Menu
Menu
Menu
Menu
Background: Rates of adolescent HIV and unintended pregnancy in southern Africa are amongst the highest in the world. Gender-transformative interventions that address underlying gender inequalities and engage both males and females have been emphasised by the World Health Organisation, amongst others, to target prevention. However, few such gender-transformative interventions have been rigorously developed or evaluated. Objective: To expedite potential impact and reduce development costs, we conducted a needs assessment to inform the co-design, in consultation with local stakeholders, of adapted versions of an existing gender-transformative Relationships and Sexuality Education intervention for use in South Africa and Lesotho. Methods: Adaptation of the intervention was guided by a modified version of Intervention Mapping (IM). This process involved consultation with separate adolescent, community and expert advisory groups and a collaboratively conducted needs assessment, which drew on focus groups with adolescents (8 groups, n = 55) and adults (4 groups, n = 22) in South Africa and Lesotho, and was informed by our systematic review of the literature on the determinants of condom use among adolescents in the region. Results: The findings clarified how the intervention should be adapted, which individual- and environmental-level determinants of condom use to target, and actions for facilitating successful adoption, evaluation and implementation in the new settings. Conclusions: The IM approach allows for a systematic appraisal of whether components and processes of an existing intervention are appropriate for a new target population before costly evaluation studies are conducted. The findings will be of interest to those wishing to rigourously develop and evaluate gender-transformative interventions engaging men to improve health for all.
The research design was grounded in three methodological approaches: Ensuring methodological rigour was a primary concern and we were guided in this endeavour by Lincoln and Guba’s concept of trustworthiness, incorporating considerations of credibility, dependability, confirmability and transferability [35]. Our approach to incorporating each of these considerations is addressed the relevant sections below. If I Were Jack [70,83,84] is an evidence-informed, theory-based, gender-transformative (challenges gender inequalities relating to SRH) RSE programme designed to reduce unintended teenage pregnancy and promote positive sexual health. It aims to increase intentions to avoid teenage pregnancy and HIV/STIs by encouraging delayed initiation of sexual intercourse and/or consistent use of contraception and is designed to be delivered in educational settings. It specifically targets boys aged 14–15, however, it can also be delivered to girls and used in same-sex or mixed-class groups. It is designed to promote critical thinking about the social pressures that normally situate teenage pregnancy and its prevention as a female-only issue. Programme components include: an interactive film which tells the story of 16-year-old Jack, who has just found out that his girlfriend Emma is unexpectedly pregnant; classroom materials for teachers containing detailed lesson plans with specific classroom-based and homework activities; a 90-minute training session delivered by RSE specialists to teachers implementing the programme and parent components as described below. The JACK programme and Trial methods are described in full elsewhere [29,30] and more information about the project can found at www.qub.ac.uk/if-i-were-jack. The study was conducted in two countries, Lesotho and South Africa. Lesotho and South Africa differ in terms of culture, language and religion, but they both have very high levels of poverty, unemployment and gender-linked violence [36]. In Lesotho, the study was located in the Maseru district, which consists of the capital city, Maseru (population circa 330,000) and other surrounding peri-urban and rural areas. Lesotho is listed among the ‘least developed countries’ on the Development Assistance Committee (DAC) list of Official Development Assistance (ODA) recipients. Despite huge need in relation to SRH in Lesotho with services mostly centred in the capital, government expenditure on health overall is only around 5% of gross expenditure [37]. In South Africa, the study was located in the peri-urban informal settlement of Khayelitsha (population circa 500,000), which is situated on the outskirts of Cape Town. While South African law guarantees access to SRH services, there remains a large service gap with an economic status gradient evident in places such as Khayelitsha [6]. We chose to work in these countries for a number of reasons. First, they are home to some of the most vulnerable adolescents in Southern Africa who are very much in need of evidence-based SRH interventions. Second, we thought the differences between them (in particular culture, poverty levels, abortion laws and SRH services provision) would allow identification of key factors that can promote or inhibit adoption, implementation and evaluation during future phases of the study. Importantly, the project team in South Africa also had extensive experience working in both country sites. When compared to the UK, for which the original intervention was designed, South Africa is comparable in terms of reported gender equality indexes, while Lesotho shows stark differences. In 2020, the Global Gender Gap Report [36] ranked the UK 21st out of 153 countries, while South Africa was placed 17th and Lesotho 88th. Further, the report highlights that while 29% of women in the UK and 21% of women in South Africa report ever having experienced gender violence, the figure in Lesotho is 62%. Additionally, while unmet need for family planning in the UK stands at 6%, figures in South Africa (15%) and Lesotho (18%) are higher, as are incidences of maternal mortality (7 per 100,00 live births in the UK compared with 119 in South Africa and 544 in Lesotho). Due to the burden of HIV/AIDS in the region, it is imperative that SRH interventions targeting adolescents incorporate a focus on HIV reduction. For this reason, the adapted intervention will appoint equal focus on the avoidance of HIV/STIs and unintended pregnancy. While we recognise the importance of the full range of contraceptive methods available – the JACK programme encourages the use of condoms alongside another contraceptive method – we focus on condom use because it is the only method of contraception that provides dual protection from STIs and pregnancy. Our definition of ‘unprotected sex’ for this project therefore refers to sex without a condom (‘condomless’ sex). When we refer to ‘contraception’, this includes condoms and all other methods of contraception, unless otherwise specified. We used a systematic and collaborative process to assess SRH need and plan sociocultural adaptations of the JACK intervention for use in South Africa and Lesotho. We employed a participatory approach, working with local stakeholders to obtain contextually relevant information pertaining to the study objectives. This involved the establishment of separate adolescent, community, and expert project advisory groups (PAGs) in each country to consult on how best to: a) engage with the broader community for a successful project; b) address culturally-sensitive issues; and c) adapt, as needed, the JACK educational pedagogies for local contexts. Table 1 summarizes the demographic characteristics of the PAGs and study participants. Across both sites, we recruited a combined total of 53 PAG members and 77 adolescent and adult focus group participants. We employed convenience and snowball sampling to recruit focus group participants according to pre-specified inclusion criteria that ensured a mix in terms of gender, age and occupation. The PAGs acted in an advisory capacity only. While their opinions were considered during the IM process, no data were collected from them and their views are not reported in detail in any publication. Demographic characteristics of project advisory group members and focus group participants The study was approved by the Health Research Ethics Committee of Stellenbosch University [N19/07/081] and the Lesotho Ministry of Health Ethics Committee [ID 215–2019]. Prior to enrolment in the study, informed written consent was obtained from all adult participants. All adolescent participants provided informed written assent and informed written consent was also provided by their caregivers. In order to promote privacy, focus group participants were reminded that they should only share information that they were comfortable sharing publically and asked to verbally agree to ’group rules’ that everything said during the discussion should be considered private. In order to maintain confidentiality, participants were assigned a participant identification number and this number was used for all information collected. No other information that would disclose the participant’s identity was included in transcripts. All focus group participants received R160/M160 in vouchers as well as travel expenses to and from the focus group venue. The research team developed and adhered to a data storage and management plan, which detailed steps for ensuring data quality across researchers and countries and was approved by both ethics committees. We used a modified version of Bartholomew et al’s [33] Intervention Mapping (IM) approach to frame and inform the data collection and analysis processes. Key components of the IM-informed adaptation process is to conduct a needs assessment (Step 1) and then compare and contrast the components and theory of change from the original intervention with the needs of the new target population (Step 2). Divergences between the existing intervention and the needs of the new population indicate what changes need to be made to the original intervention informing adaptation, implementation and evaluation plans at Step 3. Figure 1 presents the modified version of IM used in this study. We conducted Steps 1 and 2 separately in both countries. Modified intervention Mapping process Following initial consultations, and positive feedback from PAGs regarding the potential suitability of adapted versions of the JACK intervention for use in South Africa and Lesotho, we conducted a systematic review of qualitative and quantitative literature on the determinants of condom use among adolescents in Southern Africa, and a series of focus group discussions with the adolescent and adult participants. The aim was to ensure that the adapted intervention would be based on a thorough understanding of the needs of the target population and current community capacity to implement it (Figure 1, Step 1). The findings of the systematic review, reported elsewhere [38,39] provided information on pertinent population and contextual issues that helped inform the needs assessment process and also informed, in collaboration with PAGs, the development of the semi-structured interview schedule used to collect primary data. We collected primary data from 12 focus group discussions (6 in each country), eight with 55 adolescents and four with 22 adult participants. Participants represented a convenience sample, recruited via requests for volunteers from non-government organisations on our advisory groups. To help ensure the credibility of the findings (47), the focus groups were facilitated by experienced research assistants who were provided with training to ensure they had adequate knowledge of the subject. Facilitators were also provided with extra training to help them address common challenges (e.g. dealing with embarrassment, ensuring privacy, encouraging the expression of difficult opinions) experienced in conducting focus groups on sensitive topics [40]. Further, the researchers resided in the communities where the study was conducted, and were first language speakers of Sesotho or isiXhosa, the predominant languages in the two sites. The researchers implemented reflexive journals [41] during fieldwork and had regular meetings with the project manager during the process, reflecting on experiences and preliminary findings. Following best practice [35], the focus groups used a flexible format to explore the factors that affect adolescent sexual behaviour and researchers spent time at the end of each session checking and confirming that the key messages they heard were accurate. During the focus groups, the facilitators spent 2–3 hours with the groups, using the pre-designed topic guides to ask about views on knowledge, attitudes, issues around self-efficacy, intentions, perceptions of risk, and perceptions of social norms relating to sex. They also explored the relevance of the Jack intervention materials with the participants and elicited suggestions on how to adapt them in the two African study sites. While data saturation was not a guiding feature of the research design, mainly because of time and resource constraints, we are confident that the rich data provided from 12 focus groups, triangulated with systematic review findings is adequate to support the potential transferability of findings [35]. To help ensure dependability of the findings [35], we prepared and adhered to a study protocol containing a detailed description of the methods and established an audit trail of the data collection and analysis processes. The focus group interview topic guides were developed in English by the local research teams, translated into Sesotho and isiXhosa, and pilot tested with the PAGs. The focus group discussions were conducted in the respective first languages of the participants in both countries. All focus groups were video recorded, transcribed verbatim in the language of the interview, and translated from Sesotho and isiXhosa into English. Two authors (SR, ÁA) submitted the transcripts to thematic analysis [42], coding the data to help identify patterns in the data and recurring themes and meeting regularly to discuss and confirm coding accuracy, thereby helping to ensure confirmability of the findings [35]. The analysis process followed Braun and Clarke’s [43] six steps. First, we familiarised ourselves with the data by reading the transcripts twice. Second, the transcripts were imported into ATLAS.ti 8 [44] to conduct initial coding. This involved coding sections of data using open theoretical coding linked to our overall research objectives of understanding the determinants of condom use and perceived intervention needs of the participants. Examples of the codes applied at this stage are provided in Table 2. At step three we examined the codes to see if they fitted clearly into a theme, for example, several codes related to how boys and girls were treated differently and these were combined into an initial theme of ‘gender inequalities’. At step four we reviewed the initial themes in relation to the overall data set and to see if they made sense. We followed the guidance of Braun and Clarke [43] in relation to asking questions about whether the data fits the theme, if the themes overlap, if there are sub-themes etc. For example, we considered the theme ‘lack of diversion for adolescents’ in detail and decided that there was enough data and differences for it to constitute a separate theme rather than sub-theme of barriers to condom/contraceptive use. Finally, before the sixth step of writing up the findings, we followed Braun and Clarke’s instructions to ‘define the themes’ or describe what they meant ‘in essence’. This involved, writing reflexive memos [41] while reviewing and coding the transcripts to help synthesise emerging themes across the research domains and paid particular attention to interrogating the data to test underlying assumptions to update the intervention’s theory of change model . In practice this reflexive writing about codes and categories was a part of the entire analysis process, often involving reflection on what was happening in a particular discussion or statement and comparing with other similar or different discussions or statements. The memos were often also used for the researchers to reflect on and challenge their own interpretations of the findings [41]. Codes, themes and sub-themes from focus group discussions Transcripts from each country were analysed separately and then combined in the final synthesis. Any similarities and differences between the countries were highlighted. Most of the resulting themes presented below are at the descriptive rather than abstract level. We agreed the presentation of semantic (rather than latent) level themes was in line with our aim to identify the explicit expressed needs of participants and would facilitate the synthesis of findings across the different data sources. As illustrated in Figure 1 (step 2), we used the findings from the needs assessment to conduct an IM exercise led by ÁA in consultation with the wider team and PAGs. This process relied heavily on data source triangulation (systematic review and focus group findings). The synthesis process began with a consideration of the findings from the needs assessment (review and focus group findings) and consultation with PAGs, and a whole-team discussion of these based on the questions below. The answers to these questions were used to update the project’s original logic model of the problem; developing matrices that specified changes to intervention goals, objectives, outcomes, underpinning theory and delivery methods; and producing a draft ‘theory of change model’ and draft adaptation, adoption, implementation and evaluation plans. These were finalized following consultation with advisory group members.
N/A
