Ethical challenges confronted in non-clinical,public health research with young people in England

Abstract

This opinion piece emerged from a collaboration of mental health researchers working on the National Institute of Health Research 3-schools mental health programme in underserved communities. The aim is to encourage debate and reflection on the challenges encountered with university research ethics committees when undertaking qualitative research with vulnerable young people. We explore the tension between principles of safeguarding and protection of research participants, on the one hand, and the potential for this oversight to become obstructive and thus to effectively stifle the voices and experiences of an already marginalised population, on the other hand.

Overview

The impetus for this commentary piece came from a collaboration of mental health researchers working on the National Institute of Health Research 3-schools mental health programme in underserved communities ¹ , who met periodically to share some of the barriers and facilitators to good practice. It draws on experiences of researching vulnerable groups of young people (YP) aged 11–18 years. This paper explores the challenges encountered in the implementation of non-clinical, public health research involving YP. It highlights specific areas relating to both university Research Ethics Committees (RECs) and the NHS Health Research Authority (HRA) ethics approval that, as researchers conducting health-related social research, we have found challenging. We believe that greater reflexivity, transparency and discussion in some of these areas may lead to improved processes. The paper focuses on five key areas: (i) consent; (ii) issues of flexibility in qualitative research design; (iii) transparency, accountability and communication; (iv) alignment of ethics requirements between institutions; and (v) representation on RECs. We conclude with practical recommendations directed at ethics committees.

The paper interrogates the tensions between the safeguarding and protection of research participants, on the one hand, against the potential for these principles to become obstructive, effectively stifling the voices and experiences of groups of already marginalised groups. Much debate relating to ethics is informal and goes unreported (Dyer and Demerit, 2009), and there is value therefore in articulating some of the challenges faced. Ethical aspirations on paper (procedural ethics) do not always mirror ethical practice on the ground; indeed receipt of ethical approval is no guarantee of sound ethical research (Carter et al., 2017). We propose that YP with capacity, who want to be engaged in research, including those deemed vulnerable, should be allowed the opportunity to have their voices heard and for their views and experiences to be fed into policy and practice. The unintended consequences of REC requirements can make research with these groups burdensome or even impossible. Further, we consider whether ethics committees are reflexive enough in relation to their own positions of power, knowledge and reach.

Whilst there is a significant literature looking at ethical challenges when undertaking qualitative research with YP (Cavazos-Rehg et al., 2020; Cullen and Walsh, 2020), there is a gap in practical recommendations targeted at RECs considering public health research. The intention, in setting out clearer guidance, is to promote a more effective, coherent ethics process to support public health research with vulnerable YP.

Background

This section locates RECs within a socio-political context and provides an overview of the ethical issues encountered in our work. Research ethics committees emerged as a response to various historic unethical research practices often reflecting an imbalance of power of professionals or corporate bodies over marginalised or vulnerable individuals. Whilst prominent examples of abuse often focus on inappropriate recruitment to clinical trials, there are also concerns about the recruitment of vulnerable individuals into qualitative research. Collection of data, for example, through interviews or observation, carries a number of risks including: dangers of misrepresentation; breaches of confidentiality/anonymity; the potential to divulge information respondents later regret; and safeguarding duties (Sun et al., 2023). Research ethics committees, set up to address these concerns, were driven primarily by the ambition to ensure the protection of human participants involved in research studies with secondary aims: to uphold academic standards and integrity; consider societal benefits; and protect the researcher (AREC, 2013).

Research ethics committees have responsibility to consider outcomes for participants deemed vulnerable. The concept of ‘vulnerability’ refers to groups considered at greater risk of suffering harm or exploitation arising from factors such as age (e.g., children or elderly), cognitive limitations, physical or mental disabilities, socio-economic status, cultural background or being in a dependent or subordinate position. However, protecting the rights of individuals must be weighed up against dangers of excluding their voice and thus contributing to the perpetuation of health and social care inequalities (Bonell et al., 2023; Houghton, 2015; Liu et al., 2017; Shaw et al., 2015; Spence et al., 2015). For YP to be central to the planning and delivery of services, it is imperative their voices be heard. To facilitate this, we must consider the role of ethics in the inclusion and exclusion of vulnerable groups, particularly those currently under-represented by existing research.

Within the last decade, a plethora of guidelines relating to research with YP have been issued, including BERA’s (2018) updated guidelines, RCUK Good Research Practice (February 2013), ESRC/UKRI (updated Nov 2022) and NSPCC(2023). Many of these guidelines address ways to reduce barriers to participation and improve the accessibility of the research through consideration of factors such as ethnic and gender matching, acceptability of location and modifications to research tools. However, whilst advice is given to improve access, the imperative for RECs to address the principle of equity in the research sample is often not addressed as a central component.

Whilst the establishment of rigorous systems of research and ethical governance provides an important check to maintain high levels of professionalism and the safeguarding of research participants’ interests, the system of checks now in place has been the focus of growing criticism around the bureaucratisation of the research ethics process (Brown et al., 2020; Juritzen et al., 2011; Törrönen and Petersen, 2021). Arguments have been advanced for the adoption of a broader perspective on the place of power within the REC (Juritzen et al., 2011). Whilst historically RECs were set up to protect vulnerable members of the public, the committee, consciously or unconsciously may itself use institutional influence without due regards for those in less powerful positions. Research participants, it is said, could be harmed, not only by coercive research practices but also by being subjected to sometimes over-controlling behaviours of ethics monitoring bodies. According to this analysis, RECs are one of the players exerting power in the relational interplay between researchers and participants and cannot therefore be ‘exempted from critical reflection and discussion’ (Juritzen et al., 2011, p. 641).

Commentators have observed a shift from an ethical code based on trust in the researchers’ professionalism, self-regulation and discretion, to an ethical code which is more rule-based. RECs, for some, appear to be more concerned with protecting the interests of the research institutions and forestalling loss of reputation, legal disputes or claims for compensation rather than prioritising the welfare of those they purport to protect (Cheek, 2005; Guishard and Tuck, 2013). Others refer to the growing chasm between the ‘institutional discourse of ethical research and the “real world” of qualitative research’ (Halse and Honey, 2007, p. 336; McDermott et al., 2019). The interests of ethical research can become ‘colonised or extended by other discourses’ (Halse and Honey, 2007, p. 340), and in the case of research with YP, ethical requirements are seen to add to the long list of national and local regulations protecting access to YP and their associated data. Inevitably however, the conundrum RECs face is that whilst the majority of researchers act with discretion and integrity a minority do not and in such cases regulatory systems need to be in place.

The probable reality is that most RECs are simply trying to do their best in difficult circumstances. As RECs attempt to balance increasing bureaucracy, institutional interests and trust in researchers, it appears that the exclusion of YP voices may be an unintended consequence.

Consent

Breaches of consent have a long history, with ill-effects felt particularly by those whose power and agency have been removed. Exploitative practices relating to minoritised communities, those lacking capacity and children are all too common. Yanchapaxi et al. (2023, p. 114) evoke the voice of black, indigenous women, where consent practices are often experienced as ‘extractive and superficial’. Lack of meaningful consent, both formal and operational, is experienced as violence, impacting on the multiple domains of body, story and artefact.

Children under the age of 16, who hold a different legal status from adults, face particular consideration in relation to consent and may be viewed as unable to provide valid consent without the authority of an adult guardian. Yet, Juritzen et al. (2011) argue that in some instances the suggestion that the inclusion of children, in the absence of adult consent, is ethically unsafe, creates an iniquitous power imbalance. The execution of paternalistic power to exclude marginalised groups may be as harmful as some of the coercive powers RECs set out to guard against (Juritzen et al., 2011, p. 646). Central to the imperative to include those whose voices are seldom heard in research is the challenge of identifying methods of robust and meaningful consent.

There is currently no clear age-related guidance on the issue of age of consent for participation in non-clinical research though the point of transition to secondary school (age 11–12 years) is generally accepted as an age where seeking individual consent from YP is deemed appropriate (Bonell et al., 2023). A systematic review of ethical issues in participatory research concludes that basing consent on biological age alone is problematic and that YP should be involved iteratively in the process of consent (Cullen and Walsh, 2020). A study examining the capacity of YP to make informed decisions relating to theoretical research concluded that from the age of 14/15 years YP's decision-making was sufficient to allow for informed consent (Bruzzese and Fisher, 2003). Moreover, it is argued that whilst parental consent should be encouraged, it is not a legally binding requirement – though it is accepted that thresholds for making ethical decisions may vary with each case. Worthy of note here and an interesting comparator, is the age of criminal responsibility which in England stands at 10 years old; lower than most other countries across Europe (Gov.uk, 2023).

In cases where parental consent is sought, a distinction may be made between opt-in consent (where the child cannot participate without receipt of signed parental consent) and opt-out (where the child can participate should they so wish unless the parent has expressly stated that they do not want their child to take part). The different options adopted may reflect the perceived level of risk in relation to themes and content of research.

Clinical trials, involving the use of medications or clinical interventions, require that parental consent is sought for participants under the age of 16 years (NIHCR, 2023). For public health interventions or observational research however, procedures are instead governed by common law. Particularly pertinent is the Gillick ruling (Children's Legal Centre, 1985) which states that where the cognitive capacity of minors (under the age of 16) is deemed sufficient to understand the tasks they are being asked to participate in, they should be allowed to give their own informed consent. The UN Convention on the Rights of the Child (UN, 1989) supports the involvement of children and YP and legislates that they are entitled to express their views and that their best interests must be prioritised in any research undertaken, including being given an opportunity to provide fully informed consent. This Convention is an international treaty and legally binding in the UK.

Criticism that the process of informed consent is not always appropriate or transparent has fuelled growing body of work exploring practice through the use of creative, participatory techniques (Gubrium et al., 2014; Ruiz-Casares and Thompson, 2016). Work to improve the process of consent with children, for example, engaging children in participatory visual techniques to promote engagement and understanding, has been shown to facilitate better cross-cultural communication. Such techniques explicitly address power imbalances, as well as cultural considerations which may prove coercive when adults in positions of power seek participation from children. Switzer (2020), working with YP in a peer education, harm reduction programme, refers to instances where YP chose not to consent, making the case for non-participation. Young people should not be seen as wilful or deviant for choosing not to engage in participative research, asserting they should be enabled to navigate participation on their own terms (Switzer 2020). Participation is a measured decision, central to which should be a clear assessment, on the part of the YP, of the risks and benefits involved.

Our experience

Our experience of applications to RECs for research involving secondary school children aged 11–16 has been that the REC initially questioned the assumption that children should be allowed to provide their own informed consent, instead suggesting that consent should be sought from someone with parental responsibility or a legal guardian. However, while parental consent may be necessary in some research depending on the topic, age group and specific population, we argued that insistence on parental consent for most secondary school children does not align with common law. Additionally, in some cases, we suggested that such a position effectively silences the voice of the most marginalised, particularly, as was often the case, in families where parents are disengaged or absent. The necessity of parents or guardians to provide opt-in consent would have been an unjustifiable incursion on the autonomy of YP and would particularly undermine the potential participation of those from socially disadvantaged groups.

As well as undermining YP's autonomy, the requirement to seek parent/carer active opt-in consent may compromise the integrity of research (Bonell et al., 2023; Flicker and Guta, 2008). A meta-analysis, investigating the effects of requiring parental consent on studies examining adolescent risk behaviours concluded that the use of such methods creates a sampling bias rendering the data unrepresentative of the population being surveyed (Liu et al., 2017). Results show that studies requiring active parental consent recruited significantly more females than males and less ethnic minority students.

The obligation to seek parental consent may also result in under-representation of YP with lower educational attainment, those from lone-parent and socio-economic disadvantaged families, and those engaged in violence (Shaw et al., 2015; Spence et al., 2015). Furthermore, response rates are significantly reduced in studies where parental consent must be sought (Courser et al., 2009; Eaton et al., 2004; Pokorny et al., 2001).

An unintended outcome therefore, of requiring parental consent, appears to be that disadvantaged students are more likely to be excluded from the sample, undermining the ability both to provide meaningful data for this group and to explore how the intervention being investigated impacts on health inequalities (Bonell et al., 2023). Our research did not involve primary age children; we acknowledge that for this group appropriate processes for consent will be more difficult, based on the content of research, and may vary with individual cases.

Flexibility in research design

Research ethics committees require that the ethics application detail clear plans for research methods including sampling, recruitment, participant numbers, interview style, interview guide and analysis. The undisputed imperative to safeguard research participants and privilege research quality favours certainty in design. In qualitative research, there are particularly heightened concerns that children need to be protected from incursions into their privacy or from risks related to disclosure or coercive practice. Yet qualitative research is less clearly bounded than quantitative and has the potential to stray into areas not necessarily flagged in the topic guides. Indeed, flexibility is a central component of many qualitative research methods, particularly when working with hard-to-reach or vulnerable groups where creative or evolving methods, responding to particular situations as they emerge, may be desirable. Such practices do not sit easily within REC frameworks which have primarily been built around quantitative methods (Haggerty, 2004; Halse and Honey, 2007; Hammersley, 2009). The research process is characterised as linear, unchanging and sequential. Flexibility in design may be equated to a lack of rigour, consistency and coherence of design (Holloway and Todres, 2003). These assumptions, it is argued, go counter to many qualitative methods (both established and emerging) which may favour participative, ethnographic or creative, less-conventional narrative methods (Lincoln and Tierney, 2004).

Research ethics committees tend to focus on specific requirements relating to study participants, overlooking the importance of the evolution of the research process. The often long and painstaking process of building trust in the community with stakeholders demands significant relational work, with favours reciprocated, confidences shared and a tapestry of connections built. In research of this type, the interconnections between the research questions, the theoretical perspectives, empirical findings and results develop in dynamic interplay throughout the process (Juritzen et al., 2011). Qualitative research is exploratory in a way that quantitative research is not and connection between parts of the process are not always linear. Its methods may need modification as a result of findings emerging from the analysis (especially in the case of grounded theory approaches).

Our experience

The dilemma we experienced when submitting REC applications lay in having to counter the requirement from the REC to provide sufficient detail of the methods we intended to use throughout the study, with the desire to subject these methods to an element of iterative testing and scrutiny. Specifically, the REC generally requires a detailed description of methodologies and plans prior to the collection of data; an injunction on any deviations from these plans; the use of standardised procedures for obtaining consent; prior approval of questions in interview guides and detailed regulation of observational studies undertaken (AREC 2013). In our experience of working with underserved populations, particularly YP, they are more responsive to less structured data-gathering processes, particularly those focusing on more relational aspects of engagement. It may be that common conceptions of what constitutes data and how best to gather consent need to be re-examined with a view to understanding the culture and perspective of extremely marginalised sections of the population.

Transparency, accountability and clear communication

Whilst there is an imperative in all research to clearly articulate the aims, goals and associated risks – when working with young and potentially vulnerable participants it is vital that this stage is entirely transparent. Ethical procedures must be an iterative and on-going part of the process – rather than a one-off phase, negotiated at project outset and then put aside (Gubrium et al., 2014). Transparency is also key, not only in relation to research in practice but also at the level of organisational research governance. Scrutiny of leading social science research institutions revealed that publicly available information on ethics review and governance procedures was often hard to locate and highly variable in terms of detail, format and procedures. This was particularly the case in relation to children deemed vulnerable (Parsons et al., 2015).

Alignment of ethics requirements between institutions

Public health research often works across organisational and disciplinary boundaries. The practicalities of this mean that it is often necessary to negotiate a number of ethical and R&D layers, each requiring slightly different processes and requirements across NHS, social care, education and voluntary organisations. The process is often time-consuming, placing a burden not only on the researchers but also participating organisations with each individually repeating similar procedures to assure themselves of the ethical integrity of the research. There have been calls for less duplication of effort and a more streamlined ethical process. Further challenges emerge in instances when contradictory advice is offered, where there is a lack of clear guidance or a sense that at times the stipulations of different bodies are not aligned (Thompson and France, 2010).

Representation, training and support of ethics committees

A framework, put forward by The Association of Research Ethics Committees (AREC, 2013), designed to support Highter Education Institutions (HEIs) in developing rigour and quality in their policies and procedures to manage ethical issues, recommends that RECs should include members from a wide range of disciplines, including representatives from outside the immediate faculty or academic unit as well as ‘lay’ or external members. AREC also advises that committee members should be offered systematic training and that their role should be compensated in the form of reduced teaching commitments. Yet despite such guidance, significant variation in practice is reported in committees across the country (Trace and Kolstoe, 2017).

Our experience of alignment and system competence

Research ethics committee guidance appeared lacking or unclear on necessary processes for some qualitative or ‘light touch’ quantitative research such as an online survey. One study struggled with apparently contradictory advice from the study sponsor and the NHS representative around protocols for ethical approvals for research within GP practices. For example, lack of clarity over whether participant identification centres or research sites were needed created delays, with multiple meetings between researcher, Sponsor and Clnical Research Network (CRN) study support. Apparent differences between documented online guidance provided via the HRA ethics website and ‘accepted practice’ as understood by the study Sponsor for research in primary care made it difficult for researchers to resolve queries, and applications had to be rewritten on multiple occasions. All of this caused time delay, inefficiencies and ultimately created barriers to conducting engaged and reflexive research with a currently underserved population. Systems where the understanding of ethics processes of one organisation does not align with another (in this case NHS ethics and the research sponsor) and where ‘grey areas’ exist within the guidance, create extra work for study Sponsors, researchers and practitioners working on the ground.

Whilst theoretically, it is possible to put in amendments to RECs once permissions have been given, time and resource constraints often mitigate against this. A scoping phase, used to help test and develop appropriate research methods and data-gathering tools, is often seen as a useful means to ensure that the main phase of research uses sensitive and appropriate instruments. Yet, given the time frames involved in putting in two ethics applications, this staged approach often proves prohibitive.

The time frames from ethics submission to securing a favourable opinion, amongst projects submitted by the authors of this paper varied between 13 and 26 weeks for university RECs and six months in the case of NHS SRA approval. Until ethics approval is secured no research can take place; therefore, within the two-year time frame of the projects reported here, applying twice for ethics is generally prohibitive, or if done would certainly impact on research outputs. Furthermore, even when permission is granted for the main phase of fieldwork, using a reflexive approach to the methodology often entails the need for a number of tweaks being made to project design. In one case, for example, the decision to extend the targeted age group from 11–18 to include 18–25-year-olds was seen as a substantial amendment, with such amendments each taking at least a month to process. The limited time available, therefore, once amendments have been processed, within a time-boundaried research programme, begins to have a significant impact on the quality and depth of the research. Time for the building of trusting relationships is eroded which in turn affects the quality and integrity of the outputs.

Discussion

Reflecting on the experiences reported in this paper leads us to conclude that despite a sizeable body of literature from public health and related social sciences raising concerns about research ethics, work still needs to be done to develop more streamlined, consistent and equitable practice. Research ethics committee processes within the field of health sciences were developed primarily to oversee quantitative research and clinical trials, yet the ethical requirements for non-clinical public health research (quantitative and qualitative), even when covering non-controversial topics, often appear disproportionately high.

Experiences presented here suggest a tendency of RECs to view children as lacking agency, autonomy, capacity and competence – a practice which leads to vulnerable children being further silenced and marginalised. We suggest that YP often demonstrate a complexity of understanding and insight beyond the expectation of their linear age (Singh 2007) and, as long as they are deemed to have cognitive capacity in line with what is being asked of them, should not be prevented from sharing their views, experiences and insights. If we are to carry out research related to YP's health we must value and honour their perspective and strive to continue to develop more flexible, sensitive methods to incorporate their voice into our work. We must also be mindful of, and facilitate the ability of YP to choose not to participate. Moletsane et al. (2021), in their account of transnational participatory visual research with girls, highlight the tensions in recognising the complex challenges faced and the need to safeguard, whilst acknowledging the imperative to nurture the voice and agency of YP.

Qualitative research does not always fit seamlessly into the ethics process – and the REC system is experienced by some as risk-averse and paternalistic. Ensuring that RECs are populated with members holding different specialisms across both qualitative and quantitative methods, or at least having specialists who can be referred to as required, may help guard against the appearance that certain methodologies face tougher or disproportionate levels of control. As it stands, the failure of some forms of qualitative research to offer certainties may result in such studies being perceived as lacking in rigorous ethical standards. This reinforces a methodological orthodoxy that may place unnecessary constraints on academic freedom, restrict access to groups who would benefit from their views and experiences being heard and impact on diversity in the choice of methodologies (Juritzen et al., 2011). Moreover, by focusing attention on lines of enquiry which may be better suited to clinical research, there is a danger that RECs may detract researchers from other more important ethical issues including for example: treating respondents with empathy; spending time talking to the respondents to ensure they have really understood the research, rather than focussing on the signing of densely typed forms; responding with sensitivity and reflexivity to participant needs; or securing meaningful follow-up support.

Addressing issues of flexibility of response, we suggest that: (i) RECs clarify in their procedures that qualitative research guides can legitimately have themes and probes rather than verbatim questions and (ii) RECs need to be reconfigured or better resourced, in order to respond faster to amendments. Taking these steps will reduce significant delays which may ensue when studies require often minor modifications following interim analysis.

Good ethical practice must be the cornerstone of research, with the welfare of the population at the forefront, yet we have found that at times the bureaucratisation and rule-based REC processes lend themselves to a ‘tick box’ response, where the researcher may become more concerned with getting the project through ethics, rather than meaningful participation in the process. The ultimate outcomes may be contrary to the intentions of the REC; researchers become more concerned with what will get them through the deliberations of the ethics committee, rather than with what constitutes appropriate ethical practices with regards to the participants. We also note, as have others (Guishard and Tuck, 2013; Törrönen and Petersen, 2021) the disincentive effect that rigid review systems may have on more participative, innovative and creative methods of research. It may be that if RECs were to review qualitative research projects via a different pathway to quantitative/mixed methods projects, a system could be developed where panels can become more responsive to proposed changes in method.

Whilst we do not contend that the process of research is universally beneficial to the target group, indeed there are valid concerns and challenges relating to interviewing marginalised and or vulnerable YP (Guishard and Tuck, 2013), we suggest however, that at times the obligation to ‘do no harm’ has led to over-protectionism which may trump the goal of doing good. We recognise and strongly support the need for regulations to be in place and agree that seeking consent from YP must be done with care and regards to well-being and justice. Nonetheless, we suggest that many YP not only benefit from having their stories heard (Halse and Honey, 2007) but also from having their voice represented in public service planning and design. The desire of the REC to protect from harm may be to the detriment of those who, as a result, are excluded from participating in research. The obstacles put in the way of this type of research may also dissuade researchers both from embarking on any work with YP under the age of 16 and from developing new creative, participative methods allowing YP to become an active part of the research rather than its passive recipients (Lincoln and Tierney, 2004). Paradoxically, work done in the name of ‘public involvement’ is not regarded as research and therefore not subject to the same ethical regulations yet may involve similar methods of interaction. Learning from these disciplines and examining the ethical principles followed may inform the current debate in research (Hagell and Benniche, 2022).

Finally, it may be worth considering the requirement for parallel ethics approvals when working with local statutory and voluntary organisations – many of whom have their own set of ethical requirements. Aligning these processes in a way that will provide the assurance that the required ethical standards are being met should help streamline processes and reduce replication.

Recommendations to RECs

• Decisions around parental and child consent should be aligned with the law and RECs need sufficient training to ensure this is the case.

• Those reviewing qualitative research applications should be aware of the need for flexibility in research design. For example, proposed ‘themes’ may be sufficient in topic guides rather than specific questions.

• To improve transparency and accountability, RECs should have a clear rigorous process where the views and experiences of those submitting applications are regularly sought.

• Consideration should be given, at the macro level, to align ethics requirements between universities, NHS, statutory and voluntary organisations. Ethics committees across different sectors and settings may benefit from consolidating some of the work required to provide safety and quality assurance and thus reduce wasted resource through replication of processes.

• Research ethics committees should ensure that each committee represents the necessary diversity of expertise including experience in emerging methods of creative and participative research.