Aim and objectives
This project comprises three specific aims. Aim 1, which was completed during the study’s formative phase, involved identifying and randomizing eligible clusters (study clinics) and engaging community stakeholders through a Community Collaborative Board (CCB) and Youth Advisory Board (YAB) to assist in the adaptation of the YWHC and S&D reduction training program to the South African geographic and cultural context. This study protocol is based on Aims 2 and 3, the experimental phase of the study. Aim 2 is to evaluate the impact of the HPP S&D reduction training among clinic staff on the use of HIV and SRH services by AGYW, including PrEP, and staff attitudes and behaviors toward AGYW at 4- and 8-month follow-up. Aim 3 is to test the efficacy of a multilevel HIV prevention strategy that addresses structural (S&D reduction), interpersonal (peer social support), and individual (personal agency, substance use, and GBV) factors on PrEP readiness, uptake, and adherence; HIV status; and other risk-taking behaviors at 3-, 6-, and 9-month follow-up among vulnerable AGYW. We plan to recruit 900 AGYW across 12 communities in clinic catchment areas for the experimental phase of study.
Setting
We conducted feasibility assessments of potential study sites in provincial and city clinics in Tshwane, South Africa; 15 clinics were selected as the final study clinics. Each assessment included engaging the clinic manager and requesting permission to conduct the assessment. The selected clinics had (1) community ward-based outreach teams (WBOTs), (2) comprehensive SRH services, and (3) designated youth-friendly nurses and/or services. Other criteria included (a) the jurisdiction of the clinic—city or provincial, as this determines the allocation of resources and infrastructure, (b) reported average number of AGYW visiting the clinic per month, and (c) clinic location (peri-urban or semi-rural community) within the Tshwane district. As a result, the final sample consisted of 12 study clinics, 8 provincial clinics and 4 city clinics, with 3 additional clinics as backups if one or more of the selected clinics turns out not to be viable.
Data collection: adolescent girls and young women level
Recruitment
AGYW are recruited in communities that are in the study clinic’s catchment area. Recruitment is conducted through research study staff, who may also be assisted by trained WBOTs associated with the study clinics. However, screening is only conducted and performed by research study staff. Recruitment is conducted via street outreach at identified areas where AGYW are known to frequent. Street outreach is an established recruitment method that has been used successfully in previous research studies [
50‐
53], including Tshwane WHC studies to reach and recruit a high number of women most at risk of HIV. Study staff visit these hotspots regularly to establish a known presence in the community and gain rapport with community members. Recruitment also is conducted through referrals from the clinic or other staff who conduct HIV and/or pregnancy testing in the community, such as HIV and AIDS, STI, and TB (HAST) counselors.
Participant eligibility criteria include the following: (1) identify as female; (2) HIV negative status; (3) between ages 16 and 24 years; (4) have had condomless sex in the past 3 months with a male partner; (5) not currently pregnant and do not want to get pregnant within the next year; (6) interested in taking a daily pill to prevent HIV (PrEP); (7) not having previously participated in the formative phase of the study; (8) not previously or not currently participating in any other PrEP-related project or research study; (9) not previously or not currently participating in any other HIV study in Tshwane; (10) not on multidrug-resistant tuberculosis (MDR-TB) treatment; (11) lives in one of the target communities; (12) intends to stay in the Tshwane district for the next 12 months; (13) agrees to provide contact information; (14) willing to undergo rapid HIV testing; and (15) willing to undergo pregnancy testing.
Because of the eligibility requirement of an HIV-negative and nonpregnant status, interested potential participants must first consent to rapid testing for HIV and pregnancy to confirm eligibility. Additionally, co-enrollment in another HIV study is also checked and any individual who is currently enrolled is excluded.
Screening
Verbal consent is obtained to screen the potential participant and determine eligibility. Initial eligibility screening of potential participants is performed individually and in private by research study staff. After the initial screening, the next step is to confirm that the participant is HIV-negative, not pregnant, and not enrolled in another HIV study; this is done at the study clinic. Potential participants provide informed consent before conducting these tests. All participants receive a rapid HIV test and a pregnancy test. The only exception is if they have tested negative for HIV the same day and they have an acceptable proof of their HIV testing result, such as a clinic card. If a potential participant cannot be seen for the baseline appointment on the same day as initial screening, they are not tested for HIV and pregnancy until the day of their scheduled baseline appointment to minimize repeat testing. They are rescreened using the initial eligibility screening tool to ensure they meet study eligibility criteria on the day of enrollment.
Consent or assent
The intake process takes place within the compound of the study clinic, typically in an outdoor tent because of the limited physical space in clinics. Written informed consent or assent is obtained from eligible participants before study enrollment. For potential participants who are younger than 18, consent is required from their mother or a trusted adult woman at least 25 years old who may serve in loco parentis (“in place of a parent”). Parental waiver has been requested to protect a participant’s confidentiality if they are uncomfortable having their mother consent for them. This approach has been used successfully in our previous South African studies with adolescents [
19,
51]. In loco parentis enables the participant to select a female adult (either identified by the participant themselves or by the study staff) to provide consent on her behalf.
Once the adult woman who will provide consent is identified, study staff schedule an appointment for the potential participant and the adult to sign the mother/in loco parentis consent. The potential participant is screened separately again and assented separately from the adult woman to prevent coercion, maintain confidentiality, and provide the opportunity for the participant to decline. The adult woman is also required to sign a confidentiality agreement form as part of the consent process so they understand that their experience at the project site is confidential and that they must maintain confidentiality regarding the person for whom they consented.
Intake assessment
After consenting/assenting, participants also sign a release of medical records permitting the release of information on their SRH, PrEP screening and monitoring results, and other referrals and care pertinent to the study.
Participants take a breathalyzer test to detect recent alcohol use (conducted earlier in the appointment because of the limited detection window for alcohol breath scans) and provide urine for drug screening, locator information is collected, and a photograph of the participant is taken to identify the participant for subsequent appointments. This photo is returned or shredded at the final appointment. The participant then completes a baseline questionnaire on a computer tablet via audio computer-assisted self-interview (ACASI) in either English or Setswana (see AGYW Health CoOp Questionnaire in Supplementary File
3). Trained field staff are available to assist the participant at any time and provide referrals based on prompts related to self-reported violence, symptoms of psychological distress, or suicidal ideation experienced in the past 3 months that are triggered as the participant completes the questionnaire. The questionnaire is derived from components of the youth-specific modification of the Revised Risk Behavior Assessment (RRBA) [
54], which has been adapted and modified for many studies in South Africa. The modified RRBA for this study contains sections on PrEP knowledge, contraception and other SRH services access and utilization, sexual communication, STI symptoms, alcohol and other drug use, relationship equity and sexual control, economic dependence, personal agency, psychological distress, victimization, and peers and social support. After completing the questionnaire, urine drug screening results are provided to assess the recent use of amphetamine, methamphetamine, benzodiazepine, cocaine, marijuana, opioids, and MDMA. These screening tests are repeated at 3-, 6-, and 9-month follow-up assessments.
After the intake appointment, participants interested in initiating PrEP are referred to the youth-friendly nurse in their respective study clinic or the study’s roaming nurse to screen for PrEP eligibility using the National Department of Health (NDoH) guidelines. Baseline screening for PrEP involves reporting of no MDR-TB (part of the study eligibility criteria), HIV and pregnancy testing (conducted prior to the intake process to confirm HIV-negative and not pregnant), estimation of creatinine clearance, and hepatitis B virus (HBV) screening. Contraindications for PrEP use includes poor renal function (estimated creatinine clearance < 60 mL/min) and absence of the hepatitis B surface antigen and antibody. Presence of HBV is not a contraindication for PrEP; however, liver function monitoring is advised if PrEP is initiated. Participants not eligible for PrEP are linked to the public health clinic to receive the standard of care for each respective medical condition. Individuals who are not eligible for PrEP can still remain enrolled in the study.
Participants may decide to initiate PrEP at any time during the study. However, those who do not initiate PrEP during their baseline appointment have to repeat all PrEP eligibility screening. Participants who decide to initiate PrEP are counselled on the importance of adherence and using other protective strategies for up to 20 days of daily dosing before protective levels in vaginal tissue are achieved. Participants are also given guidance on proper management of mild side effects that may occur after initiation and are advised to continue to use condoms and birth control to protect themselves against other STIs and unplanned pregnancy, as PrEP does not provide this type of protection.
Clinic visits to test for HIV and pregnancy status are scheduled 1 month after initiation of PrEP and every 3 months thereafter. At 6 months post initiation of PrEP, screening of creatinine levels is conducted to estimate creatine clearance. If a participant tests positive for HIV at any point in the follow-up appointments, PrEP is stopped for that participant and an active referral is provided for test and treat according the South African HIV guidelines. Nonclinical behavioral study data collection follow-up visits continue to occur at 3, 6, and 9 months post enrollment regardless of whether or when PrEP is initiated or when a participant’s PrEP clinic visits take place.
Sexual and reproductive health services
All participants who complete their baseline appointment are actively referred to the designated youth-friendly nurse or youth-friendly service at their respective study clinic for SRH services; specifically, birth control services. Follow-up on these linkages is conducted at the 3-month follow-up appointment for participants in the control groups and during monthly check-ins for participants in the YWHC groups.
YWHC workshops
The YWHC workshops comprise a 2-session, 4-module program designed to build on the nexus of substance use, HIV, and GBV by increasing knowledge, skills, and agency to reduce GBV and substance use, and ways to decrease sexual risk and HIV incidence. Previous research in South Africa has shown positive outcomes with the WHC, including that women in the WHC were more likely to use condoms with their boyfriends during their last episode of sexual intercourse, more likely to use female condoms with their boyfriends in the past month, more likely to negotiate condom use, and less likely to report daily substance use at follow-up than women who were not in the WHC [
38,
39,
52]. The workshops also include voices and quotes from the formative focus group discussion participants (see Table
2 for an overview of the workshops). Typically, workshops are conducted approximately 7 to 10 days apart, but can be completed on the same day and in either order because of limited availability of the participants; for example, some participants are full-time students. Both workshops end with participants developing an individualized and personalized risk-reduction action plan with assistance from staff. Also, any relevant referrals are offered. Participants also receive toiletry kits and pill boxes with a colorful bag to discreetly store their PrEP. Once both workshops are completed, participants complete a Satisfaction Form to provide feedback and suggestions regarding the intervention. Staff conduct further follow-up with participants individually to check in on their personal action plans at least monthly via mobile phone or in person.
Table 2
Overview of the YWHC Workshops
Workshop 1 |
Topics |
• Becoming an adult woman and influences |
• Sex, sexual expectations, and risk |
• Our reproductive bodies (female and male anatomy) |
• STIs and HIV |
• Ways to reduce risk |
• Male and female condoms |
• Ways of communicating, negotiating, and problem-solving |
• PrEP |
• Birth control |
• Action plan (complete Workshop 1 goals and steps in workshop booklet) |
Workshop 2 |
Topics |
• Inequality and gender power |
• Becoming strong women and concern for how boyfriends treat you |
• Abuse and violence |
• Safety tips for going out |
• Alcohol and drug use |
• Parenting |
• The importance of education and goals |
• Social support, especially taking PrEP |
• Action plan (complete Workshop 2 goals and steps in workshop booklet) |
PrEP navigation after PrEP initiation
Although the South African National Strategic Plan for HIV, TB and STIs (2017–2022) [
47] outlines the availability of PrEP for all AGYW who are HIV-negative and at risk of HIV, PrEP roll-out in public health clinics has yet to occur. Consequently, a majority of public health clinic staff have not been trained on PrEP prescribing and dispensing protocols. Prior to the national roll-out, the PrEPARE Project partnered with the NDoH to train youth-friendly nurses and pharmacists charged with supporting and implementing PrEP delivery from all study clinics on standardized PrEP protocols. A roaming nurse has been hired to assist in the management and support of dispensing PrEP where there is a shortage of trained clinic staff.
PrEP navigation, which occurs through phone calls and text messaging, provides participants in the YWHC intervention arms who are on PrEP with much-needed support, specifically in the first 1 to 2 weeks of PrEP initiation where many participants may experience minor side effects that often lead to PrEP discontinuation. Study staff work with participants to develop feasible plans that can support the daily use of PrEP, such as the pill boxes they receive, adherence strategies, social support, and addressing other concerns about PrEP. Participants also are reminded about their upcoming refill and check-up appointments. This navigation occurs throughout the study duration as participants are due for their refills.
Follow-up assessments
Participants in both study arms return for their 3-, 6- and 9-month behavioral and biological follow-up assessments. Research staff track participants using locator information provided during the intake appointment to help retain participants. The follow-up visit includes reconsenting; updating locator information; a follow-up questionnaire via ACASI; and biological testing for HIV, pregnancy, alcohol use (breathalyzer), and other recent drug use. Dried blood spots (DBS) also are collected at these appointments from participants who report to be on PrEP. The PrEP medication used for this study is a combination of FTC 200 mg/TDF 300 mg tablets. The DBS are used to assess the presence of tenofovir diphosphate (TFV-DP), a measure of cumulative and recent adherence, using a previously validated methodology [
55].
Data management and quality assurance
To protect confidentiality, the study assigns each participant a unique alphanumeric study participant identification number (PID). This PID is the only link between the behavioral and biological data and the identifying information collected for locating participants for their follow-up interviews. Locator forms, consent/assent agreements, and any data that can be linked through the PID are stored separately from other data in double-locked file cabinets in locked rooms at the study’s project site, with restricted access.
Data collection for this study is conducted by highly trained staff from the community who develop a rapport with the study participants to engender trust and elicit the most accurate data possible. All staff sign a confidentiality agreement and are trained on the study’s Quality Assurance Protocol and Quality Management Plan (QMP). Study data are encrypted before they are transmitted daily from the field site to secure servers in the United States. The US-based data manager reviews additional automated quality control checks that the software generates each day. If any critical inconsistencies are noted, the data manager contacts the project director and the field supervisor to resolve these inconsistencies. The Principal Investigator, other members of the research team, and the field staff receive daily field activity reports.
Data and safety monitoring plan and data and safety monitoring board
Procedures have been put in place to address adverse events (AEs) or serious adverse events (SAEs), such as improper disclosure of information or mental or emotional discomfort. As specified in our Data and Safety Monitoring Plan (DSMP), SAEs are reported to the Principal Investigator and the South African Co-Investigator/Medical Director within 24 h of an occurrence and to the Data and Safety Monitoring Board (DSMB), the funding agency, and the Institutional Review Board within 72 h, with appropriate action taken immediately. The study does not interfere with any activities or reports that are part of the public health clinics’ standard operating procedures that do not directly affect the study participants adversely. The DSMP ensures that procedures have been set in place to safeguard the security, validity, and integrity of study data, and that study staff are trained on the policies and procedures for data management according to the Quality Assurance Protocol and QMP. The DSMP also outlines the data analysis plan including preliminary analyses of data for quality assurance and to track the progress of the study.
The study established a DSMB comprising three members: a psychologist, an infectious disease clinician, and a bioethicist. The DSMB will meet every 6 months during the trial to review study progress and ensure adherence to the DSMP. This board is independent from the researchers and the study sponsoring institution. Board members discussed whether stopping rules were necessary for this study; they determined that stopping rules were not necessary.
Outcomes
The clinic-level primary outcomes for this study include clinic staff attitudes and environment, observed discrimination, and stigmatizing avoidance behaviors toward AGYW, as measured by staff surveys and the clinic audits. The AGYW-level primary outcomes include the level of PrEP readiness, uptake, and adherence (as measured by self-report and DBS), and SRH uptake. The AGYW-level secondary outcomes include frequency of substance use, as measured by the RRBA and biological drug screening and breathalyzer tests to assess recent alcohol use; GBV, assessed through self-reported experiences of emotional, physical and sexual abuse; sexual risk as measured by condomless sex, impaired sex, other sex partners; self-reported frequency of experienced stigma.
Sample size and power
For the AGYW analyses, the total sample size and number of clusters (clinic catchment areas) were selected to ensure sufficient power to detect meaningful differences in our primary outcomes, while also balancing considerations tied to reducing possible contamination (exposure to the intervention in control clusters) and implementation feasibility with a set number of clinics in the study area. Sample size estimates for tests of two proportions in a cluster-randomized design [
56] were conducted for a 2-sided test with significance level of 0.05, power of .80, and intra-cluster correlation of 0.01 (based on our team’s past research in Cape Town communities), assuming 10% attrition over the 9-month follow-up period [
52,
57]. This was not adjusted for multiple comparisons. Based on these parameters and calculations, the study sample includes 12 clusters (defined as clinics and their catchment areas), with a total sample size of 900 AGYW enrolled, 75 per cluster (clinic catchment area). This sample was chosen to ensure that meaningful differences in the primary outcomes of PrEP readiness, uptake, and adherence between groups would be detectable. We are powered to detect a difference between 9 to 13% in primary outcomes. Power analyses were first conducted in PASS software [
58] and were refined in Stata [
59] (see PrEPARE Additional Power Calculation Information in Supplementary File
4).
Analysis
Analysis of primary outcome measures will determine whether the provision of S&D reduction training for clinic staff and the YWHC intervention for AGYW will increase PrEP readiness, PrEP uptake, and PrEP adherence, and SRH uptake.
Clinic level
AGYW service utilization will be assessed and analyzed through clinic level data at each of the 12 study clinics during an 8-month period pre- and post-training (e.g., HIV testing; number of PrEP prescriptions). Baseline and follow-up clinic staff surveys will be analyzed to examine changes in attitudes and stigma at the clinic level. Initial analyses will be descriptive and evaluate group differences comparing intervention and control clinics using unpaired t-tests with unequal variance or nonparametric tests, as appropriate. Using multilevel modeling approach, we will investigate and account for clustering at the clinic level. We will use difference-in-differences (DD) methods as done in another South African study to examine utilization of services by AGYW and staff attitudes and behaviors in the 8 months between intervention and control clinics [
60]. The DD estimator reflects the average change in clinics that received S&D reduction training, after the average change in utilization in control clinics is subtracted, assuming parallel trends in clinics over time had the S&D reduction training not been implemented. DD estimates will be calculated for each outcome separately; models will include covariates for time (pre- vs. post-intervention), treatment (training vs. no training) and an interaction between the two. The clinics identified are similar in size, services, and populations served. Any clinic features for which we cannot achieve balance through the design will be included as covariates in the multivariable DD analyses. A stigma score will constitute one primary outcome for these analyses. To examine short-term changes between pre-intervention and 4-month follow-up, DD methods can be used to compare the average change in stigma scores in clinics that received S&D reduction training, removing any changes observed in control clinics. We will use generalized estimating equations (GEE) [
61] to examine provider-level stigma scores measured at 4- and 8-month follow-up comparing trained and untrained clinics, accounting for clustering at the clinic-level.
AGYW level
The testing of Aim 3 will be done by intention to treat analysis, examining the effects of intervention groups at both levels (clinic and individual) on PrEP readiness, uptake and adherence. PrEP uptake will be defined as any use of PrEP. We will consider both short-term uptake within the first 3 months after study enrollment and uptake at any point during the 9-month follow-up period. Initially, we will calculate the proportion of AGYW with any use of PrEP, overall, and within each study arm, calculating 95% confidence intervals using the binomial distribution. Given the need to consider the influence of the cluster randomization by clinic, we will use generalized estimating equations (GEE) [
61] models with a logit link and an exchangeable correlation structure to estimate the effects of randomization groups on both short-term (3-month follow-up) and long-term follow-up period (9-month follow-up). Models will include both intervention groups (i.e., S&D reduction training and YWHC) as covariates. Effects of clustering by randomization clinic may be examined by adjusting the variance by the inflation factor [1 + (m-1)r], where
m is the average clinic size and
r is the interclass correlation estimate. If the Hausman assumption of correlation between the random and fixed effects is violated, then we may include fixed effects representing cluster identification.
Finally, we will include additional baseline covariates in the models, including sociodemographic factors and behavioral risks, should initial descriptive analyses suggest differences in the distribution of these factors across study arms. In subsequent exploratory analyses, we will examine the potential mediating and moderating roles of key behavioral, social, and structural factors hypothesized to influence PrEP uptake (e.g., social support; economic dependence), as outlined in
Outcomes. To assess adherence in the past month, we will measure drug-level concentration, TFV-DF assessed through DBS samples, at 3, 6 and 9 months. The threshold for adherence will be established based on recommendations from ongoing studies and research designed to inform thresholds in women. Secondary adherence measures will include self-reported use, which will be examined as the number of days in the past 30 days when PrEP was taken. Using clinical/pharmacy data on PrEP dispensation and self-report of months in which PrEP was used, we will examine adherence (the number of months of PrEP use) and discontinuation. Initially, we will examine differences in the proportions of AGYW who are adherent at months 3, 6, and 9 in each study arm, calculating confidence intervals using a binomial distribution. Our analytic approaches will be similar to those described for uptake. We will use generalized linear mixed models to allow for multiple variance components, including clustering of clinics and repeated measures of adherence within individuals over time. We will use logit, Poisson, and linear regression approaches, as appropriate.
Analyses of secondary outcomes will use similar analytic approaches and examine hypotheses tied to intervention effects on behavioral risk endpoints (e.g., substance use) and on AGYW’s assessment of stigma (e.g., participants attending the S&D reduction trained clinics will report lower levels of stigma from clinic staff than participants who attend the non-trained clinics). Other model specifications, such as zero-inflated Poisson or negative binomial will be considered for count outcomes (e.g., the number of condomless sex acts) that have excess zero values or overdispersion.
From the AGYW outcomes, it is hypothesized that AGYW enrolled from clinics who received clinic-level S&D reduction training will have higher PrEP and SRH utilization than AGYW enrolled from clinics that did not receive training. AGYW enrolled from clinics that are randomized to receive the YWHC intervention will have higher PrEP readiness, uptake and adherence than those enrolled from clinics that are randomized to PrEP and SRH provision only. AGYW enrolled from communities in clinic catchment areas that are randomized to receive the YWHC will report less HIV-related risk at follow-up (e.g., GBV, substance use, condomless sex) than those enrolled from clinics that are randomized to PrEP and SRH provision only.
Missing data in terms of data management may occur because of nonresponse and study attrition. We will analyze differential attrition in relation to participants’ key demographic characteristics. Based on our prior research conducted in South Africa with AGYW who engage in sex risk behaviors, we estimate an average retention rate of 90% over the repeated follow-ups [
19,
37,
62]. We will address missingness by including demographic covariates that will serve as proxies for dropout and by conducting a sensitivity analysis.