The ethical ‐ legal requirements for adolescent self ‐ consent to research in sub ‐ Saharan Africa: A scoping review

Support for the enrolment of adolescents in research has been constrained by uncertainties in parental involvement, and the lack of clarity in the ethical and legal frameworks. We conducted a scoping review to examine articles that explored the opinion of scholars on the question of adolescent consent and conditions for parental waivers in research in sub ‐ Saharan Africa (SSA). Guided by the Preferred Reporting Items for Systematic Reviews and Meta ‐ Analyses (PRISMA) tool, we searched electronic databases (PubMed, EMBASSE, EBS-COHOST) and also reviewed the references of articles identified for additional relevant literature. We included full text English articles focusing on adolescent consent and parental waivers in SSA that were published between 2004 and 2020. We excluded studies focusing on healthcare, theses, and reviews. We reviewed a total of 21 publications from South Africa ( n = 12), Kenya ( n = 4) and Botswana, Malawi, Nigeria, Uganda and Zimbabwe ( n = 1 each). We identified four broad thematic issues: the current position regarding parental waivers and self ‐ consent; parental involvement in the consent process; the role of community approval or consent when adolescent self ‐ consent approaches were used; and complexities and ambiguities in legal requirements and ethical guidelines on adolescent consent. Our findings show inconsistencies and ambiguities in the existing legal and ethical frameworks within and across different countries, and underscore the need for consistent and clearer guidance on parental waivers and adolescent self ‐ consent. Harmonization of the legal and ethical frameworks taking into account varying contexts is critically important to ensure research on adolescents in SSA meets adolescents' specific unmet needs.

studies were screened independently, and later as a team.Differences and discrepancies were resolved by discussions until a consensus was reached.Following this level of screening, the remaining 92 articles were considered for a full text screening using the same strategy to identify a final list of papers meeting the study criteria.Out of 92 studies screened for full text, 71 were excluded because the articles included non-primary literature (n = 9); the population was either infants or above 18 years old (n = 8); were not from SSA (n = 28); and the focus did not involve adolescent self-consent, parental waiver, or parental consent/adolescent assent (n = 26) (see Figure 1).

| DATA ANALYSIS
Using a matrix (word document) B. N. and B. Z. categorized articles meeting the criteria into four main themes, (a) the current position regarding parental waivers and self-consent approaches; (b) parental involvement in the consent process; (c) the role of community approval or consent when adolescent self-consent approaches were used; and (d) and complexities and ambiguities in legal requirements and ethical guidelines on adolescent consent (see Table 1).Studies addressing more than one theme were charted in one category using a consensus approach.
We identified a total of 21 studies, including those from South Africa (n = 12); Kenya (n = 4) and Botswana, Malawi, Nigeria, Uganda and Zimbabwe (n = 1 each).The studies included: social science studies; willingness to participate (WTP) studies; simulated clinical trials; intervention programmes focusing on parental waivers and adolescent consent; as well as articles on the ethical and legal guidelines surrounding adolescent participation and parental waivers /parental consent.The current position regarding parental waivers and self-consent approaches 2006 PLoS Medicine 3 (7): 984-988   No guidelines for parental waivers.

| THE CURRENT POSITION REGARDING PARENTAL WAIVERS AND SELF-CONSENT APPROACHES
Source: Day et al. (2020).See note 1.
who are homeless.In South Africa, conditions for parental waivers included the sensitive nature of the study, older adolescents, 16 and older, minimum risks and community consent.There was no data on parental waivers in Botswana, Malawi and Zimbabwe (see Table 2).
We identified two Kenyan studies, where parental consent was waived and an adolescent self-consent approach was used. 6The studies involved hard to reach adolescents including orphans and vulnerable children and children living on the streets between the ages of 10-19 years.The researchers obtained community approval through consultations with the local leaders, and professionals including teachers, and social workers, working closely with the ethics committee to obtain a waiver of parental consent.The ethics committee found that the self-consent approach would not adversely alter the risk-benefit ratio for participants.The self-consent approach was justified on the basis that the studies could not have been carried out without the parental waiver.Involving this cohort was essential to the study objectives to identify adolescent health needs and broader challenges they face in the community.
We also found that a study involving an HIV vaccine trial in South Africa did not get approval from the ethics committee.The investigators set out to employ a self-consent strategy by including adolescents 14 years or older. 7Adolescents were to be recruited from pre-natal and family planning clinics without seeking parental consent.Consequently, the researchers had to change the age range of participants to those above 18 because the study posed more than minimal risk.Most of the clinical studies in this scoping review were WTP or HIV clinical simulation studies, which may reflect the difficulties in obtaining parental waivers, particularly for higher risk or interventional research. 8

| PARENTAL INVOLVEMENT IN THE CONSENT PROCESS
We found divergent views regarding the level of parental involvement in adolescent participation in research.These views were based on the type of study and potential to breach adolescent confidentiality, as well as adolescent-parent relationships.Our findings showed a growing recognition that adolescent research participation should be a joint decision-making process between the parents, or caregivers and the adolescents. 9This joint approach strengthens understanding of the study because both parties (parents or caregivers and adolescents) receive the same information, regardless of the consent approach.The role of parents in these instances is seen as the providing of support for their children. 10 A WTP study in South Africa showed that although adolescents reported that parents should provide consent for adolescent participation in vaccine trials, they expressed concern about a lack of control over access to their study results. 11Studies on adolescents' views about parental consent in Kenya showed heterogeneity across youth and their preferences regarding parental consent. 12Adolescents supported parental consent and their protective role as a safeguard in case something went wrong during the study and from community gossip about their involvement in the research study.
A randomized controlled trial (RCT) examining the efficacy of a behavioural intervention targeting adolescent sexual risk behaviours in Botswana, reported that adolescents were likely to report pressure from parents or relatives, and that younger adolescents aged 13-15 were likely to refuse. 13In this study, parents who provided consent for their adolescents during the RCT had mixed reactions when informed that they would not have access to a participant's results unless their children voluntarily shared the information with them or voluntarily involved them in the return of results process.Ultimately, the desire of parents to be involved and to access their children's HSV-2 results overrode the importance of protecting the confidentiality of the adolescents, breaching the adolescents' privacy.In the South African WTP study some participants were of the view that adolescents should have the autonomy to enrol without parental consent. 14Their rationale was that parental consent may prevent some adolescents from study participation if the parent was not readily available to give consent or refused.

| THE ROLE OF COMMUNITY APPROVAL OR CONSENT WHEN ADOLESCENT SELF-CONSENT APPROACHES WERE USED
We found that community approval of an adolescent-self consent strategy is an ethical norm unique to the South African framework. 15Although this was not a requirement in Kenya, we found two studies that used community consent and were granted a parental waiver by the ethics committee. 16The studies 6 involved adolescents living on the street and orphaned children, and parental consent could not be obtained.We also found that there are no guidelines or definition for community approval when using the selfconsent strategy.In the two studies in Kenya, community approval was obtained through consultations with the local leaders, and professionals including teachers, and social workers.
In most of the studies reviewed, community consultation was key in respecting and gaining access to the community, and adolescents. 17The communities' input was used to improve the informed consent process such as clarifying study materials and strengthening referral linkages in clinical trials. 18Community consent was therefore viewed as an appropriate and necessary strategy even before individual consent to extend protections from the individual to the community-level, and was considered as a potential alternative to parental consent in cases where parental consent was not feasible, or when the child is not well cared for. 19study in Nigeria showed that culture rather than the law carries greater weight.This was evidenced in the importance of consultations within the more immediate family circle.In this study it was reported that some parents wanted to talk with their families or respected people in their community before reaching a decision about providing consent for an adolescent to participate in a sexual and reproductive health research, especially when such research involves invasive procedures such as regular blood draws and vaginal examinations. 20This approach is in line with the view of other authors who argue that ethical guidelines that focus on parental involvement specifically for the purpose of enrolling an adolescent minor into a study fail to consider that adolescents are embedded in relationships with partners, peers, families and communities. 21ne of the authors defined the scope or breadth of community consultations.However, a study from Kenya cautioned against reliance on community leaders as this might marginalize individuals and groups who may not feel able to speak freely in group deliberations, and thus their opinions may go unheard in community discussions. 22

| COMPLEXITIES AND AMBIGUITIES IN LEGAL REQUIREMENTS AND ETHICAL GUIDELINES ON ADOLESCENT CONSENT
Most of the authors argued that existing ethical and legal norms act as a barrier in enrolling adolescents in clinical trials. 23The divergent approaches to consent taken in law and ethical guidelines add a layer to these complexities (Table 2).In South Africa, the divergent consent approaches in ethical and legal frameworks present challenges regarding enrolment of adolescents in clinical trials. 24The legal guidelines (National Control Act requires that the parent or legally authorized representative provide consent on behalf of a minor. 28y problems within these frameworks include the use of age as a proxy for maturity.Most authors argued that age is not always an appropriate proxy for adolescent self-consent and authors advocated that researchers and ethics committees should look at maturity, or mental capacity; and or use competency tests 29 to take into account the evolving maturity of adolescents, and subsequent participation in research.However, findings from a study in Kenya showed that some of the adolescent participants felt that adolescents 16 years and over were capable of making their own decisions about research participation, therefore should be granted parental waiver. 30 did not find clarity on the issue of assent from adolescents.
A study in Malawi reported the difficulties and the lack of clarity and guidance regarding age at which assent and consent should be sought. 31The authors highlighted the diverse household structures and adolescents' living arrangements including children living in boarding schools, or adolescents living on their own as a result of educational and employment opportunities.In this context, obtaining parental consent becomes difficult.

| DISCUSSION
Our review shows wide support for adolescents' participation in research; however, there are mixed views about the level of parental involvement in decision-making regarding research participation.Four broad thematic issues emerged.First, delineating the role of parents in the consent process during adolescent research 32 and finding the balance between the protective function of parental consent and its potential to act as a barrier to research. 33The absence of objective or biological markers to define when an individual becomes an adult and at what age they have the actual capacity to give consent underscores the importance for the field of bioethics to pay greater attention to adolescents.
Second, if parental consent is to be waived, it is not clear under which circumstances this would be ethical and what ought to be the procedural obligations that should be met. 34The study risk level was often found to influence the granting of parental waivers.This is demonstrated in the HIV vaccine efficacy trial in South Africa in which parental waivers were not granted because the risk level was more than minimal. 35Studies posing more than minimal risk, such as the South African clinical trial, used WTP due to the inability to obtain parental waivers.While WTP studies inform future recruitment and retention efforts, they are not a good predictor of future successful recruitment and retention of adolescents into HIV vaccine trials. 36In Kenya, the parental waiver was granted in two exploratory studies where the risk level was minimal and parental consent was not feasible. 37 Studies reviewed showed that there are no objective or biological markers to define when or what age an individual becomes an adult and at what age they have the actual capacity to give consent.We found that there is variability in when adolescents are permitted to self-consent and what conditions make it ethically acceptable.Kenya, Nigeria and Uganda granted parental waivers for emancipated minors including orphans and married adolescents.Furthermore, Kenya made provisions for children experiencing substance abuse and those F I G U R E 1 PRISMA flow diagram T A B L E 1 List of studies and thematic domains Third, support for wider community engagement and consultation as a protective mechanism for adolescent participation, and the lack of guidance on community consent, resulted in synonymous use with community participation.Fourth, there are inconsistencies and ambiguities in the existing legal and ethical frameworks within and across different countries.38Our study highlights gaps and ambiguities in national ethical and legal frameworks regarding adolescent participation in research.The authors in most of the papers argued that parental consent has the potential to act as a barrier in research involving sensitive topics such as sexual orientation, SRH research and HIV prevention studies.39This was demonstrated in the RCT study in Botswana, where the desire of parents to be involved and to access their children's HSV-2 results overrode the importance of protecting the confidentiality of the adolescents, breaching the adolescents' privacy.40We also found that cultural values play a major role in decision-making as shown by family and community participation in decision-making in many SSA countries.41Several scholars discussed the role of community consultations.Several argued that consultation and engagement of the community prior to research protocol submission to the ethics committee creates the opportunity to discuss the rationale underpinning the inclusion of adolescents in the research, and waivers of parental consent in studies where such is a requirement.42Community endorsement of research plans is perceived as being a major factor in research ethics committee considerations on whether to allow adolescents to provide autonomous consent for participation in a study.43Although our study focused on the SSA region, complexities in legal and ethical frameworks, which take different approaches regarding the autonomous participation of adolescents in research, have been reported in other places including North America and the United Kingdom.44While most studies from the SSA region underscore broader community engagement and consultation, such consultation is limited to the parents or immediate family in North America.International research guidance increasingly supports the proactive inclusion of children and adolescents in health research in recognition of the need for more evidence-based treatment.This is reinforced by international agencies including the Global Strategy for Women's,Children's and Adolescents' Health of the United Nations, and the Global Accelerated Action for the Health of Adolescents of the World Health Organisation, and the Global Accelerated Action for the Health of Adolescents of the World Health Organization.45Despite the growing interest and divergent views surrounding parental waivers, and adolescent self-consent in SSA, our study showed that there is limited research involving prospective adolescent HIV research participants in SSA.46 10 | STRENGTHS AND LIMITATIONSThis comprehensive review highlights differing approaches in adolescent participation, as well as emerging trends including broader community engagement in addition to parental consent in SSA.It is possible that our search did not detect all publications that covered issues relevant to adolescent research ethics, for example, due to inclusion only of studies conducted in English.By excluding studies from North America, we may have inadvertently omitted information from studies that included data about SSA.Although the majority of the studies propose community engagement and consultation, defining these concepts was beyond the scope of this review.11| CONCLUSIONOur findings show a complexity and variance in how adolescents are included in research without parental permission and underscore the need for consistent and unambiguous guidance on parental waivers and adolescent self-consent.Harmonization of the legal and ethical guidelines taking into account varying contexts is critically important to ensure research on adolescents in SSA meets their specific unmet needs.
Lindgren et al., op.cit.note2; World Health Organization, op.cit.note3.legal guardian to give proxy consent despite the reality that many adolescents do not live with biological or legal guardians.While there no specific laws that are dedicated to research involving children in Zimbabwe, the Medicines and Allied Substances