Ethics of witnessing in clinical case notes of children in state care: From data extraction to narrative care

Abstract

Clinical case notes provide researchers with a rare yet unsettling insight into the intimate realities of children’s mental health. These records capture fragments of lived experience that demand rigorous ethical attention in how they are accessed, interpreted, and re-presented. This paper examines the ethical tensions between data extraction and narrative witnessing through the reflections of a Social Work practitioner, through a collaboration wth professionals, and Experts by Experience. Drawing on a year-long reflexive study of clinical case notes from children in state care, this paper explores how professional language, institutional framing, and systemic pressures influence both the production and interpretation of such records. Divergent professional positions and discourses often collide, shaping children’s representation and obscuring their voices within clinical and administrative documentation. The analysis also considers the emotional labour and moral unease experienced by researchers when encountering recorded suffering, including the risks of vicarious trauma and moral distress. By reframing clinical case notes as testimonies rather than data alone, the paper argues for the practice of narrative care, an approach that prioritises attentive anonymisation, reflexive interpretation, and dignified representation. Narrative care acknowledges the methodological and emotional dimensions of engaging with trauma-saturated texts, enabling ethically responsive research that bears witness to children’s experiences. In doing so, this study highlights how researchers can uphold humanity within the constraints of research governance, transforming analysis into an act of ethical witnessing that recognises, rather than erases, the suffering and resilience contained within clinical records.

Introduction

The number of children in state care in England has now exceeded 81,000 (Department for Education, 2025), a trend that has led to growing concerns regarding their mental health (ADCS, 2022; Evans et al., 2021; WWfCSC, 2022). These children are particularly vulnerable, with most having been exposed to neglectful parenting or abuse and disrupted care experiences (Department for Education, 2025; Hiller et al., 2020; Phillips et al., 2024). To understand the kinds of intervention these children receive from services, a year-long study exploring care trajectories using clinical case notes held within a large NHS trust in the South of England was conducted. This data documents how children’s need for a service is recognised, responded to, and often mediated through multiple professional lenses.

Findings from the analysis of this data supported the view that professionals’ responses and interventions are shaped by institutional logics and narrative silences, rather than reflecting the truths of children’s circumstances (O’Keefe et al., 2025). This was seen, at times, to influence the trajectory of how other professionals understood children’s needs, making interpretation of case notes production problematic or seen as fact, without careful consideration of how understanding is constructed (Lillis et al., 2017). Clinical case notes have been shown to shape decision-making about future service delivery, including interventions offered (Morgan et al., 2025). Additionally, recording is typically completed by adults and therefore centres on professional perspectives, which may marginalise the voices of children and overlook their lived experiences, with little room for children to input their own voices into decision-making (Hawkes et al., 2024; O’Keefe, 2024). As such, they are often viewed as places of professional analysis or repetitions of action, or lack thereof (Lillis, 2020). Although the understanding of case notes also holds potential to predict needs (Yin et al., 2023) based on behaviour and reported presentations and also allows for researchers to evaluate the effectiveness of interventions (Trubey et al., 2024), providing rich insights into decision making and outcomes for children (O’Keefe et al., 2025). While children’s own accounts are less common, clinical case notes provide windows and snapshots of children’s lives, with occasional records capturing children’s own direct accounts in letters, quotations, texts, or emails (Lillis et al., 2017).

Inevitably, clinical case notes also detail risks that escalate and capture moments when harm occurs. Engaging with these documents as a researcher entails an exposure to others’ trauma, and so a risk of absorbing the harm of others, contributing to vicarious trauma, desensitisation and a level of ethical burden (Lillis, 2020). Little is known about the impact of vicarious trauma on researchers, including its effects on their welfare and its impact on their work (AbiNader et al., 2023; Dickson-Swift et al., 2006). Research is starting to indicate that engaging in emotionally demanding research with underserved communities can lead to a spectrum of adverse effects, including physical health issues such as headaches and nausea, as well as significant emotional and psychological challenges such as anxiety, depression, fear and anger (Dickson-Swift et al., 2006; Mallon and Elliott, 2019; Smith et al., 2023; Taylor et al., 2016). Researchers also face potential costs associated with co-production, including interprofessional conflict and the management of others’ emotions (Oliver et al., 2019). Existing guidelines on how best to manage emotionally demanding research have been principally oriented towards academic researchers (Burrell et al., 2023).

The exposure to indirect traumatic incidents is known to be harmful both in the short and potentially longer term (AbiNader et al., 2023) requiring measures to support researchers’ process within their work. Researchers must remain alert to how their positionality shapes interpretation and the importance of human connection (Braun and Clarke, 2024), recognising that we do not enter the record as neutral observers but as practitioners and scholars with our own unique past experiences. Reading about children whose trajectories might have been different, where unmet needs were evident yet unaddressed, can bring feelings of frustration and helplessness, but also joy and fulfilment when a practitioner engages and supports a child (AbiNader et al., 2023). It can also desensitise and shape findings while overlooking and creating unintended distance.

For practitioners who record in clinical and social care settings, high-quality case recording is essential for ensuring continuity of care, particularly in complex systems where multiple professionals may be involved over time. Accurate and detailed records allow practitioners to understand a child’s history, needs, and progress, enabling coordinated and responsive interventions (SCIE, 2025). Practitioners’ confidence and consistency in recording vary, and while many strive for child-centred approaches, systemic pressures often limit their ability to do so (O’Keefe et al., 2025). Case recording is frequently treated as a ‘nesting activity’, something done in between other tasks, resulting in limited time and attention (Lillis et al., 2020). This can lead to significant encounters or disclosures of serious harm being reduced to brief factual statements, with little analytical depth or contextual detail. Such recording practices can obscure the emotional and relational dimensions of trauma, making it difficult for researchers and future practitioners to fully understand the child’s lived experience or the details of how the encounter was felt by the child and the practitioner (AbiNader et al., 2024; Richardson and Zolkos; Stumm, 2022).

In such moments, practitioners are confronted with the challenge of bearing witness to harm (Bosack et al., 2025; Causer, 2021) while trying to understand our own role in the experience, without the possibility of intervening. Witnessing, therefore, is both a burden and a form of ethical responsibility (Richardson and Zolkos; Stumm, 2022). To witness is to resist the erasure of suffering by naming it, by making it visible in spaces where it might otherwise be hidden. While this cannot change what happened, acknowledging the learning and potential from records alongside situating experiences within broader patterns of systemic challenges offers potential to transform the act of reading into a form of ethical recognition. These tensions form the focus of this paper: reflecting on what it means to analyse case notes not only as data but as testimonies, and how researchers can adopt practices of narrative care. Narrative care here is understood as any form of care (formal or informal, professional or personal) that is grounded in and guided by a sensitivity to the stories of people’s lives: the stories they tell, the stories they live by, and the stories they are caught up in (Randall and McKim, 2008). This analysis aims to uphold the dignity of those whose lives are inscribed in the record, recognising their experiences and attending carefully to how they are represented.

Methods

This reflective piece draws on a study which included a reflexive thematic analysis of Child and Adolescent Mental Health Service (CAMHS) clinical case notes from one NHS Foundation Trust led by myself, a Qualified and registered social worker. In October 2023, a sample of 20 individual mental health records was extracted, with each record relating to a single child living in state care. Conditional random sampling was used to ensure diversity in age, gender, placement situation, and ethnicity among children who had received CAMHS interventions within the previous 2 years.

The records include data from one of the largest mental health providers in Europe, as well as from national and specialist outpatient services. This area has greater ethnic diversity and a higher proportion of deprived areas than national data (Wickersham et al., 2024). The system provides secure and anonymised access to case records, assigning children ZZZZZ to uphold anonymity and carers and place settings are often represented as QQQQQ. The NHS trust operates an opt-out option and, consequently, separate consent is not requested for individual studies. Records consisted of information from other agencies, reports, risk assessments and case notes. Case records included in this study were also assigned pseudonyms following extraction, and quotations alongside this paper were reviewed by the CRIS oversight committee, given the low sample size, and to further protect confidentiality.

Children’s records ranged in length from 59 to over 600 notes, with notes varying from brief observations to detailed multi-agency documentation. The data were analysed using reflexive thematic interpretation (Braun and Clarke, 2023), which I undertook in collaboration with a diverse professional and academic research team comprising social workers, psychologists, academics, students, and Experts by Experience (EbyE). This diversity of perspectives was integral to the analytical process, supporting critical reflection and helping to form a more ethically attuned interpretation of the material, including the challenges of how emotions shaped reactions to encounters in case notes (Wright et al., 2021). The findings presented here reflect on selected records and explore the tensions, emotional labour, and challenges encountered throughout this research process from different positions. As part of this process, I kept a daily reflective journal, noting feelings, emotions, and reflections at the time of analysis and after analysis to better understand the influence of reading and analysing data throughout the study.

Reflections on findings

Incompleteness as a witness - Joe

Across years of referrals, assessments, and planning meetings, Joe’s record captures professionals’ decisions, but rarely his own or his family’s words or feelings. His ethnicity is misrecorded, and when his carer speaks of her hopes, the notes appear to describe his deficits and struggles, rather than seeing his needs. As an extended family foster carer, his carer appears to be seen as a family member first and a professional and expert second. CAMHS referrals are declined because long-term legal planning is perceived as required to provide stability, the carer feels unheard, and the record documents their upset with the professional network. Meanwhile, Joe’s needs remain suspended as professionals navigate legal and institutional gateways. Reading Joe’s record as a researcher, it is striking less by what is present than by what is missing. His life is processed, assessed, and administered through an increasing number of measures. Reading such experiences inevitably provided only windows into children’s lives, and at times involved sitting with uncertainty and gaps in children’s experiences. As a practitioner and a researcher, I recognise that capturing the child’s voice is now considered an integral aspect of child-centred case recording (O’Keefe, 2025). Although without seeing a child, a void in knowing left a feeling of discomfort, suppressed through professional filters (McMahon, 2023). Such feelings were shared within the research group, with one EbyE recounting their own experience and comparing this to searching for children’s voices, frustrated that patterns from their own childhood were still apparent.

As with other children’s records, case entries for many children rapidly increased following their accommodation into state care, alongside, for most children, a simultaneous reduction in direct mental health interventions. It was acknowledged, at the onset, that Local Authorities and carers believe children in their care have unmet needs for their mental health (ADCS, 2022; Hiller et al., 2020; Phillips et al., 2023), which is often attributed to a lack of stability in care arrangements (Holland et al., 2005; Morgan et al., 2025; O’Connor et al., 2024; Phillips et al., 2023). However, locating attempts to access support was equally challenging, as this showed how professional language shaped the positioning of children’s needs, leading to increased administrative burdens for children in care. At such times, reflection with peers and the Experts by Experience was crucial. Discussions on how professional language could exclude needs and children’s voices with limited help, meant holding strong feelings and a sense of helplessness. As children entered care, their records became task and document-heavy, distancing from capturing their voice and reducing direct interventions, notably when services stepped back (Morgan et al., 2025).

From a practitioner-researcher perspective, this led to described moments of frustration with a sense that more was needed to support children in the records. In addition, it caused myself and others in the group to question their own use of language in interactions and how this lingers in children’s records, sometimes not helping or supporting, but instead framing how others see a child. This moment illustrates the necessity of reflexivity in qualitative work, to help position oneself against the research task (Braun and Clarke, 2006, 2022). Bringing our own history, training, and emotional responses into the act of interpretation is not just a methodological detail but an ethical challenge that needs to be embraced (ibid).

Yet reflexivity alone does not capture the full weight of encountering harm in case notes. Causer (2021) uses the language of bearing witness to describe the experiences of higher education staff after a student suicide: the responsibility of seeing and knowing, without being able to prevent. Similarly, Bosack et al. (2025) write of moral distress among clinical research professionals who witness potential harm but are constrained by their role from acting. These concepts resonate with experiences of reading children’s records. To see the evidence of need, but not being able to step into the past to intervene, our role is to analyse, not to treat. The weight of witnessing lies in holding these fragments of suffering, knowing that the possibility of action is foreclosed.

The act of witnessing, therefore, is both a burden and a form of ethical responsibility (Cox et al., 2023). To witness is to resist the erasure of suffering by naming it, by making it visible in spaces where it might otherwise be hidden. This perspective reframes encounters such as with Joe’s record. While we cannot change what happened to him, we can bear witness by refusing to treat his case notes as mere data points. By writing reflexively, by acknowledging the silences, and by situating experiences within broader patterns of systemic limitations, we can transform the act of reading into a form of ethical recognition.

At the same time, witnessing is not without cost. As Clay et al. (2024) note in their study of educators and therapists working with childhood trauma, the act of hearing or reading about suffering leaves an imprint on the witness. It is emotionally demanding to sit with stories of unmet need, systemic gatekeeping, and cultural misrecognition. In their own work, author one recognises the importance of supervision, peer discussion, and reflective writing as essential supports in processing these responses. Without such structures, the risk of secondary trauma or emotional fatigue is possible (Bosack et al., 2025).

Fragments of distress - Ethan

Ethan’s case was striking for its moments in which the raw presence of a young boy was vividly documented. One record describes him as clawing at professionals while hiding under a table, screaming, scratching, and smashing objects in sheer distress. The entry is recorded in the neutral language of incident reporting: “child distressed,” “property damaged,” “risk managed.” Yet reading it, I could not help but picture a child overwhelmed, his body telling a story of trauma and fear that the words do not capture. The image lingered not because the file dwells on it, but precisely because it does not. A moment of profound distress is collapsed into a handful of behavioural descriptors, leaving one to imagine what was felt but not said. As a researcher, this left me feeling empty, helpless, and unable to act, imagining what had happened, and again reiterated to me the importance of researching as a group, not alone. This is the paradox of case notes: they are simultaneously dense with detail and thin with meaning. As Lillis (2016) argues, institutional texts operate within prescribed genres that privilege certain forms of description, risk categories, behavioural summaries or action plans, while silencing others. Ethan’s clawing creates an image in one’s mind, but its emotional and relational context is absent. It is detailed what he did, not what it meant. The incident becomes data for risk assessment, rather than testimony of trauma.

For the researcher, such fragments pose an ethical challenge. They are powerful precisely because they resist clear interpretation. While instinct may be to see it as an embodied cry for help, there is a need to acknowledge that it is a layered reading already shaped by the observer who completed the recordings (Braun and Clarke, 2023). Therefore, witnessing episodes of distress also raises the question of how to re-present them in research. Testimony, even when partial, carries an ethical demand: it asks to be heard with care. To cite Ethan’s as “data” risks dehumanising him, reducing his suffering to a token of behavioural difficulty. Yet omitting it erases a powerful expression of his lived experience. Too much detail or quotations places the child’s identification at risk. Therefore, the challenge was to hold the fragment as testimony: to write about it in ways that preserve dignity, anonymise carefully, and resist sensationalism while still naming the harm and the challenges professionals are faced with assessing and planning supports for children in acute need.

Professionals working with traumatised children are affected by what they hear; the same holds for researchers reading what is written (Clay et al., 2024). Ethan’s clawing was unsettling. It evoked empathy and sorrow, but also unease at the distance between what one might feel and what the record shows. The file moves on quickly to referrals, care planning, and placement notes. Against this, there is a methodological opportunity. Stepanow et al. (2019) have demonstrated that narrative notes can predict seclusion in inpatient settings. Garriga et al. (2023) showed that combining structured data with narrative notes improves the prediction of future mental health crises. Although in Ethan’s case, the crisis can only be interpreted through hindsight. Ethan’s fragment might similarly be read as a signal of escalating need. Yet prediction alone is not enough. For Ethan, the account appeared archived without a follow-up therapeutic response, highlighting how needs can sometimes be visible without being documented as acted upon.

For researchers, then, fragments can draw attention to the need for research with narrative care (Randall and McKim, 2008). This involves not only analysing what is in the record but attending to how it is written, what is absent, and how we write about it ourselves (Braun and Clarke, 2024). It requires balancing fidelity to the text with recognition of the humanity behind it. Narrative care asks us to pause with Ethan’s case note, not to sensationalise it or reduce it, but to acknowledge its weight as testimony of a child’s distress. In this way, moments and traces of distress become more than behavioural incidents. They become moments that press on the researcher ethically and emotionally, encouraging a rethinking of what it means to analyse case notes to support representations of children like Ethan, positioning them as subjects of experience and learning to inform improvements to systems of care.

Containment and care - Leo

Accommodated into state care from a young age, Leo’s file constructs a different pattern: a child built almost entirely through the lens of risk and containment. Removed from home after years of exposure to violence and neglect, Leo’s placements quickly broke down. One note describes how several police officers were called to forcibly remove him after a carer felt unable to manage his behaviour. Another notes a move into residential care framed as necessary for safety and “risk management.” The language of the file is striking: Leo is sometimes written as “a danger to himself and others.” He was not considered stable enough for therapeutic engagement; only escalating levels of containment.

This mode of representation reflects what Lillis (2016) describes as the institutional shaping of texts: case notes reproduce organisational logics. In Leo’s case, the institutional priority is risk management. The record, therefore, does not narrate a child in need of care, but a risk subject who must be controlled. The administrative apparatus of trying to find a safe home, termed “placements” becomes a substitute for therapeutic support. What is preserved are descriptions of crises, police interventions, and transfers between placements, rather than narratives of healing or growth. This reading of Leo’s file is shaped by a sense of discomfort at the sheer number of professionals involved in his containment and the absence of evidence that he was offered trauma-informed therapy. Here, practitioner-researchers noticed how their own identity inclined them to look for signs of care and to be pained when they could find only control. Reflexivity requires naming this lens, but also points to the need to acknowledge, by design, a chronicle of risk. Having diverse input within the project posed challenges but also raised more questions as each perspective added layers.

Witnessing Leo’s story evokes a different ethical burden than Joe’s absence or Ethan’s fragment. Here, the harm is not in what is missing, but in what is over-represented. It can be said that sometimes we need to step back and reflect on the need for testimonies to have an ethical hearing (Causer et al., 2024). But what happens when the “testimony” we receive is written through institutional risk frameworks rather than the child’s voice? To witness ethically here suggests the need to not simply reproduce the descriptors but explore how they function, what they silence, and how they contribute to a trajectory where containment substitutes for care. Professionals in the group, grounded in values and a professional ethical base, supported understanding and guided the need for supervision, reflection, drawing attention to the importance of dignity and capacity, while also highlighting the limitations of acting when harm is witnessed (BASW, 2021).

This tension resonates with research on moral distress and moral injury. Professionals experience distress when they witness harm yet are unable to act due to systemic constraints (Bosack et al., 2025). Bearing witness to such situations and being left feeling unable to act, is recognised as a form of moral injury and draws attention to the need for support systems (Barker and Macleod, 2018; Birze et al., 2025). Leo’s file embodies the reader encountering a child navigating care systems where containment dominates and therapeutic care recedes.

As a practitioner researcher, I experienced my own moral unease: I see what could have been different, but I am bound by the role of researcher, not practitioner. Some professionals refer to Leo by his name, others as a ‘child’, and others as a ‘case’, with more senior staff appearing to see the child less, with language that distances the child’s individuality. Hospital records for many of the children in this study showed how professionals wrestled with differing understandings of children’s needs in acute situations. From social workers seeing need in the home, to doctors seeing patients in need of treatment and CAMHS practitioners navigating complex gateways among professionals.

Narrative care draws attention to the overemphasis on risk, acknowledges the absence of therapeutic narratives, and seeks to re-inscribe dignity by recognising the humanity that institutional texts often obscure (Randall and McKim, 2008). It also demands reflexivity, an awareness of how our own responses shape what we see, and a commitment to choosing language that resists reproducing harm. The involvement of EbyE introduced both challenge and grounding, connecting the work to lived realities beyond the text. The practitioners within the records, too, had stories to tell, stories that revealed the organisational and professional pressures shaping their practice. These included systemic constraints such as waiting lists, the frustration of recognising limitations without the power to intervene, and the impact of external agency decisions on children’s lives, particularly around notions of stability and need.

Leo’s lived experiences (alongside others in this study) detailed secure or residential placements functioning as perceived endpoints when acute needs remain unmet. The records bear witness to this substitution of containment for care, perhaps omitting the positive aspects of such homes and seeing them as a failure to meet social needs. As researchers, our ethical responsibility lies in making this visible, prompting an understanding and not sensationalising.

The ethics of non-intervenability - India

Perhaps the most ethically charged moments in the records were when harm was seen to escalate beyond the capacity of professionals to contain it, leading to sudden harm. India’s records detail the journey of a child through hospital and care settings who attempted to end their life, citing not seeing a way out of their situation. The incident was initially noted in a single sentence detailing their attempt to jump from a building after refusing to return to their placement. A desperate act appears flattened into a case chronology. As a social worker and researcher, I found the sentence in the record was difficult to move past. The inability to reach back into the past to call the social worker or carer, a constraint tied to my researcher role, sat uncomfortably. At such moments, a grounding in ethics and values (BASW, 2021) suggests a need to understand the boundaries and responsibilities that are possible. Here, the responsibility is to analyse the record, not to change the outcome, which cannot be changed. While the researcher’s distress is different in scale to the moral injury in direct practice, it is parallel in kind: seeing harm, imagining alternatives, yet the role as a researcher rendered one powerless to intervene.

The ethical question here is not only about feelings as a researcher but about what it means to bear witness in research. It consists of recognising and naming suffering, so that it is not erased (Richardson and Solkos, 2022). But how should this recognition occur when the available testimony itself is muted? The act of witnessing becomes an act of translation into a narrative that acknowledges the weight of what is described (Barker and Macleod, 2018). However, when faced with governance that ensures enhanced confidentiality, balancing the protection of the child is a priority, and therefore, some voices cannot be brought forward.

This raises a second dilemma: to what extent should researchers act when case notes reveal potential or ongoing harm? Deshpande and Surendra (2020) outline the ethical and legal challenges of working with mental health records, where confidentiality protections are strong and governance frameworks tightly bound. In archival research, especially with historic documents, there may be no clear path for reporting, nor any way to know if the risk remains. As research practitioners and EbyE, this produces a tension between safeguarding instincts and research constraints. Qualitative researchers need to minimise harm in their practices (Buchanan, 2024). Yet, in the context of case notes, the line between research responsibility and safeguarding obligation is less clear.

The researcher’s non-intervenability is thus both a safeguard and a burden. It protects confidentiality and ensures that archival research does not morph into clinical practice. But it also places the researcher in a position of helpless witness, reading suffering without response. Supervision to encourage processing of information and emotions, holds important here (McCarthy et al., 2020; Stout et al., 2020): to name how my practitioner identity pulls the researcher toward wanting to act, to admit the discomfort of not doing so, and to consider how that discomfort shapes the way I analyse and write due to my own positionality within research (Braun and Clarke, 2022).

Finally, there is the question of impact. Here, again, the concept of narrative care becomes crucial. By writing about children’s experiences with attentiveness, anonymity, and respect, accounts can be analysed and constructed into thematic testimony. By linking individual children’s stories to systemic patterns, research has the potential to contribute to broader understanding, policy critique, and potential changes, even in small ways. The EbyE provided a grounding and further direct windows into how services interact with individuals and the impact on the lives of both good and poor care. Here, they also provided motivation and commitment to further the project to publication. Witnessing, then, is not futile; it is a form of indirect advocacy. In this way, research can honour the children whose lives appear in case notes, even when direct intervention with the child is impossible.

Concluding comments

Clinical case notes are more than administrative records alone; they capture trauma-saturated experiences and in doing so, hold traces of the past, including pain and unmet needs. This demands careful research. Case records can be read as institutional artefacts shaped by genres of risk, accountability, and professional exchange (Lillis, 2020). Joe’s absence, Ethan’s fragment, Leo’s risk descriptors, and India’s search for a way out are all constructed through the language and logic of systems and through multiple voices. At one level, these notes can be treated as data for understanding patterns of access, delay, and intervention. Yet they also function as testimonies, preserving traces of children’s suffering, resilience, and unmet needs alongside the tenacity and performance of professionals striving to engage and support children towards a positive future.

To read case records as data alone is to risk erasing their human weight; to read them only as testimony is to ignore their institutional production. Research can hold both frames in tension in considering the use of these records against the need to uphold confidentiality. This presents an ethical dilemma of wanting to highlight situations where learning can enhance future practice, while ensuring that children’s needs and their right to privacy remain central. The value of such records lies not only in what they reveal but also in what they fail to capture, as recording evolves under the influence of risk management and legal accountability, which can expose systemic factors that are otherwise invisible to quantitative analysis (Reamer, 2025; Saxena et al., 2023).

Across the four cases, different ways of witnessing as a researcher emerged. With Joe, witnessing occurs through silence: the absence of his voice, deferrals, and the reduction of a child to an administrative process. As practitioners, there is recognition of their limited ability within a system to do more, for some witnessing involved processing. Ethan represents witnessing through fragments: a brief description of clawing and screaming that conveys intensity while withholding context. Leo’s file exemplifies witnessing through framing: institutional texts constructing him primarily as a risk subject, rather than as a child in pain. Finally, India not wanting to return to their placement at all costs, illustrates witnessing through reduction, where an act of desperation is condensed into a chronology, leaving the reader with a sense of incompleteness. These align with what Stumm (2010) and Cox et al. (2023) describe as the ethical demand of testimony: to recognise even partial or muted accounts as calls to be heard. However, there is also a need to contextualise not only the researcher’s position but also the challenges faced by practitioners. The researcher is in a position of responsibility: to acknowledge suffering without sensationalising, to name silences without presuming to speak for the child, and to situate fragments within systemic contexts. As practitioner researchers, there is also a level of recognition that professionals bound within systems are practising in action, a level of humility and empathy to the limitations and attempts made to practice ethically, not all likely to be recorded.

What unites these stories is the researcher’s inability to intervene. The researcher is intervening in a very different way by writing a new narrative through the production of knowledge. This non-intervenability is both a safeguard and a source of burden. It protects confidentiality and ensures that archival research does not slip into practice, yet it leaves the researcher to absorb knowledge of harm without recourse to action. For practitioner-researchers, this dynamic is sharpened by professional identity. Having worked in children’s services, I cannot read these records as detached data. My training pulls me toward intervention, my instincts toward safeguarding. Naming the frustration, sorrow, and helplessness that arise in reading case notes is part of the analytic process (Braun and Clarke, 2024), and as a social worker, supervision held a role in processing experiences and feelings. Rather than treating these responses as bias to be eliminated, reflexive practice acknowledges them as constitutive of meaning-making. To attend to needs as a social worker, I found supervision and reflection with others were crucial. Navigating emotional fragments left by witnessing harm without the possibility of intervention is a dynamic that Clay et al. (2024) and AbiNader et al. (2023) identify as a source of secondary trauma in research with vulnerable populations and an area that is of significance for working with data of children in acute need, and something recognised in some helping professions like social worker.

This analytic process also surfaced tensions between the roles of practitioner and researcher within a diverse research team. My familiarity with the system enabled a form of insider reading, attuned to what was unsaid and what was omitted. Yet this proximity also risked desensitisation, a normalisation of harm through repeated exposure. The act of slowing down, of reading with care and discomfort, became a methodological stance. It allowed me to resist the pull of abstraction and instead remain with the granular, the affective, and the ethically charged. As Richardson and St Pierre (2017) suggest, analysis is not merely about coding but about crafting meaning, attending to the emotional and political dimensions of the text. However, to do so requires a grounding in ethics and values (BASW, 2021; IFSW, 2018), which includes the rights of children to participate, be heard, and be supported in their recovery from harm (UNICEF, 1991). In bringing these cases into dialogue, I do not claim to speak for the children or to offer definitive interpretations. Rather, this analysis is a form of ethical witnessing, foregrounding the limitations of institutional texts and highlighting the need for ethical care. The case notes, while constrained, still carry traces of relationality, resistance, and pain. To read them well is to read beyond them: to ask what is missing, who is unheard, and how it might do better. Here, there are opportunities to bring further rights and needs into practice awareness and potential improvements.

Lessons to take forward

Drawing on this exploration of witnessing within children’s clinical case notes, there can be a suggestion in the ethical use of case records. There is a need for care in how such records are maintained, repeated and analysed. Firstly, writing with consideration to the principle of narrative care, there is a need to recognise an increased commitment to attentiveness, dignity, and reflexivity when engaging with stories of lived experience within institutional records. Secondly, to witness ethically is to resist the erasure of suffering by naming it, with care. This is not an easy task and one which relies on others to hold the mirror back and question our thinking and representation, along with realising the limitations of case records representing an aspect only of practice. Embedding narrative care within both research practice and professional recording provides a pathway toward more humane, reflective, and ethically sustainable engagement with the lives of children, who themselves access their case records into the future. Thirdly, the act of reading clinical records carries profound emotional and ethical weight; structured reflexive supervision to explore the relational aspects of researching, peer debriefings, and collaborative research processes become essential to support researchers’ wellbeing alongside supporting analytical integrity. These were integral to the project. Lastly, the inclusion of Experts by Experience and professionals from multiple disciplines deepened interpretation and reflection, though such diversity necessitates both individual and collective structures of care.