facility’s phlebotomist conducts the POC test according to the manufacturer’s SOP. In the case of a test kit stock out, the urine is transferred to two 1.5-mL cryogenic vials and stored at − 80 °C in freezers at NHRTI. Frozen samples are thawed and brought to 37 °C before performing the POC test.

Statistical methods

Statistical methods for primary and secondary outcomes {20a}

The primary outcome analysis will be a modified intent-to-treat analysis such that only enrolled participants with a detectable viral load and/or high self-stigma at baseline will be included in the primary outcome. We will use generalized linear mixed models to compare the composite outcome of HIV management between the two study arms at the 6-month data collection time point, immediately after the intervention is completed. Random effects will be included to account for within-participant repeated measures across baseline and midline timepoints, and clustering by health facilities. The models will adjust for non-time varying baseline covariates that are believed to be associated with the outcome.

Analyses of all caregivers and AYA secondary outcomes generated by various questionnaires will include summary statistics for categorial or continuous variables as appropriate. Comparisons of scores between study arms at a cluster level will use t-tests with suitable transformations if needed. Generalized linear models will be used to evaluate associations with covariates of interest, adjusting for clustering by clinic.

Interim analyses {21b}

No interim analysis is planned.

Methods for additional analyses (e.g., subgroup analyses) {20b}

This research will provide sufficient data for additional analyses that will contribute to our generalizable knowledge of AYA living with HIV and their caregivers in an intervention context. Examples of sub-analyses include:

• Subgroup analyses by gender and age to understand differential intervention impact,

• Dose–response among the intervention participants,

• Examining the relationship between how a young person learned about living with HIV and their HIV VL status,

• Subgroup analysis examining factors related to AYA with a detectable VL versus those with an undetectable viral load, at baseline,

• The long-term impact of the intervention on AYAs (measured at endline).

Methods in analysis to handle protocol non-adherence and any statistical methods to handle missing data {20c}

The primary and secondary outcomes will use an as-randomized analysis approach. The effects of protocol non-adherence will be explored through sensitivity analyses including models of the dose–response of the intervention to the outcome. Missing data is anticipated to be minimal due to the quality assurance measures in the trial. If any missing data is found, excluding missing observations, multiple imputation will be used to fill in missing data with sensitivity analyses to assess the influence of imputed missingness.

Plans to give access to the full protocol, participant-level data, and statistical code {31c}

The trial protocol is available at ClinicalTrials.gov (NCT05358795). De-identified data will be placed in a repository following guidance from both the Zambian National Health Research Authority, and the funding agency, NIMH.

Oversight and monitoring

Composition of the coordinating center and trial steering committee {5d}

The coordinating center for the trial is the National Health Research and Training Institute in Ndola, Zambia. The coordinating center includes a country-level Principal Investigator (PI) and two co-PIs, a behavioral scientist and a clinician, who, under the guidance of the study PI (Denison), oversee the trial’s ethical implementation. Responsibilities include management of the study team (intervention facilitators and data collectors) and the ethical collection and management of the trial data (including blood samples, urine samples, and survey data). This team meets weekly within NHRTI and weekly with the JHU PI and co-investigator. The study also has a youth advisory board, consisting of 6–12 youth who provide guidance and insight into trial implementation and will provide guidance regarding interpretation and dissemination of trial results.

Composition of the data monitoring committee, its role and reporting structure {21a}

The trial has an independent safety monitor who is a JHU public health faculty in the Department of Mental Health. The ISM’s role is to monitor participant safety, evaluate the progress of the study against targets and any deviations from those targets, review procedures for maintaining the confidentiality of data, and review the quality of data collection, management, and analysis. To fulfill these responsibilities, the ISM meets every 6 months with the trial team (JHU and NHRTI PI, co-PI and co-I).

Adverse event reporting and harms {22}

Adverse events are collected during all data collection points and intervention sessions by study and documented on standardized forms. A log of all adverse events is compiled and maintained, with the event description, time of occurrence, relationship to study procedures, and details on when the event was reported. Possible serious adverse events (SAEs) could occur due to accidental HIV status disclosure due to participation in the research and subsequent ramifications. Any SAEs will be reported to the IRB. All adverse events have their relationship to study procedures, including the intervention, assessed by the study Principal Investigators and the Zambia-based study team based on temporal relationship and first-hand knowledge. All adverse events occurring during the study will be documented appropriately, reported to ethical and oversight bodies, and followed to adequate resolution regardless of relationship to the research.

Frequency and plans for auditing trial conduct {23}

Study co-investigator site visits are conducted twice a year to ensure trial conduct follows the protocol and standard operating procedures. Routine study calls are also used to assess fidelity to study operations and procedures.

Plans for communicating important protocol amendments to relevant parties (e.g., trial participants, ethical committees) {25}

All protocol amendments must be approved by the ethical committees in Zambia and the USA before enacted upon, following amendment procedures. Once approved, amendment changes are also shared with Zambia’s Ministry of Health National Health Research Authority. Contact information for trial participants is maintained in a locked cabinet and will be used to reach out to participants as needed if an amendment will impact their involvement in the trial.

Dissemination plans {31a}

The study team is committed to sharing evidence and data to increase knowledge and advance programs and services for adolescents and young adults (AYA) living with HIV. This study will comply with the National Institutes of Health (NIH) Public Access Policy, which ensures that the public has access to the published results of NIH-funded research. It requires scientists to submit final peer-reviewed journal manuscripts that arise from NIH funds to the digital archive PubMed Central upon acceptance for publication. This study will also comply with the NIH Data Sharing Policy and Policy on the Dissemination of NIH-Funded Clinical Trial Information and the Clinical Trials Registration and Results Information Submission rule. As such, this trial is registered at ClinicalTrials.gov, and results information from this trial will be submitted to ClinicalTrials.gov.

We will also share the data generated by this project with study participants, the Youth Advisory Board (YAB), study clinics, community members, and relevant Ministry of Health (MOH) staff and committees through Zambia specific dissemination efforts and globally through scientific meetings and publications. The YAB will have a critical role in shaping dissemination material and approaches. We will also work with the NIH Program Officer to identify an appropriate data repository for the data collected under this award.

Discussion

The Family Connections study evaluates the impact of a family-based group intervention for adolescents and young adults (AYA) living with HIV and their caregivers. The intervention seeks to increase social and family support and decrease self-stigma among AYA, so they may improve their medication adherence and achieve an undetectable viral load and low self-stigma. Findings will fill a critical gap in available evidence-based intervention options for improving the HIV-related outcomes and wellbeing of HIV-positive AYA in East, West, and Southern Africa. Specifically, this intervention seeks to shift HIV care strategies from individual level and peer group approaches to approaches that engage caregivers in collaboration with health clinic staff in the care and treatment of AYA. Caregivers in Zambia provide a key pillar of support for AYALHIV and may also serve as a barrier, to some of the health challenges faced by AYALHIV. For example, in terms of support, adolescent girls living with HIV in a 2019 Zambian study reported that their family members were their primary source of support in addressing the multi-layered barriers to HIV care that they experienced [41]. In Zambia, informal caregivers provide the bulk of care for AYALHIV, this includes accompanying the youth to the hospital, administering medication, and providing personal care needs such as, cooking, bathing, and washing when necessary [10, 41].

A key challenge in implementation of the study is missing data in the SmartCare system where potential study participants are being drawn. This is mainly due to the power outages that are being experienced in the country that make it challenging for health staff to update the system. In some cases, this may result in potentially eligible participants not being included in the study. Study teams try to mitigate this challenge by going through paper-based records to ensure as many potential participants who meet the initial screening criteria are included on the recruitment list as possible.

Trial status

Current protocol is Version 7. Recruitment is rolling, with two clinics undergoing recruitment and enrollment at the same time. A total of 6 weeks is needed for recruitment and enrollment to be completed at a pair of clinics. Recruitment and enrollment data collection began on April 17, 2024, and is scheduled to be completed by June 30, 2025.

Abbreviations

AYA

Adolescent and young adult

AYALHIV

Adolescent and young adults living with HIV

ART

Anti-retroviral therapy

BART

Balloon analog risk test

CA

California

CI

Confidence interval

HIV

Human immunodeficiency virus

ICC

Intra-class correlation

IDI

In-depth interview

IRB

Institutional Review Board

ISM

Independent safety monitor

JHU

Johns Hopkins University

LC–MS/MS

Liquid chromatography with tandem mass spectrometry

mL

Milliliter

MOH

Ministry of Health

NCT

National Clinical Trial

NHRTI

National Health Research and Training Institute

NIH

National Institutes for Health

NIMH

National Institute of Mental Health

NIAID

National Institute of Allergy and Infectious Diseases

NCI

National Cancer Institute

NICHD

National Institute of Child Health and Development

NHLBI

National Heart, Lung, and Blood Institute

NIDA

National Institute on Drug Abuse

NIA

National Institute on Aging

FIC

Fogarty International Center

NIGMS

National Institute of General Medical Sciences

NIDDK

National Institute of Diabetes and Digestive and Kidney Diseases

OAR

Office of AIDS Research

PI

Principal investigator

PID

Participant identification number

POC

Point-of-care

PYD

Positive youth development

RCT

Randomized control trial

REDCap

Research electronic data capture

SAE

Serious adverse event

SCT

Social cognitive therapy

STI

Sexually transmitted diseases

TFV

Tenofovir

VL

Viral load

WHO

World Health Organization

YAB

Youth advisory board

Authors’ contributions {31b}

Trial design: JD, DRM; Intervention design: TP; design of research protocol: JD, DRM; design of analytical plan: BAN, JLS; preparation and conduct of surveys: JD, JLS, DRM, TM, SM, contributed to manuscript preparation: JD, GC, TM, JLS, DRM SM, BAN. All authors read and approved the final manuscript.

Funding {4}

This work was supported by the National Institute of Mental Health (NIMH) under grant 1R01MH121157; and The Johns Hopkins University Center for AIDS Research, an NIH-funded program under grant P30AI094189, which is supported by the following NIH Co-Funding and Participating Institutes and Centers: NIAID, NCI, NICHD, NHLBI, NIDA, NIMH,NIA, FIC, NIGMS, NIDDK, and OAR. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

Data availability {29}

Upon completion of data cleaning/readiness, we will make our data available to our team and main partners. The study funder (NIH-NIMH) requires that results and accomplishments of the study should be made available to the public. All mechanisms used to share data will include proper plans and safeguards for the protection of privacy, confidentiality, and security for data dissemination and reuse (e.g., all data will be thoroughly de-identified and will not be traceable to a specific study participant). Plans for archiving and long-term preservation of the data will be implemented in consultation with the study funder.

Declarations

Ethics approval and consent to participate {24}

The protocol, informed consent form(s), recruitment materials, and all participant materials were submitted to the following Institutional Review Boards (IRBs) for review and approval: Eres Converge IRB in Zambia (#2022-June-003), Johns Hopkins Bloomberg School of Public Health IRB (IRB00020325), and the National Regulatory Authority in Zambia. Approval of both the protocol and the consent form(s) was obtained before any participant was consented. All amendments to the protocol require review and approval by the IRBs before the changes are implemented to the study. Any changes to the consent form(s) are IRB approved. Written informed consent or parental permission and assent will be obtained from all participants.

Consent for publication {32}

Example consent forms can be found in supplementary material.

Competing interests {28}

The authors declare that they have no competing interests.