Adolescent and young adult patients as co‐researchers: A scoping review

Abstract Background As part of a research project aimed at evaluating a hospital‐based adolescent transition programme, we asked ourselves what is known about the ethical and methodological challenges of research involving adolescent patients as co‐researchers. The aim of our review was to summarize empirical evidence and identify knowledge gaps about the involvement of young patients as co‐researchers. Methods We conducted a scoping review through searches in MEDLINE, EMBASE, PsychINFO, AMED. Results We found reports of young patients being actively engaged as co‐researchers in any stage of a research project, although commonly they were not involved in every stage. Including young patients as co‐researchers is resource demanding and time‐consuming. Involving young patients as co‐researchers contributes to the fulfilment of their right to participation and may improve the relevance of research. Benefits for the young co‐researcher include empowerment, skills building and raised self‐esteem. Few authors go into detail about ethical considerations when involving young co‐researchers. None of the included articles discuss legal considerations. Discussion and conclusion No lists of recommendations are given, but recommendations can be deduced from the articles. There is need for time, funding and flexibility when including young patients as co‐researchers. Knowledge gaps concern legal and ethical dilemmas of including a vulnerable group as co‐researchers. More reflection is needed about what meaningful participation is and what it entails in this context. Patient or Public Contribution This review is part of a research project where the hospital youth council has been involved in discussions of focus area and methods.


| INTRODUC TI ON
The idea of citizen participation is a little like eating spinach: no one is against it in principle because it is good for you'. 1 What Arnstein pointed out in 1969 was that people were talking the talk, but were not necessarily willing to walk the walk of true citizen participation. It was a question of empty ritual versus real power to influence. 1 Arnstein's original typology of the eight-rung ladder of participation illustrates different forms of participation. The ladder is frequently referenced and has been adapted by many, also to health care. 2,3 It is now usually referred to by five levels: (a) information, (b) consultation, (c) advice, (d) collaboration and (e) control. 3 Only the two last levels-collaboration and control-refer to approaches where power is truly shared.
We are seeing an increasing focus on patient participation in health research, with an increasing number of national guidelines being issued, for example in the UK, 4 in Norway 5 and in Belgium. 6 However, as Malterud  for the project and to provide training and support to the patients so that they are sufficiently prepared. Furthermore, one should continuously monitor and provide feedback on the interactions within the project and be prepared to refine procedures as necessary. 9,10 As part of a research project evaluating a hospital-based adolescent transition programme, 11 we asked ourselves how the ethical and methodological challenges identified in research with adult co-researchers apply when adolescent patients are co-researchers.
Children up to 18 years of age should be given the right to participate in research within the overall framework of human rights.
Furthermore, the ladder of participation has been adapted also to children's participation. 12 14 concern research projects where young people were consulted or allowed to give advice, thus referring to levels 2 or 3 of the participation ladder and not to co-research. A recent scoping review by van Schelven et al 15 also points out that more research is needed to expand the evidence base of involving young people with chronic conditions in projects regarding their health and social care.
The aim of this review was to summarize empirical evidence and identify knowledge gaps about the involvement of young patients as co-researchers. By co-research, we refer to full participation in the research team, that is levels 4 and 5 of the simplified participation ladder. 3 Our review aimed to address the following questions:

| ME THODS
To explore these topics, we performed a scoping review following the outline of the process as described by Peterson et al 16 A scoping review gives an overview of a broad topic 16,17 and is a good method for exploring new topics. 16

| Data collection
A search of databases (Table 1)  In stage two of the process, the first and last author screened the abstracts. The following inclusion criteria were used: articles concerning patients 12 years and above; patients are co-researchers; the research topic is in relation to health care and health systems research; describes 4th and 5th level of participation according to the participation ladder as amended by Teunissen and referenced by de Wit et al 3 ; and publication date 2000-current. The first author and one co-author performed the full-text review. In case of disagreement, a third author was consulted. All co-authors were asked to consider whether they knew other publications that should be reviewed. The reference lists of the included articles were handsearched. This did not lead to any further inclusions. The search was updated in June 2020, and an additional 5 articles were included.
The characteristics of the included articles are presented in Table 2.

| Quality of reporting
Although not mandatory for a scoping review, we chose to perform a quality assessment of the included articles. As argued by van Schelven et al, it helps to put the results in context. 15 The quality of reporting was assessed by the last author. Assessment of the article in Spanish was done by a Spanish-speaking co-author. The quality appraisal was intended as indicative and was not part of the selection process. A protocol designed by Kmet et al 18 was used, which includes ten items. Details are outlined in Table 3. Each article was assessed according to the fulfilment of items and was rated as 'yes' (=2), 'partial' (=1) or 'no' (=0) on each item. Each study was assigned a summary score between 0 and 1 (with higher scores indicating better quality of reporting) by adding the scores of all individual items and dividing by the maximum possible score (20 for qualitative studies).

| Data analysis
An extraction grid was developed. Categories were created to reflect the research questions and the stages of a research process. The extraction grid was tested by the first and last authors. Amendments  or adolescent* or youth* or teenager* or teen* or "young people") adj15 (co-research* or "participatory research" or "action research")).ab. AND exp *Hospitals/OR hospital*.ti,kf,kw,id. OR exp *Health Services/OR (health service* or health care setting*). ti,kf,kw,id. limit to (danish or english or french or italian or norwegian or portuguese or spanish or swedish) limit 22 to yr="2000 -Current" remove duplicates TA B L E 1 Overview of search string were made before application to the included articles. An integrated form of analysis was performed 19 with the help of the extraction grid.

| Characteristics of the included studies
None of our 14 articles were included in the review by Haijes & van Thiel. 14 Two overlap with those included by van Schelven et al 15 Four articles were from the mental health context. [20][21][22][23] Three reports included patients with somatic conditions followed up in hospital. [24][25][26] Eight articles were from Northern America, three from the UK, two from the Netherlands and one from Colombia. Twelve projects included adolescents aged 17 years and above. Kramer and Schwartz 27 included co-researchers aged 14-21, and Moules 26 had co-researchers as young as 12-16 years of age. All studies reported on projects with a mainly qualitative approach.

| Quality of reporting
Quality assessment scores ranged between 0.65 and 0.95. All scores but three were over 0.8, which indicates a high quality of reporting for most articles (Table 3).

| Participation may happen at any stage, but seldom at every stage
There is evidence of adolescent co-researchers being involved in all six stages of research (Table 4), but only two projects report on coresearchers being part of all stages. 20,22 In the preparatory phase, there is focus on research training. Van Staa et al 24 did a field trip with the co-researchers to a newspaper to gain insight and discuss the interview protocol, while Dunn 20 reports on formal literature search training. Some do not go into details on the training given 26,28,29 while Lincoln and colleagues 22 report on extensive training and formal testing of knowledge. It is argued that research training addresses the imbalance of power between researcher and co-researcher and that the training provides the adolescents with a wider knowledge base. 25 Four articles report on involvement in the recruitment phase.
This is the area with least reported participation. These articles report on co-creation of the recruitment tool 21 or participation in design and dissemination of the recruitment material. 20,30 Lincoln et al 22 report that the recruitment strategy was discussed by 'the whole research team'. Most articles report on involvement in the design phase including project development meetings, 25 protocol writing, 30-32 decision on methods, 22 discussions of research questions, 26,29 development of questions for interviews, focus groups or preparation of workshops. 20,21,23,24,[26][27][28][29] Nine articles report on participation in the data collection phase. Young researchers were involved in data collection such as interviews or focus groups. [20][21][22][24][25][26][27][28][29] In the case of focus groups, they either led or co-hosted the group.
All articles report on involvement in the analysis phase. Some report briefly that the analysis was done by the team, 21,23,26,27,29 that adolescents were involved in 'continuous review of all work ' 25 or that the results were discussed. 33 Others are more specific, for example describing that the co-researchers transcribed their own interviews after which the group as a whole performed the analysis. 22 Mitchell et al 28  to the dissemination stage, some simply state that dissemination was done 'as determined by the team' 22 or was shared. 28 Others report co-presentations in different forums, including co-design of conference posters or conference presentations. 29 Co-writing publications ranged from co-writing the published article 23,25 to popular or 'unspecified' publications. 20,24,30

| Methodological, legal and ethical issues and challenges
The reasoning behind choice of methods revolves around the importance of involvement to better meet the needs of adolescents 21 and arguing the need to do research with rather than on patients. 20,24,27 It is argued that participatory or qualitative methods are suitable because they give rich opportunities for adolescents to voice their opinion and share their experience. 21  challenges or the role of parents in co-deciding or giving permission for participation as co-researchers.

| Benefits and drawbacks of involving young patients as co-researchers
Benefits for the young co-researchers, the professional researcher and the research process are discussed. Arguments used are primarily for the research itself, including relevance of questions, methods and findings, as well as recruitment. 22

| Recommended strategies for meaningful involvement
No article draws up a list of recommendations, but the importance of clarifying roles has been emphasized, 30  processes that are generally more time and resource demanding than more traditional forms of research. A second recommendation is to address, in advance, questions of who will benefit from the research. 24,30 There are three particular recommendations to safeguard the interests of the young co-researcher. The first is proper training. 25 The second is to develop a self-care plan so that possible challenges are discussed and addressed ahead of time. 22 Finally, it is recommended to be flexible, that is, that the young co-researchers should be made aware of the possibility to step out of the research, and/or choose which stages of the research to participate in.

| D ISCUSS I ON
This is the first scoping review of adolescent patients participating at level 4 or 5 according to the simplified ladder of participation. We identified only 14 articles matching the inclusion criteria. We find it interesting that we do find evidence of young co-researchers being part of all stages of the research process. This is a similar to findings reported in the review by Van Schelven et al 15 However, what topics are reported on still raises a number of questions for further discussion.

| Ethical issues
It is our opinion that the included articles would have benefited from a more in-depth portrayal of ethical challenges of involving young Collected data from own life (written, photo, video and field note). Co-facilitated focus groups.
Took part as storytellers of vignettes.
Took part in focus groups.
Read transcripts and analysed data as group.
Involved in discussion of results.
Invited to discussion but unable. Consulted on draft through email.

Participated in team.
Analysed as a group.
Done by the whole group.

Involved in coding.
Member check.
'Shared' -Not specified. et al argue 'Ethical issues to be considered when carrying out research or service evaluation with children may include: power relationships, consent issues, confidentiality and dissemination of results'. 34 They acknowledge that these issues are not unique to research with children (meaning up to 18 years of age), but argue that there may be a need to pay particular attention to them in this context. Furthermore, Gilchrist and colleagues 34 state that we live in an adult-centred society and that the power adults have can be carried over into research. This is an argument also made in a literature review by Kirk. 35 In health care, this is influenced by the authority that adolescents are used to impart to their health-care providers. They are used to assigning themselves a subordinate role as patients and of lesser knowledge and age. Whether adolescents and their health-care providers are used to seeing young patients as valid knowledge holders in the context of research will play an important role in their involvement. Are they judged by health professionals as mature enough to take part? 36 Or are the rights to participation lagging behind because health professionals are used to thinking in terms of the right to protection and that safeguarding the best interest of the child is more or less equivalent to protection? 36 The power relation between young patient and health professional is such that in most cases the adolescent would need an invitation from the adult researcher to join research. They would most likely not be the one to take the initiative. Thus, it is essential that the adult researcher see participation in research as relevant. These ethical considerations should be discussed to advance participation.
A related issue is research ethics boards' project assessment.
How can we ensure that they properly assess not just the research process the co-researchers are to be involved in, but also the role as co-researcher? Researchers wanting to involve young co-researchers should consider and have to account for ethical questions suitable for their projects. This was exemplified by Lincoln et al who developed a self-care plan for the young co-researchers' well-being. 22 Similarly, Kirby 13 highlights the importance of recognizing that taking part in research may impact on emotional well-being and that one should discuss this with the young co-researchers at an early stage. The INVOLVE guide 13 also addresses issues of safety and well-being when adolescents take part in fieldwork.

| What rungs of the ladder do we aim for?
The discussion on ethics ties into reflections we should have when utilizing the ladder of participation. It is easy to get the impression that participation at the higher rungs of the ladder should be pre- also what boundaries to establish for the safety of the adolescent co-researcher. 37 Furthermore, Hart states that 'It is not appropriate that some children feel that they must always only follow the initiative of others any more than it is good for any child to feel that they should always be a leader'. 12

| How do we prepare participants for coresearch?
A question that arises after reviewing the included articles is what is relevant or adequate research training for young co-researchers?
It has been suggested by others that fully involving children as coresearchers can be problematic due to their lack of the theoretical knowledge needed. 35 The way research training was approached, differed greatly in the reported projects. Therefore, the included articles do not provide a clear roadmap for other researchers to follow. The larger the training package, the more time and resource demanding the process will be. This is not an argument against involvement and training of the adolescents, but an argument for thinking this through and accounting for it before endeavouring on such a process. There is a fine balance between providing research training to the extent that the adolescents feel comfortable to take part and 'demystifying' research for them, while still preserving their commitment to the required time and effort. 38

| Challenges and drawback
Van Schelven et al report in their review that some of their included articles problematize that lack of research experience can lead to lack of depth in for instance interviews. 15 This is also a concern raised in the article by Van Staa et al included in this review. 24  Is it as Hart puts it liberation of the young co-researcher from adults, or is it to recognize the rights of others to participation and therefore involve them? 12

| Learning points for future research projects
Finally, what can be learned from the included articles is the need for flexibility. The need for flexibility was also pointed out by van Schelven and colleagues. 15 A young person may not be able to commit for a long period for different reasons. This brings us back to the discussion of what we aspire to achieve through the participatory process. Is it to have the same adolescents follow the whole process or is it having someone follow each step? And if adolescents are invited to full participation, are well informed, but opt for a lower level of participation such as being consulted, does that mean that the project is not truly participatory? We argue that research would benefit from a mix of forms of participation. 36 Young people themselves state that there is not one 'right' way of involvement, because preferences vary from person to person. 41 Similarly, the INVOLVE guide cautions against assuming that young people cannot be involved in certain stages of the research, but also against assuming that they wish to be involved in all the stages of the research. 13 Furthermore, there is a need to acknowledge that desire to participate might shift through time.

| Strengths and limitations
One limitation of the review could be specific databases indexing that may have prevented us from identifying relevant publications.
Van Schelven et al 15 problematize that there is a lack of conceptual clarity of PPI resulting in different definitions of the same concepts.
This could also have contributed to potential challenges in identifying all relevant studies.

| CON CLUS ION
No lists of recommendations are given in the included articles, but recommendations can be deduced from the articles. There is need the competencies to properly assess not just the research process that the adolescents are to be involved in, but their role as co-researchers.

ACK N OWLED G EM ENTS
The authors want to express their gratitude to librarian Åse Marit Hammersbøen, Akershus University Hospital, for her invaluable contribution to the database searches.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

AUTH O R CO NTR I B UTI O N S
Kjersti J. Ø. Fløtten designed and conducted the work presented in this paper, analysed the data and wrote the paper. Isabelle Aujoulat designed the work, analysed the data and has co-written the final version of the manuscript. Ana Isabelle F. Guerreiro has participated in the design, offered advice on data collection and has co-written the final manuscript. Ilaria Simonelli has participated in the design, offered advice on data collection and has co-written the final manuscript. Anne Lee Solevåg participated in the design, offered advice on data collection and has co-written the final version of the manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data sharing is not applicable to this study as no new data were created or analysed.