‘The human toll’: Highlighting the unacknowledged harms of prison suicide which radiate across stakeholder groups

Internationally, prisoner mortality rates are up to 50% above those in the community. Although prisoner deaths are frequent and have significant implications across a broad range of stakeholder groups, these harms are rarely acknowledged. We address this by presenting original data from semi-structured interviews with 19 senior Prison Service staff (representing 8 prisons and 11 regions) and 16 Ombudsman investigators in England and Wales. These professional groups have received limited consideration in previous research. Without negating implications for bereaved families and other prisoners, we demonstrate that scholars have grasped neither the impacts of prisoner deaths on investigators, nor the extent of the harms experienced through investigations. All stakeholders benefit from reducing prisoner suicides, but death investigations do not enable stakeholders’ ‘shared ground’ to be mobilised. Currently, death investigations compound the harms of deaths.


Introduction
Internationally, prisoner mortality rates are up to 50% above those amongst comparable populations in the community (UNOHCHR, 2019). Death numbers are rising globally because prison populations and numbers of older prisoners are growing (Roulston et al., 2021). Types of prisoner death include 'natural' deaths, homicide and suicide. Suicide forms our focus in this article. Prison suicide has been a recognised public health, criminal justice and ethical crisis for many years (Brown and Day, 2008;Zhong et al., 2021). In England and Wales, record-breaking prison suicide numbers in 2016 bereaved many and consumed around £400 million of public funds (Tomczak, 2018). Given the frequency and likely implications of prisoner deaths, the 'paucity of research on the experiences of a death in custody on staff and relatives', inter alia, is problematic (Roulston et al., 2021: 227).
We demonstrate that scholars have grasped neither: the breadth and depth of harms from prisoner deaths, which affect stakeholders around the world including bereaved families, prisoners, prison staff and societies (Tomczak, 2018(Tomczak, , 2022; nor the substantive impacts of prisoner deaths on investigators; nor the extent of the harms experienced through investigations. Deaths and death investigations can have catastrophic effects on well-being, as we highlight with reference to prison staff and Prisons and Probation Ombudsman (PPO) staff who investigate prisoner deaths. All stakeholders would benefit from reducing prisoner suicides and there is, amongst all stakeholders, ubiquitous desire to prevent prisoner suicide, but this 'shared ground' is not mobilised by the PPO's investigations, which currently create harms and compound the harms of deaths. Death investigations may inadvertently be 'doing more harm than good', potentially stimulating vicious cycles of harm and blame across stakeholder groups, which risk undermining prison safety and increasing the likelihood of further deaths. These harms must be mitigated.
Scholarship reports that suicide(s) and subsequent investigations have extensive negative impacts, for example, prison(er) safety and relations between staff and prisoners (e.g. Barry, 2017Barry, , 2020Ludlow et al., 2015;Roulston et al., 2021). In this article, we specifically examine the experiences of prison Governors (Wardens) and regional Safer Custody Group Leads (SCGLs). SCGLs are little studied but provide regional support to improve safety in custody, seeking to reduce deaths, self-harm and violence by capturing and sharing good practice and learning from deaths and serious incidents. The intersecting impacts of deaths and investigations may vary across staff roles (e.g. custodial, psychology, healthcare), levels of seniority (e.g. Operational Support Grade, Governor, regional staff and national staff) and the characteristics of individual staff (e.g. gender, race, ethnicity). Future work could valuably explore these variations. We also offer a novel illustration of impacts on the PPO staff who investigate prisoner deaths. Our focus on these three groups of professionals neither obfuscates nor minimises harms experienced through deaths and investigations by prisoners, 1 bereaved families 2 and prison staff in other roles, which scholars should consider in depth.
International laws, for example, Article 2 of the European Convention on Human Rights, require that all deaths in state detention are independently investigated (Rogan, 2018). In England and Wales, prisoner deaths are externally investigated by at least the police, PPO and Coroner. The police examine whether evidence could support prosecution, for example, for homicide or corporate manslaughter. The PPO investigate deaths by examining custody records, interviewing staff, interviewing prisoners and reviewing CCTV evidence (Tomczak, 2018). The PPO report informs the Coroners' inquest, which is an investigation and court hearing seeking to find out what happened from those involved in the care of the person who has died (PPO, 2021). These police, ombudsman and coroner investigations can be very disruptive and cause 'lengthy periods of uncertainty and anxiety' for all involved (Liebling, 1998: 69-70).

Harms for prison staff
Exposure to suicide is a primary trauma that can have a negative 'transformative' impact on prison staff: affecting their practice, coping mechanisms and relationships (Barry, 2017: 59; see also McKendy et al., 2021;Ricciardelli, 2019). The 'circles of vulnerability' model indicates that those at the centre of the circle, who discover the body, were in close familial or social circles, or felt psychologically close to the deceased are particularly likely to be affected by the death (Ayalon, 1993). Staff-prisoner proximity means that staff experiences of suicide can reflect familial bereavements (Snow and McHugh, 2002). Suicides form a significant risk factor for staff developing psychological trauma, which can affect staff long after the death (Barry, 2017;McKendy et al., 2021). Staff may develop prolonged and complicated grief (Roulston et al., 2021), involving 'absent, excessive, distorted, or unending' responses, delayed grief reactions and 'masked' symptoms or behaviours that are not easily recognisable as grief (Joy and Baron, 2008: 136). This is significant, as complicated grief may lead to suicidal ideation, depression and post-traumatic stress disorder (PTSD) Vaswani, 2014). The risks of developing complicated grief, stress and trauma are heightened for staff with high levels of involvement in the death, who have experienced prior or concurrent losses, and who perceive that the death could have been prevented (Joy and Baron, 2008;McKendy et al., 2021;Wright et al., 2006). Staff exposed to suicide have high prevalence of PTSD Wright et al., 2006). Indeed, Ludlow et al. (2015) highlight Governor perceptions that staff were traumatised by every suicide they experienced.
Prison culture can exacerbate trauma and affect abilities to grieve (Barry, 2020;McKendy et al., 2021). Prisoner and staff emotion management is shaped by the 'feeling rules' of the prison (Arnold, 2005;Crawley, 2004;Hochschild, 1983), within which 'emotions such as fear, sadness, anxiety and guilt' are deemed unacceptable (Barry, 2020: 3). As such, staff may use coping strategies to manage emotions and conceal the impacts of death, for example, using black humour to mask feelings and avoiding going to certain areas within the prison (Barry, 2017). Traumatic experiences, such as suicide, and employing coping strategies can lead to staff becoming 'hardened or disengaged' (Ludlow et al., 2015: 57) and developing 'attitudes of distrust and cynicism' (McKendy et al., 2021: 11). Indeed, deaths and investigations risk compounding prison staff tendencies to 'promote security and punishment over social justice and human rights of prisoners' (Roulston et al., 2021: 240). Suicide can cause 'institutional anxiety', leading to 'increased surveillance and security', which ultimately 'negatively affect a prison's capacity for [suicide] prevention' (Ludlow et al., 2015: 58). Traumatic experiences can thus establish vicious cycles, impacting staff ability to support prisoners and thus increasing the likelihood of further suicides (Ludlow et al., 2015).
Impacts are not limited to staff with direct involvement in deaths. Workers including correctional officers, probation and parole officers and health/psychology staff who are 'less likely to bear witness or respond to […] suicide […] are nonetheless impacted ' (McKendy et al., 2021: 6). Regarding executions in the United States, vicarious trauma-affected staff who were only indirectly involved, such as prison chaplains, wardens and correctional officers who escorted prisoners (Chiappetta and Johnson, 2021). Vicarious trauma results from empathic engagement with trauma experiences (Pearlman and Mac Ian, 1995: 558) and leaves staff susceptible to PTSD, major depression, burnout (exhaustion due to prolonged exposure to work-related problems) (Canu et al., 2021) and suicidal ideation, potentially requiring 'considerable leave from work' (Ricciardelli et al., 2021: 2), which can again establish vicious cycles that compound harms.
Investigations can also compound the harms of deaths. For prison staff, being under investigation has myriad well-being implications (Ludlow et al., 2015). As with the death, emotion management in accordance with prison 'feeling rules' influences how staff react to investigations (Barry, 2020). After a death, staff may become risk-averse, potentially becoming obsessed with procedures due to fear of further suicides and being blamed during investigations and inquests which can leave staff feeling 'defensive, resentful, and exposed' (Liebling, 1998: 81). Subsequent 'over cautious' suicide prevention efforts then negatively affect prisoner care and support (Ludlow et al., 2015: 58). Moreover, the likelihood of complicated grief is increased by staff feeling and being criticised in investigations (McKendy et al., 2021). Prison staff consider the PPO's recommendations as 'inappropriate, unrealistic or a reflection of being poorly informed about the 'realities' of prison life' and unhelpful for managing suicide risks (Ludlow et al., 2015: 58). Long investigation delays cause uncertainty and fear, and interviews and giving evidence can result in staff recalling memories that were suppressed as a coping mechanism (Liebling, 1998;Ludlow et al., 2015).

Harms for investigators
The experiences of prisoner death investigators have not yet been explored in scholarship but there is informative research with analogous groups indicating that professional investigation work can harm workers, substantiating the inclusion of prisoner death investigators in our study. For example, research in Australia on death investigations undertaken by the police briefly explored that 'death investigations are almost always challenging, emotional' and 'uncomfortable' for police officers, and noted the protracted nature of death investigations (Carpenter et al., 2016: 12). Research on medico-legal death investigators described how their 'experiences can be associated with symptoms of extreme stress [and] depression […] which can lead to poor job performance and a reduced quality of life' (Ritter, 2013: 2). 'In New Zealand, Coroners were personally impacted by the 'volume and diversity of suicides' and experienced vicarious trauma (Jenkin et al., 2019: 6). Our study develops Jenkin et al.'s (2019) findings about the impact of suicide on Coroners, highlighting the harms experienced by prisoner death investigators.

Methodology
We report findings from a research project running from 2019 to 2021, which examined how the PPO (seek to) effect change in prisons following prisoner suicides and how death investigations could have more impact on prison policy and practice. Harms was an emerging analytical theme rather than the focus of the study. Within this project, 35 semi-structured interviews were undertaken with: 16 PPO staff (who work across England and Wales from a base in London), 8 prison Governing Governors (representing 8 prisons) and 11 regional SCGLs (representing all but two regions nationally). The sample was purposive for all groups, as appropriate for our exploratory analysis and the resources available (Bryman, 2012); however, the sample is not representative of all staff in the groups we interviewed. Each stakeholder group makes a valuable contribution in and of themselves, but our combination of empirical data, perspectives and multiple observers 'adds rigor, breadth, complexity, richness, and depth' to our analysis (Denzin and Lincoln, 2000: 6).
Interviews were selected to elicit rich, detailed and wide-ranging data, and enable participants to express their point of view, along with complexities and contradictions (Valentine, 2005). The interview schedule was written formally but used flexibly and acted as an aide-memoire for topics to cover. The interview guide contained 18 questions (e.g. 'what is it like doing your job?') and 10 additional prompts (e.g. 'do you have any coping strategies for difficult emotions?'). The research project also included substantive document analysis, the results of which informed separate analytical themes and are being published elsewhere. Breadth of participant groups and numbers was prioritised for this exploratory research rather than the depth that ethnography could provide, and the less invasive interview technique increased the likelihood of recruiting participants.
Volunteer PPO participants were recruited through an email to all Ombudsman staff across Senior Investigator to Senior Management roles, and we gained participation across these roles. Face-to-face interviews were conducted in December 2019. Governors volunteered to participate following an invitation sent to 10 prisons that had recently experienced multiple suicides. SCGLs volunteered to participate following an invitation sent to all 13 SCGLs in England and Wales. Due to the COVID pandemic, Governor and SCGL interviews were undertaken by telephone and Microsoft TEAMS audio calls (at the participant's preference) in July and August 2020. Ethical approval was obtained from Her Majesty's Prison and Probation Service and the University of Nottingham. All interviews were audio-recorded with participant's consent. Data have been anonymised.
All interviews were thematically analysed. Our inductive and deductive content analysis approach drew upon theoretical a priori issues through the thematic framework of 'harm' and inductive emergent themes from patterning of stakeholder experiences (Ritchie and Spencer, 2002). Thematic coding illustrated negative impacts on stakeholders' well-being through (in) direct involvement with suicides and subsequent investigations. These harms included (i) primary and vicarious trauma, prolonged and complicated grief; (ii) internalised responsibility; (iii) the burdens of investigations; (iv) helplessness/powerlessness; (v) broken relations; (vi) burnout. These harms represented the most prevalent themes across the data set, as measured by the number of participants articulating the theme (Braun and Clarke, 2006). Analysis indicated three sources of harm: (i) the death, (ii) being investigated and (iii) conducting investigations. In the analysis which follows we discuss these themes. Our principal findings are that prison suicide results in frustration and professional burnout for prison staff and death investigators who both felt unable to deal with the 'overload' of death and questioned whether they could continue to do their work. In turn, prison(er) safety is further compromised.

Primary and vicarious trauma from the death
Whilst the primary trauma of prison suicide is particularly likely to affect those discovering the body or responding to the emergency (Ayalon, 1993), in our study, trauma was both primary and vicarious, and spread beyond people directly involved at the scene. In this section, we examine (i) how primary trauma, vicarious trauma and prolonged and complicated grief affected prison staff as a collective, with implications for prisoners and (ii) how PPO investigators also experienced (prolonged) primary and vicarious trauma, and unlike prison staff deal relentlessly with deaths.
Discovering a body, potentially at a harrowing scene, is a significant primary trauma that deeply and directly affected frontline prison staff. Governor 6 explained that their last death was due to a prisoner 'cutting his throat and his wrists, so it was really, really traumatic'. Ombudsman 5 referred to 'a horrible case where a chap set fire to himself', which was 'particularly distressing' for 'the prison and health care staff [who] had to deal with him'. Staff could be so affected by experiencing primary trauma like this that they were unable to return to work: "One guy had cut his throat with a tin lid, which was really graphic and we directly lost about four staff on that who just were so unwell they couldn't return to work, one of which was our Head of Healthcare". (Governor 7) Suicide formed a significant source of stress that was held to (in)directly affect the whole prison staff. Governor 5 described suicide as probably 'the biggest source of stress and upset' in prison work, explaining: "If the 'phone rings at 6 o'clock in the morning I know […] we have found someone dead on roll check.
[…] When the 'phone rings your heart stops".
Governor 6 described deaths as 'really tough for all of us, […] devastating' and Governor 8 confirmed that suicides were 'hugely traumatic for everybody, […] it damages people […], they are absolutely awful, they are traumatising'. The depth and breadth of prison staff pain following a suicide was confirmed by PPO staff, who voiced that during their interviews, prison and health care staff 'were very distressed' (Ombudsman 5), with Officer after Officer burst(ing) into tears' (Ombudsman 1). Together, these quotations highlight that prisoner suicides create primary trauma which spreads beyond staff directly involved at the scene across prison staff collectives, with implications for staff and prisoners. This supports McKendy et al.'s (2021: 14) finding that 'direct or indirect exposure to suicide' may cause psychological distress to correctional workers 'across professional positions'. Suicides cause significant disturbance and may lead to long periods of leave or even staff never returning to work (Barry, 2017;Roulston et al., 2021). This can result in core roles, such as the Head of Healthcare, being vacant, with substantive implications for prison regimes and staffing levels that are often already stretched (Ludlow et al., 2015). Staff absence is likely to have negative effects for prisoners, interrupting regimes, leading to extended periods being locked in cells and impeding provision of purposeful activity, family contact, vital goods and services, whilst creating additional work for the remaining staff (Ludlow et al., 2015).
Consequences of suicide were significant and could also last a long time. Within the study, there was evidence of staff suffering from prolonged and/or complicated grief. In an interview, Governor 5 explained 'I know people now still cry when talking about that case […] I'm getting emotional thinking about that'. Even with more distance from the scene, breaking the news of a death to families could produce vicarious trauma for prison staff if the families' grief was absorbed. SCGL 6 explained that multiple 'phone calls to 'parents and wives […] breaks my heart', and Governor 8 expanded: "The worst part of my job is having to go out and tell somebody that their loved one has passed away in my care and I have done that on too many occasions unfortunately". Although our participants did not spell out how prolonged, complicated grief and vicarious trauma affected staff and prison regimes, literature indicates that they can have similar implications to primary trauma, potentially requiring 'considerable leave from work' (Ricciardelli et al., 2021: 2), with all the disruptive implications described above.
There was consensus regarding the gravity of these experiences for prison staff, which is even more significant given that harms formed an emerging theme in our study rather than its focus. It is problematic that there was not sufficient recognition of how primary trauma, vicarious trauma and prolonged and complicated grief affect prison staff and, in turn, prisoners. SCGL 1 opined that it was 'easy to forget how deeply affecting a death is on an establishment and on the individuals' and SCGL 5 stated: "I can't recall the last time anybody said it's your fifth suicide, how do you feel about that now? Do you think it's impacted upon you?".
The trauma of suicide radiated beyond the prison, with PPO investigators also experiencing (prolonged) primary and vicarious trauma. PPO staff explained that their investigations included: watching CCTV footage of the death; reading reports; listening to 'phone calls; reading personal letters and notes; and speaking to bereaved relatives. Notably, this Ombudsman described other members of PPO staff discussing their traumatic experiences together, perhaps indicating emotional expression that is distinct from prison staff traumatic experiences and maladaptive coping mechanisms. Nevertheless, Ombudsman 1 reminded us that their 'job is quite solitary' and did not lend itself to 'going home and […] offloading', leaving staff 'probably processing it themselves'. PPO investigators also alluded to the prolonged nature of their primary traumas, which 'can stay with you and feel very stressful' (Ombudsman 1). Indeed, Ombudsman 3 explained: It is significant that PPO staff deal relentlessly with deaths, which prison staff do not. Ombudsman 6's described the impact of repeatedly absorbing others 'pain […] loss and […] anger': "You go into a prison and it might be your 500th interview of someone who has been first on scene but you […] have got to treat everyone with […] sincerity and interest and focus and that is fucking tiring […] I think that it is a relentless job".
Although PPO staff appeared to acknowledge the impact of trauma more than prison staff, PPO staff are experiencing endless deaths. In analogous literature, paramedics' trauma from multiple deaths 'just mounts up' and could have a cumulative 'overwhelming' psychological and 'long term emotional impact' (Harper, 2013: 21).
Investigators also experienced vicarious trauma. PPO participants described how both prisoner and staff distress transferred to investigators, who 'ingest an awful lot of […] sad stories' (Ombudsman 6), which then combined with the primary traumas of deaths they witnessed. Death investigations are important, but risk causing significant primary and vicarious trauma for investigators, potentially leading to emotional exhaustion.

Internalised responsibility
Compounding the primary and vicarious traumas of deaths, prison and PPO staff referred to internalising responsibility for the suicide. Governor 8 explained 'I want everybody who comes into my prison to serve their sentence and to leave at the end. […] I don't want anybody to die during that. So it really matters, it really does matter'. For Governor 7, deaths led to harsh self-talk: Internalised responsibility was similarly expressed by PPO staff. Ombudsman 6 described feeling: 'a massive responsibility to all sorts of people' as a result of primary and vicarious trauma from: "the many horrible pictures I've seen and horrible things […] described to me and how many people have sat in front of me and bawled their eyes out […] and how many times I have spoken to a bereaved mother or partner or daughter.
[…] You feel their pain and their loss and their anger […] it all goes in".
Literature from analogous settings highlights that nurses can perceive patient suicides to stem from their failure to care for the deceased (Robertson et al., 2010;Whitter, 2021). Paramedics also suffer from feelings of guilt following inability to prevent a death (Harper, 2013). This is very significant, because internalisation of responsibility and blame can have a 'catastrophic effect' on 'wellbeing and mental state' with staff suffering distress, anxiety and complicated grief (Whitter, 2021: 13).

The burdens of police, PPO and coronial investigations on prison staff
The impacts of being under police investigation were very significant amongst prison staff although criminal prosecutions are relatively rare and there has not yet been a conviction for corporate manslaughter or corporate homicide (Tomczak, 2018). Two Governors discussed staff spending months under police investigation, explaining the negative impacts of being under investigation on themselves and their staff. Staff experienced uncertainty as a result of the lengthy delays in investigations, which caused significant distress and fear: For Governor 4, the police investigation resulted in criminal trial of other staff members, which had left a lifelong impression and had been 'very, very difficult' and 'distressing' for the whole prison staff to manage, 'affecting the focus of the prison' and having significant implications for the regime and prisoner experience. Governor 1 highlighted that 'staff facing trial at the Old Bailey […] had a significant impact, […] people being very, very risk averse', thus increasing the likelihood of deaths (see also Ludlow et al., 2015).
Protracted police investigations are unusual, but there is a PPO investigation and inquest into every prison death, leading prison staff to feel intense professional scrutiny (Roulston et al., 2021). There appeared to be widespread substantial 'fear' (SCGL 3) and even 'terror' of PPO investigations (Ombudsman 6) amongst prison staff. Governors and SCGLs in our study were afraid that they themselves, their work colleagues and/or their whole institutions would be criticised or blamed by investigators. Governor 6 likened PPO interviews to being under 'interrogation' and felt that the PPO 'were really trying to apportion responsibility by the type of questioning'. Moreover, the target of 26 weeks for PPO investigation report, which could be delayed, created anxiety and risks: "If something is not right, I need to know about it now, […] (not) in 8 months' time (when) I could have had another 3, 4, 5 deaths […] but you wouldn't have told me about the issue because it takes so long to write the Report". (SCGL 8) Finally, the Coronial inquest concludes the investigation process, but overall this spans a significant time period, which can be distressing, contribute to prolonged and complicated grief (Roulston et al., 2021) and leave stakeholders to ruminate about the death (Spillane et al., 2019) and the potential outcomes of the investigations. Reflecting descriptions in Ludlow et al. (2015: 61) of how unhelpful it was for staff to have inquests 'hanging over' them, Governor 8 described the burdens of a long investigation process: Police, PPO and Coronial death investigations put a large accountability strain on staff, which can affect their well-being, interrupt their work, and in turn affect prisoner well-being and prison safety. Death investigations led to some staff being 'in Coroner's Court all the time' and high numbers of staff 'off with stress' (Governor 2). Governor 2 explained that this created 'a real churn in Heads of Safer Custody', which in turn creates disruption and limits actual Safer Custody work in prisons.
Staff commonly felt (and feared being) personally blamed for not preventing suicides. Being blamed and 'criticise(d) for someone dying' (Governor 5) has implications for staff well-being and prison(er) safety. Ombudsman 12 confirmed that 'everybody thinks "oh they are coming to blame us"'. Ludlow et al. (2015) highlighted that prison staff felt 'unfairly blamed' for deaths, and McKendy et al. (2021: 10) suggested that blaming individual staff following a suicide 'precipitated more stress than the incident' itself (McKendy et al., 2021: 10). This is significant because feeling blamed can leave staff feeling 'unsafe and/or ill-equipped' (McKendy et al., 2021: 11), in turn undermining prison(er) safety and well-being.
Prison staff also felt investigations made little acknowledgement of their often extensive efforts to prevent suicide. Working in prison was akin to playing 'Russian Roulette all day' (Governor 4) given high levels of prisoner self-harm, leaving 'the Prison Service […] creaking under the pressure' (Governor 7). SCGL 1 voiced frustration at the PPO's lack of acknowledgment of the 'huge amount of work' by prison staff attempting to prevent suicides, whilst SCGL 8 highlighted lack of PPO acknowledgement that prison staff 'save a hell of a lot of lives'. Indeed, staff may compromise their own safety to prevent prisoners, with Governor 4 stressing 'numerous examples where my staff have been in covered in blood, trying to prevent prisoners harming themselves'. Being 'covered in blood' is itself a potential harm to prison staff, in terms of trauma and bloodborne diseases. However, Governor 5 detailed: This lack of acknowledgement is very important. Prison staff generally feel 'unvalued by their managers and the general public', causing deep disappointment (Crawley and Crawley, 2012: 147). 'Lack of recognition of success in suicide prevention' causes frustration for prison staff (Ludlow et al., 2015: 61). For Correctional Workers, feeling unappreciated is a risk factor for mental illhealth, correlating with increased emotional exhaustion, sleep problems, work-family conflict and decreased self-perceived health (McKendy et al., 2021;Meier et al., 2015), thus increasing the likelihood of staff absence and threatening staff-prisoner relations.
In mitigation, Governors and SCGLs called for a more 'human approach' to staff in investigations, requesting consideration of the 'humanity of the work that we do' (SCGL 8). SCGL 2 asked that the PPO 'look behind my eyes' and seek to understand the impact of suicide prevention and bereavement on prison staff. Moreover, SCGL 6 described how PPO staff needed to acknowledge that: "These are members of staff that have found somebody dead and been on the floor trying to give him CPR for 45 min (and) we have had some (PPO staff) that have […] have been incredibly rude".
Governor 4 said: 'all I ask for, is the context'. It appears that that PPO staff experiences of primary and vicarious trauma, as illustrated above, was a contributing factor to rude behaviour or impatience with prison staff. Whatever the cause, these quotations indicate a (perceived) lack of human dignity towards prison staff from the PPO, which risks negatively affecting staff-prisoner relations and broader prison regimes (Liebling, 2004). Human dignity is the 'ability to establish a sense of self-worth and self-respect and to enjoy the respect of others', which is central to worker well-being and the quality of daily work lives (Hodson, 2001: i).

Helplessness and powerlessness
Working within challenging 'frontline' environments (including prisons) can induce 'hopelessness and helplessness' (Shaw, 2020). Feelings of helplessness, that is, the 'inability to mobilize energy and effort' (Shea and Hurley, 1964: 32) may result from 'elevated levels of stress' and vicarious trauma, but importantly these can be mitigated by organisational support, clear guidelines (Crivatu et al., 2021: 2;4) and feelings of dignity (Skinns et al., 2020). Fear of further deaths, being blamed and exposure to prisoner injuries negatively impact work performance and contribute to staff feeling powerless to 'avoid negative outcomes ' (McKendy et al., 2021: 12). Prison and PPO staff all revealed feelings of helplessness and powerlessness to prevent suicides. Ombudsman 10 described 'spending your entire time […] shutting the stable door after the horse has bolted', feeling that 'you can't really make things better'. SCGL 3 likened the feelings of powerlessness to: 'pushing treacle uphill' generally and referred to a particular case: "He told us he was in debt, a big sizeable debt […] we knew about this. We were kind of helpless. […] I'm not saying we could have fixed the debt but we could have heard the desperation and […] could have cared differently for him and he might not have killed himself". These references to hopelessness, despondency and helplessness illustrate stakeholders' problematic perceptions that death prevention efforts are 'doomed' before they are attempted (Shea and Hurley, 1964: 32). Feelings of helplessness and powerlessness diminish staff well-being by reducing autonomy and dignity (Skinns et al., 2020), and likely affect energy and commitment for work. Nevertheless, there was a ubiquitous desire for, and potential to benefit from reduced prisoner suicides. However, death investigations do not currently enable this 'shared ground' for suicide prevention to be mobilised, instead they compound the harms of deaths through extensive 'broken relations'.

Shared desire for prevention yet broken relations
Governors and SCGLs were motivated to prevent prisoner deaths -'to stop it altogether' (SCGL 6). Governor 1 stressed 'the last thing anyone wants is another death in custody.
[…] reducing fatal incident investigations, […] that's what we all want'. As such, SCGL 3 welcomed: 'the deepest, fullest investigation possible', considering them the only means 'of stopping it happening again'. Similarly, all 16 PPO participants sought to improve prison safety, expressing motivation to save prisoner lives and 'protect people' (Ombudsman 2) by applying their investigations to prevent deaths (Tomczak and McAllister, 2021: 220-221). However, even though prison and PPO staff shared the desire to reduce prison suicide, PPO investigations did not prevent deaths (Tomczak and McAllister, 2021).
Moreover, all of the harms described above were compounded by 'broken' relations between prison staff, prisoners and investigators after deaths and investigations. Building and maintaining positive relationships is critical for the overall prison climate, staff and prisoner well-being (Liebling, 2004;McKendy et al., 2021), and 'creating the energy and vision for change' (Liebling, 2016: 63). Positive relations between staff and prisoners improve well-being and dignity (Skinns et al., 2020) and mitigate the risk of suicides (Liebling, 2004). However, deaths can compound us/them divisions, with staff increasingly perceiving prisoners as risks that needed to be managed through 'increased surveillance and security' (Ludlow et al., 2015: 58). In our study, Governor 5 described an increasing focus on 'policing' prisoners, partly due to high staff absence, which deaths are one cause of, as highlighted above: "Where your sickness rate […] is quite high, […] stuff like key work doesn't happen as regularly".
Several staff cited concerns that focusing on security impeded staff-prisoner relations, with negative implications for suicide risks. Our study also identified tensions between 'risk' and 'care', and the potential for staff to become more punitive after traumatic experiences, for example, viewing self-harm as a disciplinary issue and responding with force and lack of humanity: "In one state of mind they can see when the one self-harming is needing support, care, concern, in another state of mind it's non-compliance, it's […] naughty […] and they should be punished. […] You see that where you see restraints on people for self-harming and […] it's that they were 'noncompliant'. Well they were trying to kill themselves; can we really call that non-compliance?". (SCGL 3) When prisoners are 'over-policed' trust is eroded, communication breaks down and tensions build between prison staff and prisoners, with distrust fuelling anger and alienation and damaging well-being (Liebling, 2016). SCGLs in our study expressed that, rather than 'over-policing' prisoners frontline staff needed to 'start refocusing a little bit' (SCGL 6) 'just talking, building relationships [with prisoners]' was 'a big thing' they 'should do better' (SCGL 11). Relationship building may be seriously challenged by deaths and investigations, but consistent personal contacts between officers and prisoners facilitate hope, decency, humanity and prevent suicide (Justice Committee, 2015;Liebling et al., 2019).
Prison staffs' 'generalised distrust of others ' (McKendy et al., 2021: 13) was exacerbated by deaths and extended to SCGLs, the PPO and Coroners, underpinning an 'us' and 'them' divide. SCGLs felt rejected by operational prison staff as the 'bad guy' (SCGL 8) after deaths and had to try to mediate prison staff 'clos(ing) ranks to look after each other' (SCGL 8). SCGLs' work to change staff practice in individual prisons and 'deliver those messages' of learning from deaths made 'maintaining positive relationships […] very, very difficult' (SCGL 2). If SCGLs became more zealous, perhaps as a reaction to their primary and vicarious trauma and started to 'bang the drum', they became even 'more unpopular', which could even 'feel personal sometimes' (SCGL 8), setting up another vicious cycle where prisons could become more unsafe after deaths.
Despite 'all trying to achieve the same thing', there was an 'us and them thing about the Prison Service and PPO' (SCGL 3). PPO investigations compounded prison staffs' generalised distrust of others. Prison staff felt the PPO targeted individual prison staff for criticism and that the PPO's mindset was that prison staff are 'all a bit shit' (Governor 5), which in turn explains why prison staff have 'such negative perceptions of the PPO' (SCGL 8) which reinforces divisions and impedes learning. Prison staff hostility and fear of inquests also negatively impacted relations with Coroners. SCGL 1 described a case where prison staff were not forthcoming with information for the inquest, which 'irritated the Coroner immensely' and likely reinforces all the hostility and negative feedback loops we have set out.
Prison staff hostilities were reciprocated by PPO investigators, who were frustrated by prison staff appearing 'blasé' about investigations, perhaps due to fear, but overall this reinforced divisions: PPO staff were angered due to 'see(ing) the same issues happening again' (Ombudsman 2) in the deaths that they review relentlessly, in graphic detail. The following quotation from Ombudsman 5 exemplifies the PPOs' sense of frustration with prison staff: "You feel so annoyed and frustrated because these are things that we have said over and over again.
[…] I said the same things a year ago and they promised […] they were going to action them and […] clearly haven't".
Ombudsman 3 interpreted this lack of action rather simplistically, as prison staff not knowing 'what has gone wrong or they are not that bothered'. In response, SCGL 8 explained that the PPO had 'picked up an awful habit […] of naming individual staff', but this blaming approach only compounded prison staffs' 'negative perceptions of the PPO'. As previous work has highlighted, consistently repeating recommendations and heightening blame does not reduce vulnerability to deaths but establishes vicious cycles by demoralising staff, thus threatening well-being, productivity and motivation to adapt practice (Tomczak and McAllister, 2021: 224).
Deaths and investigations are also likely to impact personal relationships. Although we did not gather data on this, prison work can shape social relations if staff become, for example, emotionally detached, hypervigilant, overly sensitive or fearful for loved ones' safety (McKendy et al., 2021). Moreover, primary and vicarious trauma risk profoundly impacting prison and PPO staff members' social ties, including 'familial, social and psychological proximity networks' (Crivatu et al., 2021;Spillane et al., 2019: 1). Common effects can include staff feeling isolated from, and unable to speak, to family or friends (Crivatu et al., 2021).
Many participants also described how dealing with death and investigations could result in burnout.
Cumulative 'overload of death' and burnout Many stakeholders described how the continuous cycle of dealing with suicide and investigations led to 'burnout': 'a psychological strain caused by continuous stressors which erode individuals' coping resources over the long term, causing physical and emotional and mental exhaustion' (Etzion, 2003: 214). The risk of burnout is intensified for prison staff due to cumulative suicides being commonplace (McKendy et al., 2021;Roulston et al., 2021), which also applies to PPO investigators. Dealing with multiple suicides had a 'cumulative effect' (SCGL 5), staff were left 'really struggling' due to 'years of dealing with death and the emotions around it' (Governor 7). Dealing with suicide on an ongoing basis was 'mentally' and 'emotionally draining' (Ombudsman 12 and 13) for staff with staff being off work for long periods of time. Staff described being 'up to here with death' (Ombudsman 11) and questioned whether they could continue to cope with the 'overload of death' (Ombudsman 6).
Moreover, repeating recommendations that failed to initiate change and prevent prisoner suicide was one of 'the biggest frustration(s)' for PPO staff (Ombudsman 2). As illustrated above, repeating recommendations was a 'continuous stressor' that placed a psychological strain upon PPOs wellbeing, productivity and motivation (Tomczak and McAllister, 2021): Dealing with prisoner suicide and investigations on an ongoing basis could add up over time, becoming insurmountable and resulting in prison and PPO staff burnout, with implications for staff performance and retention, for example, burnout is one of the main reasons ambulance workers often take early retirement (Rodgers, 1998).
Furthermore, the risk of burnout was compounded by personal issues. Difficulties in PPO investigators' personal lives made doing death investigations 'very difficult' (Ombudsman 12) and potentially led to complications such as 'panic attacks' (Ombudsman 5) and to staff not being able to continue in their role: Personal 'adverse situations' can impact abilities to cope professionally (Crivatu et al., 2021). Trauma workers who 'personally experienced adverse situations […] presumed traumatic' were more likely to be professionally vulnerable to vicarious trauma (Crivatu et al., 2021: 2). Personal bereavements were particularly difficult and led to a cumulative 'overload of death' (Ombudsman 6), with staff having to take significant time off work. We now discuss the implications of all these findings.

Discussion
This article has explored the breadth and depth of the harms of both prisoner suicides and their investigation upon prison staff, with a particular focus on Governors and previously unacknowledged SCGLs. We have deepened our understandings of the harms of suicide and investigation, illustrating that although it is 'easy to forget how deeply affecting a death is on an establishment and on the individuals' (SCGL 1), harms are not limited to those who directly witness suicide but affect the whole prison staff, with implications for staff and, in turn, prisoners over potentially long time periods. We have also examined how these harms radiate beyond the prison to affect PPO death investigators, a group who deal relentlessly with deaths and are particularly vulnerable to burnout, but whose experiences have not previously been studied. This is not merely an 'incremental' advance in knowledge, but rather opens up analysis of the (mis)application of death investigation processes, which risk creating and compounding harms and broken relationships between multiple groups, in turn constraining the learning from death investigations and their implications for practice.
Staff exposed directly and indirectly to suicide and subsequent investigations suffer a myriad of harms. Primary trauma, vicarious trauma and prolonged and/or complicated grief were evident across prison and PPO staff, across professional positions, (in)directly exposed to deaths, which potentially set up continuous cycles of 'absorbing' others 'pain […] loss and […] anger' and an 'overload of death' (Ombudsman 6) to the extent that some staff were no longer able to continue in their role due to burnout. Prison and PPO staff internalised responsibility for deaths, yet both often felt helpless and powerless to prevent death. Internalising responsibility is an important area for mitigation as this can have 'catastrophic' effects on 'wellbeing and mental state' (Whitter, 2021: 13), which are counterproductive for prison safety.
Research in analogous settings has identified the personal responsibility, guilt and blame that healthcare professionals (including nurses and paramedics) can feel for not being able to prevent a death (Harper, 2013;Robertson et al., 2010;Whitter, 2021). For prison staff, this internalisation of responsibility appeared to be compounded by fear of further deaths and being blamed in death investigations which direct responsibility for suicides towards individual staff and individual prisons. Whilst suicide prevention is of utmost importance and must be the concern of all members of prison staff (HM Inspectorate of Prisons, 1999), deaths are frequent in prisons and not all suicides are preventable (Walter and Pridmore, 2012). Moreover, prison suicides are produced by multiple systemic factors including rates of imprisonment, staff cuts, often decrepit Victorian prison buildings, and inappropriate substitution of imprisonment for secure psychiatric care. As such, it is important to reconsider the 'dominant professional and political' accountability discourse (Robertson et al., 2010: 6) that 'all suicides are avoidable' (Whitter, 2021: 13), and to problematise the causes of prison suicides, moving beyond the actions of individual members of staff. The PPO could do much more to highlight these multiple systemic factors (Tomczak and Banwell-Moore, 2021).
The PPOs repeated recommendations failed to problematise the causes of prison suicides and instead left prison staff feeling targeted and blamed. Despite our findings revealing a ubiquitous desire to prevent suicide and potential to benefit from reduced prisoner suicides, death investigations, in their current form, do not enable stakeholder 'shared ground' to be mobilised. Death investigations compound the harms of deaths by contributing to us/them divisions and 'broken relations' between stakeholders. Dignity is key to the establishment of trust, which can improve social order and relationships, and facilitate progress and change within prisons (Liebling et al., 2019). The PPO and prison staff must consider how to foster positive relations built on trust and dignity, which could create 'the energy and vision for change', improve prison(er) safety and prevent suicide (Liebling, 2004;2016: 63).
Us/them divisions and the fear of being blamed and further deaths risk compounding prison staff tendencies to 'promote security and punishment over social justice and human rights of prisoners' (Roulston et al., 2021: 240), which in turn 'negatively affect a prison's capacity for [suicide] prevention' (Ludlow et al., 2015: 58). Moreover, the PPO's lack of 'acknowledgement of what has gone right' (Governor 5) may contribute to prison staff feeling helpless and powerless to remedy problems, eroding confidence and leading to staff becoming defensive, consequentially further undermining prison(er) safety and the success of suicide prevention efforts (Ludlow et al., 2015: 57-8). Mitigating tendencies to over-police after deaths is an important area for attention in practice, and it is essential to consider how investigations could mitigate and exacerbate this tendency.
There is clear potential for the Prison Service nationally, regional staff and staff in individual prisons to actively acknowledge the implications of suicide. Doing so might be challenging given prison 'feeling rules' that discourage acknowledgment of the implications of suicide, but is essential work that could interrupt the negative transformative impacts of suicide that impact staff ability to support prisoners and thus increase the likelihood of further suicides (Barry, 2017;Ludlow et al., 2015). Furthermore, the PPO and broader Civil Service as an employer must rigorously engage with means of mitigating these harms to investigators themselves. Investigators' primary and vicarious trauma may (partially) explain the limited impact of PPO investigations and failure to reduce prisoner deaths (Tomczak and McAllister, 2021). Reflecting on how the trauma of investigations influences PPO findings and impacts would be a useful exercise for the PPO.
Moreover, providing relief from continuous job stressors and burnout (Etzion, 2003) by interrupting the vicious cycles that prisoner suicide and investigations can produce is critical to staff retention, stakeholder well-being and prisoner safety.

Conclusion
This study demonstrates that failure to acknowledge the myriad harms of prisoner suicides and death investigations, as affecting prison staff and death investigators inter alia, maintains the 'black box' that conceals potential reasons why long and expensive prisoner death investigations do not currently reduce suicides or improve prison safety. By demonstrating how the harms of prisoner deaths and investigations are broadly unacknowledged and radiate widely we seek to stimulate both (i) more substantive support for all those caught up in prison suicides and death investigations and (ii) reconsideration of how prisoner deaths are investigated. We call on the Prison Service and PPO to refocus on how prisoner suicide and death investigations affect both their own staff and their counterparts, and adapt their practices accordingly. This is particularly relevant for the PPO, whose counterpart Prison Service staff are the subjects that they seek to influence. Indeed, we wish to direct the PPO and also Coroners to international guidance including the Minnesota Protocol on the Investigation of Potentially Unlawful Death (OHCHR, 2016). This highlights that: 'investigators must take care to minimize the harm that the investigation process may cause, especially regarding the physical and mental well-being of those involved in the investigation' (OHCHR, 2016: 11, s.42); 'investigators conducting interviews should approach all witnesses with an open mind and observe the highest ethical standards. A careful assessment of risk, strategies and adequate human and financial resources must be in place to ensure the safety and security of all witnesses […] Careful attention should also be paid to the safety of the investigator' (OHCHR, 2016: 11, s. 71); and that 'interviews should be conducted in a way that […] minimizes as much as possible any negative impact the investigation may have on interviewees. Special care should be taken […] to prevent their retraumatization' (OHCHR, 2016: 18, s.85).
We have argued that greater recognition of the harms of deaths and investigations, and the vicious cycles that they establish is required, and that a new approach to death investigations informed by these harms must be developed and implemented. This new approach must reconsider the extent of individual staff responsibility in producing deaths and mitigate the consequences of the PPO's default approach of blaming individual members of staff and individual prisons for deaths. Just as prisons require positive social relations between staff and prisoners, effective prison oversight requires positive social relations with prison staff to better influence conditions and safety for prisoners.

Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/ or publication of this article.

Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/ or publication of this article: This work was supported by the UK Research and Innovation, Economic and Social Research Council under the University of Nottingham Impact Accelerator Award and Nottingham Impact Accelerator Knowledge Exchange Prize, Research England (grant number MR/T019085/1, SPF-QR funding awarded to the University of Nottin).

Data accessiblity statement
Anonymised interview transcripts for participants who consented to data sharing, plus other supporting information, are available from the UK Data Service, subject to registration at http:// doi.org.

Notes
1. Witnessing or being in proximity to suicide can leave prisoners mentally scarred: distressed and suffering from PTSD (Wainwright and Decodts, 2020). 2. Both the death and subsequent investigation can create and compound significant harms for bereaved relatives. The substantial burden of investigations alongside lack of trust in their ability to provide answers and/ or stimulate change risks producing secondary victimisation, prolonged and complicated grief, trauma and frustration in bereaved families (Roulston et al., 2021;Shapland et al., 1985;Author, forthcoming).