WhatsApp-based sexual and reproductive health education for adolescents with perinatally acquired HIV: lessons learned from an mHealth pilot intervention in South Africa

ABSTRACT

Adolescents with perinatally acquired HIV in South Africa have limited access to relevant sexual and reproductive health (SRH) education. Mobile health (mHealth) can discreetly deliver such education but has not been tested with members of this population. In this study, we describe the effects of an mHealth intervention to support transition to adult care – Interactive Transition Support for Adolescents with HIV (InTSHA) – on SRH attitudes. Between May 2021 and February 2022, we enrolled adolescents with perinatally acquired HIV aged 15–19 from a government clinic in urban KwaZulu-Natal, South Africa. Baseline and six-month SRH questionnaires were adapted from a WHO adolescent SRH survey, consisting of nine positively worded items on a 4-point Likert scale. Transcripts of the mHealth intervention modules were thematically analysed. Of 80 participants enrolled, 47 completed the SRH survey at both time points (mean age: 17.2 years); 25 (53.2%) were female and 19 (40.4%) were randomised to InTSHA. Likert scores improved from baseline to follow-up for the InTSHA group for specific SRH questions, overlapping with the frequency that topics such as ‘Defining Sexuality’, ‘Discussing Sex with Caregivers’, and ‘Comfort Saying No to Sex’ – were discussed by adolescents. mHealth is a tool for interactive SRH education that should be further developed.

KEYWORDS:

• Sexual health education
• adolescent sexual health
• adolescent HIV
• mHealth

Introduction

Sub-Saharan Africa (SSA) accounts for 20.6 million of the 37.7 million people living with HIV in 2020 (UNAIDS 2023) While the global prevalence of HIV in young people ages 15–24 is less than one percent (The World Bank 2022), South Africa is home to the highest prevalence of 12–19 year olds with HIV – at 5.7% in girls/young women and 4.5% in boys/young men with Black South Africans experiencing the majority of cases (Mabaso et al. 2021).

Most South African adolescents living with HIV (APHIV) have acquired HIV perinatally. Members of this generation now face the dual challenge of adolescent sexual health risks on top of the social and medical challenges of living with HIV (World Health Organization 2022). Most South African adolescents living with perinatal HIV report being sexually active and nearly a quarter do not graduate from high school, many due to teenage pregnancy (Ndlazi and Masango 2022). A meta-analysis of studies conducted in sub-Saharan Africa found that the pooled prevalence of sexual risk behaviour in adolescents with HIV was 36.6% (Wondmeneh and Gebeyehu Wondmeneh 2023). According to a study in Uganda, 45% of APHIV had ever been sexually active, 57% of them did not consistently use condoms, and 31% of them had disclosed their HIV status to sexual partners (Ankunda, Muhimbuura Atuyambe, and Kiwanuka 2016). This calls for adolescent-friendly services encouraging retention in HIV care and tailored to the unique sexual and reproductive health (SRH) needs for APHIV in South Africa (Zanoni et al. 2016).

Mitigating sexual risk for South African adolescents requires several factors – including a supportive and abuse-free home, food security, and HIV knowledge as well as access to condoms, contraception, and antiretroviral treatment (Toska et al. 2020). Sexual health education is imperative to decrease sexual risk behaviours in adolescents, including APHIV, but certain forms of education have proven more effective than others (UNAIDS 1997). Metrics of high-quality, sexual health education programming include having well-qualified staff, using engaging and relevant strategies, being inclusive of all genders and sexualities, connecting youth to health services, collaborating with caregivers, and building communication skills (Centers for Disease Control and Prevention 2023). APHIV particularly require comprehensive sexuality education to begin to challenge HIV and sexuality stigma, bridge knowledge and action, and focus on psychosocial, behavioural, and physical factors of safe sex and healthy relationships (Kirby, Laris, and Rolleri 2022).

Mobile Health (mHealth) interventions have been shown to successfully educate adolescents in low- and middle-income countries about SRH risks, and improve SRH outcomes, despite occasional technological barriers (Feroz et al. 2021). This format became critical and increasingly popular during the COVID-19 pandemic, when school-based sexual health education and other in-person programming was disrupted (PATA 2022). Interventions such as the iloveLife.mobi website and incentivisation programmes in South Africa have been effective in increasing health-seeking behaviour in youth (Visser, Kotze, and van Rensburg 2020). Various other interventions, such as an SMS programme amongst young women in Bangladesh (Ahmed 2020) and an SRH text message intervention for young men and women in Kenya (Mwaisaka et al. 2021), have been shown to be acceptance effective and efficacious in providing SRH education (Williamson 2022). However, even in South Africa with 77.0% mobile phone coverage in urban areas (Miyajima 2020) and with the highest rates of young people aged 15 to 24 living with HIV in the world—22.3% of young women and 7.6% of young men (Kharsany et al. 2018), no known SRH interventions have been conducted exclusively for APHIV, who face unique SRH challenges (Onukwugha et al. 2022).

This study reports on findings from InTSHA (Interactive Transition Support for Adolescents with HIV), a 10-module WhatsApp-based (Meta Platforms) social support group that was developed to improve peer support and communication by leading evidence-based discussions with APHIV (Zanoni et al. 2022). The intervention, which included two modules on sexual health, was designed to facilitate the transition to adult care. Here, we compare improvements in SRH knowledge and attitudes between APHIV who received the SRH intervention compared to APHIV who received the standard-of-care in an urban, South African setting.

Methods

Study setting

This study was conducted in KwaMashu Poly Clinic, a government-supported HIV clinic within an urban township of KwaZulu-Natal (KZN), South Africa. South Africa, and specifically KZN, has the highest prevalence of adolescents with HIV in the world, most of whom acquired HIV perinatally (The World Bank 2022). The main languages spoken in Durban are isiZulu and English. All South African adolescents receive a basic, nationally standardised sexual health education curriculum as part of their Life Orientation classes, which is meant to provide information about puberty, peer pressure, and the prevention of STIs, HIV and pregnancy. However, the curriculum, introduced in 1997 and last revised in 2011, does not mention terms such as ‘contraception’, ‘condom’, ‘homosexual’. ‘LGBT’, ‘gay’ or ‘healthy relationship’ (Curriculum and Assessment Policy Statement (CAPS) Life Orientation 2011).

Parent study design and InTSHA intervention

Between March 2021 and February 2022, the InTSHA study team enrolled 80 APHIV aged 15–19 years old, in person at the clinic. Enrolled participants were randomised after enrolment to receive either the social media intervention, InTSHA, or continue standard-of-care. Eligible participants had perinatally acquired HIV and were: (1) aged 15–19 years; (2) receiving ART for at least six months; and (3) fully aware of their HIV status. Participants were excluded if they were unable to read and/or speak English or isiZulu, or were too mentally or physically impaired to provide informed consent. Informed consent was obtained from participants ages 18 and over, and informed assent and caregiver consent were obtained by participants under 18 years old. All consent and assent forms were read to the participants in their chosen language (isiZulu or English), and participants were given the opportunity to ask questions to ensure their understanding.

InTSHA was developed using user-centred design with South African adolescents, caregivers and healthcare providers based on the Socioecological Model of Adolescent and young adult Readiness for Transition (SMART). InTSHA uses encrypted, closed group chats delivered via WhatsApp to provide peer support and improve communication between adolescents with HIV, their caregivers, and healthcare providers during transition from paediatric to adult care. A paper on the formative development of InTSHA is in preparation.

Within the intervention group, cohorts with approximately 10 participants were created and assigned to virtual text-based group chats on WhatsApp, which also included a facilitator and co-facilitator. Each virtual chat group conducted 10 modules with the same cohort – approximately one scheduled, two-hour session per week – with each module discussing a new topic. These group chats were led by a facilitator with a standardised, semi-structured discussion guide for each module conducted over approximately two hours during a time scheduled to not conflict with school hours. Participants were free to respond to the open-ended prompts during the two-hour session period, as well as use the group chat for follow-up questions any time throughout the week. The full protocol has been published elsewhere (Zanoni et al. 2022).

Two of these 10 topics – ‘Sexual and Reproductive Health’ and ‘Gender and Sexuality’ – were selected for this sub-study to specifically evaluate the feasibility of mHealth-based sexual health education for APHIV, to further develop this sexual health education curriculum for future research. Content for the curriculum was inspired by the Right to Care Flipster (Right to Care 2017), an in-person health education intervention developed for APHIV in Johannesburg. As detailed in online Supplementary File 1, the objectives of the two SRH modules included defining key terms related to SRH; improving understanding of HIV transmission; discussing the practice of safe sex; reflecting on personal biases and values related to sexuality; distinguishing between healthy and unhealthy relationships; and identifying signs of sexual abuse and assault.

Sub-study inclusion criteria

In this sub-study mixed-methods analysis, consecutive selection was used for adolescents who had 1) completed the baseline SRH questionnaire, 2) completed the six-month SRH questionnaire, 3) were randomised to the intervention group, and 4) participated in at least one of the two SRH modules. Due to a small sample size, this sub-study was not powered to detect significant differences in SRH outcomes.

Quantitative data collection

At the first recruitment visit, baseline demographic and HIV-1 plasma RNA viral load data were collected. These data included age, gender identity, and grade in school. Additionally, participants were screened with validated HIV adolescent readiness for transition scale (HARTS) (Zanoni et al. 2021). All participants then completed an in-person, paper-based pre-intervention survey that asked knowledge- and attitude-based questions on the content covered in the two SRH modules in the InTSHA intervention. This psychometric survey included nine questions assessing their attitudes and beliefs about gender, sexuality, and SRH and was adapted from the World Health Organization’s instrument ‘Asking Young People about Sexual and Reproductive Behaviors’ (Cleland, Ingham, and Stone 2001) to fit content discussed in the modules and worded to be appropriate to a South African audience. Response options were given as a four-point Likert scale (Strongly Disagree, Disagree, Agree, and Strongly Agree) (Sullivan and Artino 2013). The surveys were completed on paper in the participant’s preferred language – either English or isiZulu. The SRH questions were added to the baseline questionnaire after recruitment had already begun, so not everyone who was enrolled had a baseline SRH score. Six months after randomisation, or after completion of the intervention (if longer), participants in both arms of the study completed an in-person, paper-based follow-up survey assessing the same metrics as well as a blood test assessing viral load.

Quantitative data analysis

Paper survey data were entered into a REDCap Database hosted on the Emory University server. Using Microsoft Excel and Jamovi version 2.3, we analysed descriptive demographic, biological, intervention attendance, and questionnaire responses for both arms of the study, before and after the intervention. Questionnaire results were computed as numerical values centred around zero, with ‘Strongly Disagree’ as −2, ‘Disagree’ as −1, ‘Agree’ as + 1, and ‘Strongly Agree’ as + 2. All questions were positively worded, thus ‘Strongly Agree’ (+2) was always the most favourable answer. For those who completed the SRH questions on both surveys, the mean score was computed for each question before and after the intervention for both the intervention and control arms. The difference in scores for each question was calculated by subtracting the baseline response from the follow-up response. The average of these differences for each question (mean difference) was taken for each question. The mean differences were used in independent t test, paired samples t test, and a Pearson’s correlation to determine what demographic variables were associated with question improvement from baseline to follow-up in both the InTSHA and control arms.

Qualitative data collection and analysis

Each participant was offered the opportunity to engage in two, approximately two-hour, semi-structured, small-group SRH intervention modules that were conducted in a closed, encrypted, text-message-based WhatsApp group chat by a female South African, isiZulu and English-speaking, undergraduate level research coordinator (TS) who had facilitated all the prior intervention modules. Notes on attendance, participation, and pace were taken during the modules by a Master’s-level, US, English-speaking female researcher (SB) who observed the modules. After the completion of each SRH WhatsApp module, the WhatsApp module transcript was exported, de-identified, translated into English by the facilitator (from a combination of English and isiZulu), and stored on Dedoose Version 9.0.17, a qualitative coding software. Each module transcript was manually hand-coded and thematically analysed by two US, Masters-level research assistants, one of whom had served as a co-facilitator in each WhatsApp intervention module. Content analysis assessed major topics covered during the module, with a codebook based on the SRH questionnaire, to assess the fidelity of the intervention. In this sub-study analysis, relevant quotations are linked to corresponding quantitative findings. We followed the consolidated criteria for reporting qualitative research (COREQ) checklist for reporting qualitative research (online Supplementary File 2).

Ethical approvals

The Biomedical Research Ethics Committee of the University of KwaZulu-Natal, the KwaZulu-Natal Department of Health, the Massachusetts General Brigham Research Ethics Board, and Emory University Institutional Review Board approved this study.

Results

Aggregate baseline characteristics

Of the 80 participants enrolled in InTSHA, 40 were randomised to the intervention arm and 40 were randomised to the control arm (Figure 1). Within the intervention arm, 29 completed the SRH baseline survey due to the delayed rollout of the SRH questionnaire, 19 of whom participated in at least one SRH module of the intervention due to attendance challenges, including school examinations and family engagements, throughout the intervention. Participation was defined as sending at least three messages in the WhatsApp group chat during the session. All 19 participants also completed the six-month SRH follow-up. In the control arm, 28 completed the baseline SRH survey and 100% of those adolescents returned to complete the six-month follow-up survey. This analysis thus includes 47 participants—19 in the intervention arm and 28 in the control arm. None of the participants recruited to the study refused to participate.

Figure 1. Study sample flowchart.

Table 1 shows the baseline sample characteristics stratified by study arm. All participants were Black South African adolescents who spoke isiZulu as a first language. The average age of participants upon enrolment was 17.2 years (range: 15–19), and 53.2% of the sample were female (n = 25). Twenty-seven (57.4%) were in Grade 10 or below, while the rest were in Grades 11, 12, or had graduated. At baseline, 42 (89.4%) had an undetectable viral load (<20 copies/ml).

Table 1. Sample characteristics of included adolescents, stratified by study arm (n = 47).

	InTSHA (n = 19) N (%)	Control (n = 28) N (%)	Total (n = 47) N (%)
Baseline Age (Mean)	17.4	17.1	17.2
Gender
Female Male	14 (48.3%)	11 (39.3%)	25 (53.2%)
	15 (51.7%)	17 (60.7%)	32 (68.1%)
Baseline Grade
Grade 10 and below Grade 11 and up	12 (63.2%)	15 (53.6%)	27 (57.4%)
	7 (36.8%)	13 (46.4%)	20 (42.6%)
SRH Modules Attended
One Both	9 (47.4%)	–	9 (47.4%)
	10 (52.6%)	–	10 (52.6%)
Baseline Viral Load
Undetectable* Detectable	14 (73.6%)	20 (71.4%)	34 (72.3%) 13 (27.7%)
	5 (26.3%)	8 (28.6%)
6-Month Viral Load
Undetectable* Detectable	15 (78.9%)	27 (96.4%)	42 (89.4%)
	4 (21.1%)	1 (3.6%)	5 (10.6%)

* <20 copies/ml).

Demographics

Control arm

Of the participants in the control arm (n = 28), the mean age was 17.1 years (Table 1). Eleven (39.3%) were female. Fifteen (53.6%) were in grade 10 and below. Twenty (71.4%) had an undetectable viral load (<20 copies/ml) at baseline and 27 (96.4%) had an undetectable viral load at follow-up.

Intervention arm

Of the included participants in the InTSHA arm (n = 19), the mean age was 17.4 years (Table 1). Fourteen (48.3%) were female. Twelve (63.2%) were in grade 10 or below. Fourteen (73.6%) had an undetectable viral load (<20 copies/ml) at baseline and 15 (78.9%) had an undetectable viral load at follow-up.

SRH scores

Control arm

Table 2 highlights the average baseline and follow-up score for each individual SRH question. Four questions showed improvements in mean score from baseline to follow-up: ‘Physical Violence is Never Justifiable’ (+0.14), ‘Comfort Saying No to Sex’ (+0.07), ‘Insisting on Condom Use’ (+0.35), and ‘Comfort Accessing SRH Services’ (+0.07).

Table 2. Pre- and post-intervention survey responses⁺.

	InTSHA Group Baseline Mean (n = 19)	InTSHA Group Follow-Up Mean (n = 19)	InTSHA Group Mean Difference (n = 19)	Control Group Baseline Mean (n = 28)	Control Group Follow-Up Mean (n = 28)	Control Group Mean Difference (n = 28)
Q1: I feel comfortable defining the word ‘sexuality’.	−0.26	+0.11	+0.37	+0.21	−0.04	−0.25
Q2: I never feel pressure from others to have sexual intercourse.	+0.26	+0.16	−0.10	+0.43	+0.25	−0.18
Q3: I feel comfortable discussing sex-related matters with my caregivers.	−0.42	+0.53	+0.95	−0.18	−0.18	+0.00
Q4: Physical violence is never justifiable in a relationship.	+0.37	+0.11	−0.26	+0.32	+0.46	+0.14
Q5: I feel that I have the resources to prevent pregnancy and STIs when I choose to have sex.	+1.11	+0.89	−0.22	+0.61	+0.54	−0.07
Q6: I feel comfortable saying ‘no’ to someone if I do not want to engage in a sexual act.	+0.68	+1.26	+0.58	+0.89	+0.96	+0.07
Q7: I am confident that I can insist on condom use every time I have sex.	+1.58	+0.79	−0.79	+0.86	+1.21	+0.35
Q8: It is equally the man’s and the woman’s responsibility to ensure that contraception is used regularly.	+0.84	+0.74	−0.10	+1.11	+0.79	−0.32
Q9: I would feel comfortable visiting a health facility or doctor to receive services or information on contraception, pregnancy, abortion or STIs.	+1.16	+0.84	−0.32	+0.57	+0.64	+0.07

⁺Score ranges from −2 to 2, where ‘Strongly Disagree’ is valued at −2, ‘Disagree’ is valued at −1, ‘Agree’ is valued at ‘+1’ and ‘Strongly Agree’ is valued at + 2

Figure 2 shows that at baseline, the majority of the control group responded favourably (agree to strongly agree) to all SRH questions (60.7–92.9%) except for ‘Discussing Sex with Caregivers’. At follow-up, the control group had divisive responses for ‘Defining Sexuality’ (‘Agree’ = 50%) and ‘Discussing Sex with Caregivers’ (‘Agree’ = 50%).

Figure 2. Response Distributions Stratified by SRH Question and Intervention Arm (n=47).

Intervention arm

Nineteen participants who participated in the SRH intervention completed both the baseline and follow-up questionnaires. Scores at baseline and follow-up are reported in Table 2. Mean scores improved from baseline to follow-up for three questions: ‘Defining Sexuality’ (+0.37), ‘Discussing Sex with Caregivers’ (+0.95), and ‘Comfort Saying No to Sex’ (+0.07). The concept of discussing sex with caregivers was brought up by participants in the SRH modules (Table 3). For example, on the topic of sexual harassment, one participant said, ‘My mum is teaching me about it, so I can see if I am being abused’ (Female, Age 18).

Table 3. Topics discussed in SRH modules, by SRH question.

SRH Question	Quote from Module Transcript
Defining Sexuality	Sexuality means gender or having intercourse (Female, Age 16)
Sexual Pressure	Facilitator: How do you know if you do or do not have consent? For example – Baby you can have sex with me but if you are not ready, I will understand and respect your wish and I will wait for you bb and take your time (Male, Age 16) When other person clearly said yes (Female, Age 18) You look at her reaction (Male, Age 16)
Sex Talk	Facilitator: Wow impressive! Did you get his explanation from school or Google? Nope my mum is teaching me about it so I can see if I am being abused (Female, Age 18)
Domestic Violence	Sexual harassment is someone harassing you by looking at you badly (Female, Age 16) Sexual harassment is touching someone against their body and assault is hitting them (Male, Age 15)
Accessing Contraception	A girl should inject (Female, Age 18) Use a condom and inject the girl (Male, Age 16) Without injecting you should not have sex at a young age (Female, Age 16)
Saying No to Sex	Peers say it is important to have sex when you are in a relationship, but I don’t think so (Female, Age 16)
Condom Use	You must use a condom if you want to have sex (Male, Age 16)
Equal Responsibility	Facilitator: What are the roles of men and women in a relationship? That they are both honest and trustworthy I think (Female, Age 15) Have to love and care for each other (Female, Age 16)
Healthcare Comfort	Facilitator: How can you be a good friend if your close friend of yours had a traumatic experience being sexually assaulted? I would say that he should go to the doctor to check that he does not have any problems (Female, Age 16) He/she must go & get counselling (Female, Age 18) I say let him go to social workers (Male, Age 15) He/she can reach out for counselling (Male, Age 16)

The SRH question response frequency for each question, stratified by intervention and control group, baseline and follow-up, is shown in Figure 2. At baseline, for the intervention participants included in the analysis, the majority responded with Agree for all (57.9–89.5% agreement). For two questions, the majority responded Disagree (‘Defining Sexuality’ and ‘Discussing Sex with Caregivers’, Agree = 42.1% and 36.9%, respectively). At follow-up, the majority of InTSHA group responded with Agree to all nine SRH questions (52.6–89.5% agreement). There were no significant unpaired differences in improvement between the intervention and control group.

In qualitative analysis of the SRH transcripts, ‘Defining Sexuality’ and ‘Discussing Sex with Caregivers’ were emphasised in the design and implementation of the modules. The first SRH module began by asking participants what they knew about sexuality, as a concept and in their culture. One participant defined sexuality as, ‘Sexuality means gender or having intercourse’ (Female, Age 16) prior to an activity differentiating sex from gender. Highlighted quotations from the transcripts can be found in Table 3, organised by their thematic relationship to each SRH question tested.

Correlation between participant characteristics and SRH score improvement

According to Pearson’s correlation (Table 4), among those who attended one InTSHA SRH module, we found a significant correlation between improvement in ‘Access to Contraception’ and having a lower baseline (−0.805, p = .009) viral load. There was also a significant correlation between improvement in ‘Saying No to Sex’ and younger age (−0.774, p = .014). Many participants spoke about how they would give and receive consent for sex. One 16-year-old young man wrote, ‘Baby, you can have sex with me, but if you are not ready, I will understand and respect your wish’. There was also a correlation between never having felt ‘Sexual Pressure’ and having higher transition readiness at baseline (0.695, p = 0.038). When discussing consent, participants said consent was achieved ‘when the other person clearly said yes’ (Female, Age 18) and that you must ‘look at her reaction’ (Male, Age 16). Finally, there was a significant negative correlation between improvement in ‘Equal Responsibility’ and younger age (−0.897, p = .001).

Table 4. Correlation Matrix of Variables that Influenced Odds of Improvement for Each Question for the InTSHA Arm (n = 19).

	Attended One Module (n = 9)			Attended Both Modules (n = 10)
SRH Question	Age; p-value	Baseline VL; p-value	HART Score; p-value	Age; p-value	Baseline VL; p-value	HART Score; p-value
Q1: Defining Sexuality	0.313;	−0.024;	0.313;	0.359;	0.144;	−0.462;
	0.412	0.951	0.412	0.342	0.711	0.211
Q2: Sexual Pressure	0.109;	0.06;	0.695;	0.321;	−0.036;	−0.334;
	0.78	0.877	0.038	0.400	0.927	0.380
Q3: Sex Talk	−0.436;	−0.18;	0.507;	−0.347;	0.395;	−0.145;
	0.241	0.644	0.163	0.360	0.292	0.709
Q4: Domestic Violence	−0.065;	−0.338;	0.148;	0.334;	0.146;	0.108;
	0.868	0.374	0.705	0.379	0.707	0.783
Q5: Accessing Contraception	−0.37;	-0.805;	−0.403;	0.327;	0.300;	−0.188;
	0.327	0.009	0.282	0.391	0.433	0.628
Q6: Saying No to Sex	-0.774;	−0.33;	−0.371;	0.099;	−0.438;	−0.118;
	0.014	0.386	0.325	0.800	0.238	0.762
Q7: Condom Use	−0.514;	−0.324;	−0.608;	0.521;	-0.707;	0.335;
	0.157	0.396	0.082	0.150	0.033	0.378
Q8: Equal Responsibility	-0.897;	−0.399;	−0.303;	0.706;	-0.763;	0.109;
	0.001	0.287	0.429	0.034	0.017	0.780
Q9: Healthcare Comfort	−0.531;	0.448;	−0.143;	0.854;	-0.673;	0.099;
	0.141	0.227	0.714	0.003	0.047	0.799

Among those who attended both InTSHA modules, having a higher baseline viral load (absolute value) was inversely correlated with improvement in ‘Condom Use’ (−0.707, p = .033). When asked about condom use, one adolescent said, ‘You must use a condom if you want to have sex’ (Male, Age 16). Improvement in ‘Equal Responsibility’ was positively correlated with older age (0.706, p = .034) and negatively correlated with baseline absolute viral load (−0.763, p = .017). When discussing the roles of partners in a relationship, one participant noted the importance ‘that they are both honest and trustworthy’ (Female, Age 15). Improvement in ‘Healthcare Comfort’ was directly correlated with older age (0.854, p = .003) and inversely correlated with baseline absolute viral load. In the modules, participants discussed the concept of comfort seeing healthcare providers through a hypothetical situation of helping a friend who had been sexually assaulted. One participant suggested, ‘I would say that she should go to the doctor to check that she does not have any problems’ (Female, Age 16). Because 100% of their six-month viral load were undetectable, these correlations could not be calculated for this group.

Within-group differences in SRH score improvement

Paired t-tests were used to assess within-group differences, and no significant differences were found in the pre-post scores of all items in both groups.

Discussion

This sexual health sub-study analysis of InTSHA revealed select improvements from baseline to follow-up in SRH scores for APHIV who completed the SRH modules of the intervention. Improved comfort with ‘Defining Sexuality’, ‘Discussing Sex with Caregivers’, and ‘Comfort Saying No to Sex’ were reflected in the major topics discussed in the intervention module transcripts. At baseline, neutral responses to positively worded SRH questions, especially the ability to define sexuality and discuss sex with caregivers, demonstrated that ALPHIV were lacking in vital information and resources from their pre-existing knowledge base from Life Orientation and HIV education. This low baseline knowledge was also reflected in qualitative in-depth interviews with these APHIV (Bergam, Kuo, et al. 2022). At follow-up, in the intervention group, both mean and median scores improved, albeit non-significantly, from baseline for ‘Defining Sexuality’ and ‘Discussing Sex with Caregivers’. These results align with qualitative transcript data which showed participants speaking openly about sexuality with healthcare providers, caregivers, peers, and sexual partners. While this pilot sub-study was not powered to detect statistically significant quantitative findings, these results suggest that the content and the nature of InTSHA as an mHealth intervention show promise in de-stigmatising conversation about sexuality amongst APHIV living in urban KwaZulu-Natal. Creating a discreet, accessible, and informative group setting in which to discuss SRH issues tailored to APHIV can help engage adolescents and young people in holistic healthcare.

Higher attendance in SRH modules predicted stronger improvement in SRH scores. Amongst the group that attended both SRH modules, improvements in ‘Condom Use’, ‘Equal Responsibility’, and ‘Healthcare Comfort’ were inversely correlated with baseline viral load. The strong inverse correlations between high viral load and improvement in ‘Accessing Contraception’ detected in this study is supported by the literature on South African pregnancy rates among adolescents with poorly controlled HIV (Naicker et al. 2022). Adolescents living with HIV who know that they are not virally suppressed may feel nervous coming to the clinic for any of their health needs, including contraception. This finding suggests that control over one’s HIV with medication adherence and regular attendance of clinic appointments may be related to positive health behaviours and attitudes related to sexual and reproductive health. Additional effort must be made to engaging youth, especially those who are not engaged in care or virally suppressed, in holistic sexual reproductive healthcare where they can speak openly to providers about their sexual health and access the resources they need. Future qualitative research should seek to understand the perspectives of adolescents who have a high viral load, and ensure we are reaching these populations with reproductive healthcare.

South African adolescents typically do not feel comfortable receiving sexual health information or support from their caregivers or family members (Bergam, Sibaya, et al. 2022), and when they do, the information given typically focuses on promote abstinence and putting off relationships. As a result, adolescents tend to seek out online resources and their friends for information, advice, and support about SRH – but even online, very little information is tailored to adolescents with HIV. The InTSHA curriculum on Sexual and Reproductive Health and Gender and Sexuality offers sexual health education via mHealth, an increasingly popular intervention format that has begun to address these gaps. Qualitative research with adolescents who participated in InTSHA noted its positive, inclusive, destigmatising, interactive, and culturally competent content (Bergam, Kuo, et al. 2022). Its curriculum was tailored to the unique social environment within which they lived (Zanoni et al. 2022). InTSHA facilitators built longitudinal relationships with adolescents over the course of weeks, including through an initial in-person module and with less-sensitive topics covered in the weeks leading up to the SRH modules. As assessed in engagement in the modules as well as through qualitative assessments of feasibility and acceptability, participants were found to be receptive to discussing these stigmatised topics in a group of people who are collectively living with HIV in their community.

Two modules delivered in the middle of a health promotion intervention cannot be expected to significantly improve SRH knowledge and outcomes on their own. However, this pilot intervention has shown the feasibility and potential efficacy of mHealth, as a semi-structured facilitated group chat, as a form of peer support and sexual health education for adolescents with HIV in South Africa.

Limitations

This study has several limitations. Two modules in a single intervention are unlikely to change attitudes and long-standing beliefs about SRH. For some adolescents, the intervention was associated with improved attitudes and beliefs regarding sexual and reproductive health, but this study was not powered to detect significant differences in these outcomes. An ongoing NIH-funded clinical trial (R01MH131434) is evaluating the effectiveness of in-person and mHealth delivery of the InTSHA intervention in 16 clinics throughout KwaZulu-Natal, South Africa.

Restrictive inclusion criteria limited the generalisability of our study. Inclusion criteria included that participants were enrolled from KwaMashu Community Health Centre in a large urban township outside of Durban, where they were linked to paediatric HIV care. The results may not be generalisable to more rural environments, or to adolescents who do not have regular access to paediatric HIV care. During the intervention, challenges with module attendance around school holidays, poor Internet connectivity, and a natural disaster made it difficult for many of the InTSHA-enrolled adolescents to be present for the two SRH modules. Any significant correlations between age, viral load, and improvement must be considered with reservations given our small sample size.

Not all follow-up assessments were undertaken exactly 6 months after recruitment, due to the duration of the intervention, as well as holidays and school breaks. The potential confounder of duration between baseline and follow-up is controlled for in the adjusted model.

It is also important to note that the use of a Likert scale may have introduced acquiescence bias and habituation bias, which may be outweighed in future studies by a larger sample size (Westland 2022). Finally, the SRH questionnaire, while adapted from a WHO guideline for asking adolescents about sex, was not a pre-validated set of questions for use with this population. In hindsight, the wording of some SRH questions, such as the question pertaining to interpersonal violence, could have been misconstrued by adolescents because it could not be easily translated into isiZulu without losing some of its meaning. This is an important consideration for future groups adapting English questionnaires to other languages.

Conclusion

Utilising mHealth as a method of sexual health education for APHIV has the potential to successfully change the attitudes about and knowledge of SRH. This pilot study provides a foundation on which to base future iterations of InTSHA dedicated to complementing and scaling up sexual health education for APHIV in South Africa.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The work was supported by the Fulbright Association (SB); the US National Institute of Mental Health (BCZ) [K23MH114771]; National Institute of Mental Health (JH) [K24MH114732]; and the Center for AIDS Research, Emory University (VM) [P30AI050409].