‘Pretty cathartic actually’: Reflections on the attempt to reduce re-traumatization of researchers and nurses taking part in a longitudinal interview study

Aim: To critically evaluate the concepts of harm and re-traumatization in the research process and to explore the ethical implications of conducting research on distressing topics using our research on the experiences of nurses working during the COVID-19 pandemic as an exemplar. Design: Longitudinal qualitative interview study. Methods: Using qualitative narrative interviews, we explored the impacts of the COVID-19 pandemic on nurses' psychological well-being in the UK. Results: To reduce the potential for


| INTRODUC TI ON
Discourses regarding the ongoing impacts of the COVID-19 pandemic on healthcare workers' well-being have been increasingly prevalent in the media and academic literature with the COVID-19 pandemic exacerbating demands on a workforce already at risk of stress and burnout (Daniels et al., 2022).The need for qualitative research to inform strategies that reduce the negative mental health impact of the pandemic amongst nurses was identified as a specific research priority by a multidisciplinary panel of mental health experts and the public (Holmes et al., 2020).To explore the impact of the COVID-19 pandemic on nurses, this longitudinal interview study was undertaken over 2 years of the pandemic (Maben et al., 2022) with data collection commencing after the first wave of  in July 2020 and completing in August 2022.
Our data highlighted the extremely distressing nature of working as a nurse or midwife during the COVID-19 pandemic in the UK (Maben et al., 2022).It can be argued that systemic failings within the UK National Health Service (NHS), such as under-resourcing and poor management were accentuated by political decisions taken before and during the COVID-19 pandemic.The suffering of patients and their families, with the effects of extreme illness or mental distress, can be viewed as intrinsic to the majority of nursing work.However, due to COVID-19, nurses encountered unusually high numbers of seriously ill patients and high death rates, which led to changes in the delivery of 'essential nursing', with care quality perceived as negatively impacted by COVID-19 pressures (Maben et al., 2022).Continuing impacts on nurses' mental health and wellbeing were ongoing too.Some reported symptoms of burnout or post-traumatic stress disorder (PTSD), including flashbacks when required to don personal protective equipment (PPE) (Maben et al., 2022).
Throughout our research, we were aware of the need to balance our desire to gain qualitative data to understand the effects of the pandemic on nurses and help identify interventions to support their psychological health, with the potential for causing further distress to individual participants who consented to take part in the research.Jaffe et al. (2015) argue that due to concerns that talking to participants about traumatic experiences will induce extreme distress, some ethical review boards hesitate to approve trauma-related research.The National Institute of Health and Care Excellence (NICE) have emphasized the need for caution regarding discussing distressing subjects with those who have suffered trauma because these individuals may be at risk of further re-traumatization if asked to recall events in detail (e.g.NICE, 2005NICE, , 2018)).Thus, NICE guidance on PTSD does not recommend single-session individual debriefing as a routine practice for the same reason (NICE, 2005(NICE, , 2018)).We argue that the risks for re-traumatizing interviewees were greatly minimized due to the unique design of our methodology, the experience of the research team and the epistemology embraced.This was confirmed in the feedback we received from our participants.
In this paper, we explore the notion of ethics and ethical practice in qualitative research and outline the approach taken in our research to reduce the potential for harm.We discuss the importance of giving participants autonomy in the research and the three tenets that became central to our narrative interviewing (respect, honesty and empathy).We employed ideas from researchers who elevate the importance of reflexivity in maintaining participant and researcher well-being during the research process (Behar, 2022) and explore participants' reactions to our research approach.

| BACKG ROU N D
Some researchers and University ethics boards have questioned whether research into distressing topics with a traumatized population can ever be viewed as ethical (Sharpe & Ziemer, 2022).Ethical conduct is normatively termed, 'being concerned with what is right or acceptable in the pursuit of a given goal' and as is discussed below, is frequently preoccupied with the avoidance of harm (Bailey & Burch, 2016).Previous research has highlighted that nurses can frequently suffer from primary and secondary traumatic stress, PTSD, moral injury and burnout with work-related stress in nurses leading to decreased physical function, emotional exhaustion, desensitization, decreased personal success, low job satisfaction and low rates of retention (Ustun, 2021;Wang et al., 2020).In the context of social research, harm is often considered to mean 're-traumatization' with this being the worst consequence that could result from a research interview (Jaffe et al., 2015;Weiss, 2023).However, the concept of retraumatization is contested.Although, not recognized as a condition in the Diagnostic and Statistical Manual of Mental Disorders (a reference guide for health professionals), those who use the concept of re-traumatization frequently do so when discussing individuals who have been diagnosed with PTSD (Weiss, 2023).Re-traumatization is frequently used to refer to the individual experiencing the same level of distress that was first encountered during the traumatic event (Follette & Duckworth, 2012).However, most scholarly literature draws a distinction between being a 'survivor' of a distressing event and talking about the distressing event, where the individual is not experiencing a threat to life (Weiss, 2023).Nevertheless, in the sphere of sensitive or distressing research with potentially traumatized populations, it may be viewed as harder to specify or predict what the research encounter may entail and therefore what levels of distress are encountered.It is possible that participating in certain research may cause the participant to re-live their distress, even if only for a short time.As Hollway and Jefferson (2012, p. 99) highlight, 'the only way to the truth' may be through causing participants renewed distress.They call for 'harm to be evaluated independently of distress' (Hollway & Jefferson, 2012, p.99).Psychoanalysis is based on theoretical assumptions which stress that 'well-being depends on making the causes of distress conscious in a containing environment, where they can be discovered not to be threatening to the survival of the self' (Hollway & Jefferson, 2012, p. 98).Aside from psychoanalysis, the value of discussing potentially distressing experiences with others, such as peers, has been highlighted by previous research (e.g.Kinman & Leggetter, 2016) which found such discussions can protect healthcare workers from the negative impact of emotional demands.An example of this is Schwartz rounds, which are confidential facilitated multidisciplinary forums where staff convene to discuss the social, emotional and ethical aspects of work experiences work (Maben et al., 2018).Recent evidence suggests they are effective in making healthcare professionals feel less stressed and less isolated at work (Maben et al., 2018).The process of supervised practice has also been identified as providing healthcare professionals, such as nurses, an opportunity to 'order their experiences and manage the projections of others beneficially' (Jones, 1999(Jones, , p. 1302)).
Most professions encourage or enforce ethical conduct through adherence to ethical guidelines to ensure that professionals act according to predetermined standards.Professional bodies, such as the General Medical Council and the Nursing Midwifery Council, tend to follow Hippocrates's instruction of 'do no harm' to the individuals involved and regulators' guidelines are regularly reviewed and updated to protect the public (Bailey & Burch, 2016).Recently, medical ethicists have endeavoured to provide a 'Principlism' framework for assessing the moral foundations of ethical codes and statements of health-related professional associations (Beauchamp & Childress, 2013).They have highlighted four core moral principles ('Autonomy', 'Beneficence', 'Non-Maleficence' and 'Justice'), and four behavioural norms to guide ethical decisions involving work with patients, clients and research participants ('Veracity', 'Privacy', 'Confidentiality' and 'Fidelity') (Beauchamp & Childress, 2013) and these have been applied to research.The 'Priniciplism' framework is widely adopted in the U.S. and Europe and many social researchers strive to adhere to this framework to limit harm experienced by participants.A recent meta-analysis on participant reactions to traumarelated research found that participants generally do not experience re-traumatization from participation or regret participating in the research, regardless of the type of traumatic events they had experienced (Jaffe et al., 2015).Jaffe et al. (2015) emphasized that for participants as a whole, the potential benefits of participating in trauma research outweigh the modest amount of distress experienced.Jaffe et al. (2015: 52) went on to argue that the safeguards embedded in 'properly designed studies' give the participants autonomy and 'counteract any sense of helplessness or lack of control that are hallmarks of actual trauma events'.Indeed, the principle of autonomy has recently received attention in debates in social research ethics (e.g.Traianou & Hammersley, 2021) with the most ethical research committed to providing highly detailed information regarding all aspects of the research and the use of data to facilitate full informed consent being obtained.
In our study, the data generated by the 'parent' quantitative study, which our qualitative sample was derived from, showed that nurses' work during COVID-19 was characterized by a lack of autonomy (Couper et al., 2022).It was therefore important for the research team to ensure that autonomy for our research participants was intrinsically embedded into the framework of our study.We did this through varying methods, such as adopting narrative interviews which gave the participants the opportunity to tell their stories at their pace and in their own words, through to a participatory codesign event day which was held after the interviews.As others have noted (e.g.Conolly, 2008;Dempsey et al., 2016;Dickinson-Swift et al., 2008), research about sensitive issues has a risk of harm but this can be mitigated when the research is approached and undertaken with the utmost sensitivity.

| OVERVIE W OF THE ISSUE S
The research team were aware of the potential societal and historical value of understanding the experiences of nurses working during COVID-19, as an aid to develop support for nurses and feedback to employers, whilst being mindful of the potential risk to the individual.However, the research team faced a dilemma regarding which strategies should be used in the interview process and subsequent analysis to minimize this risk.Feminist methods were felt to represent the best-informed approach due to their empathetic, reciprocal style and effectiveness at reducing the power differential between the researcher and participants.Such methods have traditionally been used by female researchers, with female participants, to reduce the power differentials inherent in research (DeVault & Gross, 2002) and therefore to potentially make research interviews more ethical (Conolly, 2008).As Foucault (1988) argued, power is most appropriately thought of as relational, so something that is exercised from a variety of points in the social body, rather than something that is acquired, seized or shared.To counterbalance these power relations, decades of feminist research have recognized the value in the sharing of experiences (e.g.Oakley, 1981).A relational form of power, as Foucault (1988) suggests, encourages an exchange of knowledge which in this scenario seemed to enable a more equitable research relationship.
For these reasons, our approach was informed by feminist research tenets.Influenced by the resurgence of feminist interest in the ethics of care (McLeod, 2017), where notions of morality take centre stage, our approach considered Walker's, (2007: 16) argument that it is 'fruitful to locate morality in practices of responsibility that implement commonly shared understandings about who gets to do what to whom and who is supposed to do what for whom'.Other feminists have argued for responsibility to be viewed as an interaction (Barad, 2007), or for a social connection model of responsibility, which advocates practices and collectivities, focusing on responsibility for actions to address injustice (Young, 2011).We argue that modern feminist ethics involving the positioning of responsibility as a practice and interaction directly speaks to nursing, because of the gendered nature of the profession (Davies, 1995) and the difficult working practices that nurses face in the UK.In our research, only two participants were male and the majority of the research interviewers, (all bar one) were women.Although gender inequalities were not a focus of this study, it could be argued that the subjugated position of many nurses in the NHS demands their experiences are studied with a feminist lens.However, our epistemologies and ontologies were greatly influenced by the context of the research interviews and the context of the COVID-19 pandemic.We Therefore, we decided to elevate and embed participant autonomy and researcher reflexivity into our research framework.As a means to achieving this, we decided to follow Hollway and Jefferson's (2012) advocation of the importance of three main principles when conducting qualitative research, those of honesty, sympathy and respect.Although these tenets may be the keystones of all good qualitative research, as we go on to argue, we found that respect, honesty and empathy towards our participants were of central importance in our research process to minimize the potential for distress and trauma, reduce the power deferential between researcher and participants and thereby make the research process as ethical as possible.We adopted empathy over Hollway and Jefferson's (2012) use of the term sympathy as we believe the term empathy serves to highlight better the emphasis we placed on our alignment to, and understanding of, the participants' experiences and perspectives.

| DATA S O U RCE S
The aim of the study was to explore the range of experiences of nurses working during the COVID-19 pandemic in the UK and the possible impacts on their psychosocial and emotional well-being.
A social constructionist approach was utilized which viewed realities as being constructed in a relational context and with multiple insights (Burr, 2015).This paper draws on our study findings and aims to reflect upon our negotiation of the ethical challenges encountered from inception to the completion of this study.Participants were recruited via an opt-in method with individuals who had completed the parent study national nurse and midwife longitudinal surveys and expressed an interest in being contacted to take part in qualitative interviews about their COVID-19 experiences (Couper et al., 2022).The open comments from this survey, as well as social media and press accounts of nurses' experiences, informed our interview topic development.The nurses we spoke to were based in varying geographical locations throughout the UK and were sampled from varying ethnic groups, age ranges, were from a range of settings, differing experiences and levels of seniority.Out of the 50 participants, two were men.Please see Table 1 for the participant characteristics.
We became aware of the necessity to respect the autonomous manner in which the participants wished to tell their stories in flexible, open and non-directive interviews, enabling participants to lead.We realized that their narratives were not confined to their COVID-19 experiences, but also reflected on biographical information including their personal lives, identity as nurses, and what being a nurse meant to them pre-COVID-19 and during COVID-19.We embraced this approach to interviewing and began to align ourselves to the concept of 'Gestalt', or the idea that 'the whole is greater than the sum of its parts' (Hollway & Jefferson, 2012, p. 68).Thereby, it became a priority to locate interviewees working experiences within the context of their lives.This narrative interviewing approach (Hollway & Jefferson, 2012;Plummer, 1995) facilitates the production of interviewees' meaning frames.Stories were elicited to anchor the nurses' accounts to events that actually happened.Open-ended questions were used and we followed up during the interview using the participants' ordering and phrasing (Hollway & Jefferson, 2012).
We gave the participants time to relate their stories in their own way.
This meant that the interviews were long, often lasting in excess of 90 min, and one participant even asked for a repeat 'first wave' interview to provide greater clarity, which we accommodated.The benefits of adopting such a fluid approach to interviewing have been emphasized by researchers working in sensitive research topic areas and those with mental health problems (Parr, 1998).Our flexibility in being able to adopt this respectful, honest and empathetic approach, along with the team members' insight due to their professional and research experience (all six interviewers were experienced qualitative researchers, four of whom had extensive clinical nursing experience) enabled the interviewees to lead the interview process, affording them the opportunity to relate their stories at length.
The pandemic necessitated remote interviews although these were not without complications.The interviews were mostly conducted on video conferencing software, such as Zoom.Interruptions were frequent due to unstable Wi-Fi connections, homelife (such as children, pets and doorbells ringing) and other household noise.
Interruptions would normally be viewed as problematic when conducting sensitive interviews (Britten, 1995).However, the pandemic facilitated a mutual experience of conducting interviews from our own homes and our honest approach helped to bring participants into our worlds and contextualize our own home lives, therefore normalizing interruptions.The location of the interviews, being within both the researchers' and participants' homes, as well as providing an honest backdrop for discussions, may have also provided an element of empathy and safety when discussing the sensitive and distressing interview material (Parr, 1998).DeVault and Gross (2002, p. 206) highlight the flexibility of language and its productive powers, with 'subtle shades of meaning' conveyed through nuances of speech, gesture and expression.During normative qualitative interviewing situations, key indicators such as body language and eye contact would be easy for the interviewer to note.However, these staples of qualitative research are much harder to follow when the interview is conducted remotely via telephone or video link, when often only head and shoulders are visible.Therefore, the interviewers' empathetic, active listening skills became even more paramount.
The autonomous, voluntary nature of participation in the study was emphasized, with participants free to tell their stories how they wanted to and also free to withdraw at any time.The study gained ethical approval from the lead author's University ethics committee.& Jefferson, 2012, p. 98).An emphasis was placed on enabling the participants to finish what they wanted to say so nothing was 'left hanging'.It is possible that inexperienced interviewers would not be able to 'contain' the distress in such a comfortable manner, instead feeling anxious or threatened by it thus leaving the interviewee to hold the emotions and causes of distress in the interview encounter (Hollway & Jefferson, 2012).
A sensitive analysis of the data was undertaken with interviews inductively analysed for themes with a subsequent narrative analysis conducted to preserve the form of each participant's narrative.NVivo 12 was used to organize data and develop inductive codes and themes across the datasets.Pen portraits, or interview summaries, were also produced which helped to avoid fragmentation of the data (Hollway & Jefferson, 2012).We strove to identify segments of text that took the form of narrative (Riessman, 2002).
Using Muller (1999) five overlapping stages of narrative analysis, we identified whole stories, instead of segments of text, to guide the development of the pen portraits.The production of secondary level themes, which were used with the pen portraits, aided our longitudinal holistic approach to analyse the interviews from each participant (Maben et al., 2022).After comparing each participant's interview data, the lead author then compared with data from other participants at the same time point and then across all time points with co-corroboration occurring with two authors.Research participants were consulted about our findings and prior to publication of articles we checked if they were happy for their data to be used.The findings discussed in this paper are characterized by Figure 1 below.We will first discuss the researchers' perspectives before turning to the participants' perspectives.

| FINDING S: RE S E ARCHER PER S PEC TIVE S
The researcher-participant relationship is an intersubjective rela- is combined with the 'solidity of walls' to emphasize how the problems, thoughts, and identities of others may move across the sheer wall to become part of the foregrounded identity of the researcher (Carroll, 2012, p. 557).We drew on this work and found that the researcher's labour undertaken when conducting honest, respectful and empathetic research experienced by the research team meant that, as a team, we also had to confront, understand and integrate our own emotions (Carroll, 2012) into data collection and analysis.
In our research, we found that the dictum: 'if we undertake to study human lives, we have to be ready to face human feelings' (Ely et al., 1991, p. 49) held true.Previous work has highlighted the 'emotion work' undertaken by researchers (Dickinson-Swift et al., 2008).
The potentially distressing nature of the research and accompanying anticipated emotional management, did raise certain, differing issues for each member of the research team.Prior to the commencement of interviews one author felt able to confide with honesty that she did not wish to interview participants working in maternity care because, as a pregnant woman, she felt that she needed to protect her own well-being prior to giving birth.This wish was respected and treated empathetically.Q ualitative data encourage researchers to become immersed in the data or to 'become a knower' (McLeod, 2001, p. 165).But for those who research distressing topics, the normal requirement of being immersed in the data can prove problematic.For the researcher, distressing material may be heard during the interview, when checking the transcript after transcription, and several times when reviewing the material and coding or writing summaries of the interview in the analytical phase.The cumulative impact of hearing these distressing stories over and over can take researchers unawares.
All of the research team found the interviews to be extremely impactful, if not actually distressing, and similar to Carroll's (2012) gossamer wall metaphor, we found we did take on some of our interviewees' distress.Some members of the research team felt that these interviews were some of the most emotionally challenging in their career, knowing that certain aspects of nurses' stories would remain with them forever.Previous literature on researchers' emotional involvement in the research process has highlighted the emotional exhaustion and vicarious traumatization that researchers themselves can experience (Dickinson-Swift et al., 2008;Silverio et al., 2022).The empathy the team members displayed to each other enabled us to cope with some of the distress that we, as interviewers, inevitably took on during the interviews.However, the coding and writing summaries of interviews is possibly a more solitary process than interviewing.There were occasions when the cumulative effect of the distressing content of the interviews caught the lead author unawares, particularly when coding and writing the summaries of the interviews.
During another winter COVID-19 lockdown, after reading one participant's transcript for a second, or third, time tears began to flow freely.
Other research has highlighted the risk of burnout for researchers who are emotionally affected by their work and the incorporation of selfcare into research projects has been advocated as a method to mitigate such emotional impacts (Dickinson-Swift et al., 2008).We would highly recommend mindful inspired well-being breaks in order to 'connect' with ones' present surroundings to limit the immersive qualitative process to allocated times.Research diaries have been advocated as an essential tool in managing the emotional distress encountered when undertaking interviewing and the analysis of data (Silverio et al., 2022).
One author, who initially did not undertake interviews, found reading participants narratives similarly impactful and below is an excerpt from her research diary: Having not conducted an interview, receiving the raw data transcripts was the first time I was exposed to the data.The data were evocative and highly descriptive, allowing the narratives of the participants and their experiences and feelings to leap off the page.
I could visualise myself there standing beside each participant on the wards, in the care homes, trying to function in society, juggling the balance between risk and personal safety for themselves and their family members.I could feel their stresses, guilt and anxieties, understanding their struggles and decision making as they engaged with patients and colleagues.The more I engaged with the data, the more mesmerising the data became, more and more narratives became etched in my memory.They needed to be pulled apart, dissected and scrutinised through coding and the meticulous application of theory, the responsibility that the analysis did the participants narratives justice weighed heavy on my mind.
The research team was impacted differently but shared in the goal of being supportive and thorough in terms of conducting a robust study.
As the author above highlighted in her research diary, it is through the analytical process that sense is made from data.Although the process of immersing oneself in data, coding, and the application of theory may be difficult with distressing data, this is how researchers can make sense of data, process it and, hopefully, feed into academic and policy debates to bring about lasting change to try and mitigate similar future distress.
As a faceless 'resource', transcribers are often a forgotten member of the research team.Often assumed to be a somewhat mechanistic or objective job, the psychological and emotional impact of listening to and then accurately transcribing traumatic interviews is often overlooked.Some authors have argued that it is vital that transcribers' vulnerabilities and needs should also be considered (e.g.McCosker et al., 2001).Frequently women, transcribers hear the research participants' stories in the same way that the interviewer does, but due to their somewhat powerless status as a hearer, they can be viewed to participate passively in hearing participants' stories.It can be argued that this could possibly be a more traumatic role than actually interviewing research participants because transcribers are unable to use their autonomy to effect any change in the interaction.Therefore, we argue that it is key to consider transcribers' needs.Previous authors have highlighted the need for ethical and institutional review committees to act to prevent their emotional injury during the course of sensitive research.A range of further recommendations for the protection of transcribers have been made which include: informing them regarding the nature of the research and the type of data they will be transcribing; ensuring they are informed prior to receiving particularly 'challenging' or 'difficult' interviews; giving adequate time and space for regular debriefing sessions with members of the project team; providing signposting to an appropriate person for crisis counselling; and raising the possibility of using a journal to log and reflexively reflect upon their thoughts and feelings (McCosker et al., 2001).This log or journal then may be used as part of fieldwork notes if appropriate, in a similar way to researchers' fieldnotes.In the current study the authors took the steps outlined above.

| FIND ING S: PARTI CIPANT PER S PEC TIVE S
Similar to Parr (1998), we were mindful that the trauma our participants had experienced may have made them more likely to position us, researchers, as therapists.This is a position we were uncomfortable with as we were neither trained therapists nor were we consistently available for therapy sessions (Parr, 1998).We were clear about these restrictions although we provided a list of wellbeing resources and emphasized to our participants that we were available for email exchanges or updates and that we could identify further supportive resources if necessary.Participants were keen to contribute their experiences and to have their experiences and stories heard and, despite the interviews not being therapy sessions, many participants experienced them as such.The respect, honesty and empathy that we afforded our participants were met positively and there was minimal loss of research participants to withdrawals between their interviews.We remained The nurses spoke about their involvement in the research process in overwhelmingly positive terms.Similar to Camila above, many of our participants likened their research experience to that of a psychoanalytic encounter, therapy, or even the intense professional supervisions they had received when they were beginning their nursing careers: It allowed me that opportunity to just take a pause and For Isabella and Jo, the opportunity to reflect upon their feelings presented an opportunity to move on in a therapeutic manner.Their experiences and feelings had been dealt with and could be therefore put behind them.Sandra expressed a similar sentiment as she related an appointment that she had made to speak to a counsellor the day after her first research interview.However, when she got there she (Camila) Many of the nurses who emphasized the cathartic nature of the research encounter had experienced emotional distress during their interviews.We were very aware of their psychological wellbeing and the source of support we provided in the 'safe space' of interviews is very different to a sustained, constant therapeutic input from one identified therapist (Parr, 1998).We frequently reminded our participants of the sources of support that they could access after interview.The immense distress the nurses frequently re-countered, its causes and impacts, are reflected upon more fully in (Conolly et al., 2022;Maben et al., 2022).For many of the nurses who participated, the distress they encountered was not behind them.The extreme conditions made the characterization of their involvement in the interview process as overwhelmingly positive even more astounding.However, two of our participants did characterize their involvement overtly in terms of  (2008), rather than consisting of an objective process, qualitative data are produced and collected through inter-subjectivity between researcher, respondent and other significant individuals and institutions.Power relations can never be overcome but there is value in consistently analysing and rendering them visible through reflexive discussion.We were mindful of researchers' positionalities, their relationships with participants, and the research process (Behar, 2022).
We attempted to remain aware of differences that may be invisible and how these shape the research process.As such, we appreciated the value in being mindfully aware of the intersubjective identifications between researcher and participant to conduct the most ethical qualitative research (Harvey, 2017).
Empathy was key throughout the project, and the application of this tenet throughout the research helped to minimize the risks of the 'gossamer wall effect' (Carroll, 2012) to the research team and minimize the risk of (re)traumatizing the research participants.
Empathy was exhibited within team debriefs, with the valuing and sharing of experience and insider knowledge, the whole team was encouraged to treat the participants', and their own feelings, experiences and emotions with the upmost respect, honesty and empathy.Therefore, our team culture, alongside our overall ethos and analytical approach, fostered an appropriate context for the sensitive treatment of data, some of which was extremely distressing.
We argue that our approach to interviewing enabled the nurses with whom we spoke, to feel heard and to trust that we would treat their narratives respectfully.We emphasized our ability to maintain confidentiality promising to discuss any data entering the public domain which could be identified.As such, participants felt able to talk honestly about their COVID-19 experiences and the trauma they had experienced, without it reportedly being retraumatizing.By enabling our participants to relate their story at their own pace and to determine where and when to start the telling of it in a holistic manner, we were 'keeping the whole in mind' (Hollway & Jefferson, 2012).Whilst comparing the nurses' narratives, we were able to offer a holistic interpretation of the impacts of COVID-19 on their well-being.We used our combined nursing, sociological and psychoanalytical knowledge to make multiple links within and between cases.We would advocate that research into distressing subjects requires the most sensitive and ethically sound treatment and that our approach has been successful in achieving this.Although we were clear that we were not engaging in a therapeutic relationship, the majority of our participants revealed that they found the research encounter therapeutic in nature.The enthusiasm with which participants shared their stories and their expressions of gratitude at having their stories heard have provided evidence that our approach was both suitable and welcomed by our participants.After receiving further funding, the study is ongoing and the research team will endeavour to apply these values going forward.

13652648, 0 ,
Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jan.15764by Cardiff University, Wiley Online Library on [11/07/2023].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License argue that the principles of ethical research on distressing subjects can be considered universal and do not have to be limited to the label of 'feminist' research.We noted the increased emphasis on reflexivity placed by many researchers, such as Behar (2022), who was able to interweave ethnography and memoir in her work whilst encouraging researchers to be as open as research participants are with researchers about their experiences.Through documenting the shared vulnerability of the observed and the observer, Behar (2022) aimed to provide greater depth of understanding and feeling to her participants' lived experiences.Behar's (2022) boundary blending work led to the creation of hybrid genres such as self-ethnography and ethno-biography.
The research presented here involved narrative interviews with 50 nurses.Sample 1 (n = 27) took part in four interviews, over 20 months, the first in July 2020 (n = 27).Twenty-five were interviewed in December 2020, 26 in August 2021 and 21 were interviewed for the final time in March 2022.We extended our participant population with a further sample to include larger numbers of non-white, student nurses, care home nurses and community nurses.Thus, interviews with Sample 2 (n = 23) began after the second wave of COVID-19 in the UK in August 2021 and 19 of the nurses in this sample were reinterviewed in March 2022.
Data collected were ensured confidential status with identifying information removed.Every member of the team participated in interviewing.Of the predominantly female team, four are Professors of Nursing, and as such were able to draw upon their vast clinical experience, including communication skills to ensure participants were interviewed sensitively.The other researchers who undertook interviews are experienced qualitative researchers and also have considerable experience in conducting interviews on sensitive and distressing topics and approached interviewees with respect and empathy, offering opportunities to pause or stop the interview if needed.The team members' experience gave them confidence to not shy away from talking about distressing subjects, allowing the interviewees to say as much or as little as they wanted in the space of the interview and provided a 'containing environment'(Hollway tionship with each individual affecting 'the other on multiple levels throughout their interactions'(Harvey, 2017).In psychoanalysisFreud (1910)  conceptualized transference as a largely unconscious 13652648, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jan.15764by Cardiff University, Wiley Online Library on [11/07/2023].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License re-experiencing of the patient's early emotional life, directed towards persons from his or her past, and in relation to the psychoanalyst (Harvey, 2017).Psychoanalysts have argued that the process can occur in both directions and therefore the transferencecountertransference situation (Ogden, 1994) is interwoven and inseparable and frequently characterized as an intersubjective therapeutic relationship.Previous qualitative researchers have noted how the data generated in a research interview may be influenced by transference-countertransference, for example, when an interviewer / participant dynamic is reminiscent of a mother / daughter relationship for both parties (Hollway & Jefferson, 2012).Intersubjectivity refers to the reciprocal mutual influences to describe the continual, changing intersubjective nature of intrapsychic experience and it is in this way that meaning is co-created (Harvey, 2017).Whilst conducting our research, we realized the necessity of being attuned to intersubjectivity, and therefore being aware of the co-creation of the meaning process through ongoing honest, introspective engagement and reflection, or reflexivity, was of key importance.As outlined above, as a research team we were drawn to researcher reflexivity.All the members of the research team were deeply affected by our interactions with the participants of the research.Carroll (2012) appliedDoucet's (2008) gossamer wall metaphor to the research context to emphasize the relational aspects of the reflexive process during research.The 'sheerness of gossamer'

F
For the most part, working as part of a team where every member was involved in interviewing, provided an empathetic support network.Early on in the interview process (wave 1, July 2020), the team met to debrief and discuss the emotions generated by the interviewing process.The second author debriefed the third after an interview she found very distressing and she made herself available to others, thus demonstrating empathy towards one another.Similar toSilverio et al. (2022) the authors wrote reflective accounts of our interviewing experiences: The experience of interviewing nurses caring for patients during the first wave of the pandemic was deeply humbling and very emotional.I was in awe of their skills and fortitude and I was full of empathy and sorrow for all they were enduring.Several interviewees were tearful, others told the most traumatic stories in a sparse and factual way which made them all the more powerful.Conducting the interviews on zoom required particular skills.I drew on all of my nursing clinical communication skills training and experience, using active listening, strong eye contact and small gestures to convey empathy and support and at times showing my own raw emotion in response to what I was hearing.I was aware at times of my own spontaneous gestures -using my hands to tap my chest near my heart and a deep sigh to convey sincere empathy and to convey my own emotions to participants.I needed time after many of the interviews to cry, rage and recompose myself after what I had heard.On occasion I debriefed with the Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jan.15764by Cardiff University, Wiley Online Library on [11/07/2023].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License immediate team, other nurse researchers and psychologists to make sense of what I had heard and share and make sense of the horror of the situation nurses were experiencing.These research interviews were some of the most sensitive and challenging of my long career.

13652648, 0 ,
Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jan.15764by Cardiff University, Wiley Online Library on [11/07/2023].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License in email contact with our participants after their first interview and we collated the emails we received from participants during their involvement.In their interviews, we also asked participants to reflect upon their involvement in the research.The empathy we strived to display in the research process was frequently reflected upon in the emails from participants about their involvement: You have a lovely interview technique and your warmth and excellent active listening skills made me feel very comfortable and heard.(Gaby) I thoroughly enjoyed talking with you.It was like therapy.I'm proud to be able to help with your research.It was lovely chatting to you afterwards also.(Camila) The utility to the nurses themselves, either professionally or emotionally, of being involved in the research project was frequently mentioned.When informed that findings from the research project had been submitted as evidence to the Parliamentary select committee on nurses' well-being during COVID-19 one participant stated: It is very positive and thank you for letting me know -it makes such a difference knowing that taking part has made a difference.(Louise) said everything already, I'm actually okay now.Because I'd said everything to you already.So I found it therapeutic.(Sandra)As can be seen above, many of the participants reflected on their participation in the research process as valuable and in some ways therapeutic for them.The participants framed their participation as affording the opportunity for reflection on their well-being, organize their thoughts and feelings about what had happened, to tell their stories and to move on.However, Camila highlighted the very fact that people were conducting research into COVID-19 experiences in nurses as the primary positive aspect: Pretty cathartic actually (…) it just sort of helped with my resilience I think, it just helped to think that, you know, there's people out there really looking into this (…) I just found it, for me personally, very, very helpful.
distress: Rachel and Sarah: I do remember getting off the phone and I was in tears for quite a while afterwards.So I think it brought back (…) so I did 'um and ah' a little bit today as to whether or not I wanted to take a part again.But I decided that we're now in a very different place and I should take part (…) So I think it has been a positive experience, to kind of think back and reflect.(Rachel) 13652648, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jan.15764by Cardiff University, Wiley Online Library on [11/07/2023].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons LicenseI found it upsetting, it doesn't mean that everyone else would have found it upsetting, and even if you, like, showed me the questions that you were going to ask me, I might have seen the questions and thought, well, I'm not going to bring that up, so it will be fine.Whereas, like, when you're like talking about everything, like I made that decision to talk about that experience because I wanted there to be an accurate snapshot of what I'd been through.(Sarah)In the extracts above, we can see that nurses' motivation for being interviewed was to have their experiences acknowledged and valued.The key feature of Rachel's and Sarah's narratives is that even though they characterized their research participation experiences primarily in terms of distress, both of them went on to frame their distress in terms of a positive outcome.For Rachel, it was positive to 'think back and reflect' whilst Sarah wanted an accurate testimony of what occurred.For Sarah, the distress she experienced during her first interview convinced her that there was 'something wrong' and it provided her with motivation to seek professional counselling.They also both emphasized the control they felt in the process and the voluntary nature of their participation.For example, Sarah highlighted the active 'decision' she made to talk about the more distressing elements of her experience so that the research could accurately reflect the reality of her experiences.As other researchers have noted (e.g.Parr, 1998,   p. 348) the methodological lesson is not always to avoid distress or emotion, but 'to avoid the placing of participants in inferior and powerless positions'.Our work accords withJaffe et al. (2015) meta-analysis of trauma-related research which found that interviewing participants about prior trauma, whilst possibly leading to some immediate, low-to-moderate distress, represents a minimal risk to adult participants including those who have been diagnosed with PTSD.7 | D ISCUSS I ON AND CON CLUS I ON SIn this paper, we have argued that research into distressing subjects should not be avoided simply due to its potentially distressing nature.Indeed, such a position could be detrimental to any attempts to detail and record nurses' emotional experiences of the COVID-19 pandemic and indeed, elucidate social injustice in all spheres.Instead, by holding feminist theories of care and relational responsibility and ethical research practice at the fore, which for us involved embedding participant autonomy and researcher reflexivity into the research design and process through the consideration of the three central tenets of respect, honesty and empathy, researchers can formulate appropriate epistemological and ontological approaches to qualitative research when dealing with even the most distressing data.As demonstrated in this paper, being mindful of these tenets is invaluable to minimize the risks associated with conducting distressing research.Emphasizing respect, honesty and empathy in our research practice has enabled us to be mindful of power dynamics and knowledge transfer, whilst being mindful of reflexivity and intersubjectivity, and therefore a more balanced research relationship was produced.As others have highlighted (e.g.Traianou & Hammersley, 2021) participant autonomy is sometimes viewed as key in the pursuit of ethical research with the participant given the ability to choose when and where to take part in the research, how to tell their story and what happens to their data.As we have previously argued Conolly