Improving the quality of prison research: A qualitative study of ex‐offender service user involvement in prison suicide prevention research

Abstract Background Suicide is the leading cause of avoidable death in prisons worldwide and suicide prevention is an international priority. Consequently, there is an urgent need to develop evidence‐based treatments. We conducted a randomized controlled trial of a novel suicide prevention psychological therapy for male prisoners. To promote ecological validity by addressing the “real‐world” situation of suicidal prisoners, we involved a consultant group of ex‐offenders with past experience of being suicidal during imprisonment. Service user involvement in prison research is challenging and underdeveloped. Objective We aimed to investigate the ex‐offender service user consultants’ experiences of being involved in the research. Design Individual qualitative interviews were conducted and analysed using an Interpretative Phenomenology Analysis (IPA) framework. Setting/participants The study was conducted at a university in North England, UK, comprising four ex‐offenders with experience of being suicidal during past imprisonments. Results Two superordinate themes were identified: “Working Together” depicted participants’ perceptions of the pivotal role of good relationships with researchers, and “Journey of Change” outlined how participants’ involvement in the research impacted on their personal lives. Discussion Little is known about how to successfully involve ex‐offender service users in research. Our results indicate the conditions necessary for successfully engaging ex‐offender service users in research and have important implications for improving the quality of prison research. Conclusions Involving forensic service users in research is feasible and should be encouraged, as despite certain challenges, it is highly rewarding both for the research and the ex‐offender service users.


| INTRODUCTION
Suicide is a global health problem in custodial and correctional institutions being cited as the single most common cause of death for incarcerated people. 1 A study examining prisoner suicides in twelve countries found rates in Western Europe higher that in Australia, Canada and New Zealand. 2 In the United States of America, suicide is the leading cause of death in local jails and state prisons, representing 34% of all inmate deaths. 3 In England and Wales, suicide is the leading cause of preventable death in prisons with prisoners being eight times more likely to die by suicide than the general populations. 2,4 During 2015 to 2016, self-inflicted deaths by prisoners in England and Wales increased by 28% to an unprecedented high rate of 1.2 per 1000 prisoners. 5 Consequently, suicide prevention in prisons has been hailed as an international health priority by the World Health Organization, 1 and similarly is a priority for the UK Ministry of Justice 6  Few people, other than those who either work in, or have been incarcerated in, a prison have any authoritative "lived experience" of the realities of prison life. 11 Service user involvement in research is well established within the UK Department of Health policy, [12][13][14] being recognized as a hallmark of good research practice in many contemporary fields of study. 15,16 However, service user involvement in prison research remains underdeveloped 17,18 with little published research examining how to engage and retain individuals with experience of imprisonment in the research process. [19][20][21] Furthermore, from the small corpus of existing publications, some studies purporting to describe prisoner or ex-offender service user involvement have employed research methodologies where service users were research participants rather than collaborators. 22 This deviates somewhat from the definition of involvement stated by INVOLVE 16 (p7) as "research being carried out 'with' or 'by' non-professionals rather than 'to', 'about' or 'for' them." Conducting research in prisons is challenging due to the complexity of contextual variables unique to prison settings. [23][24][25][26] Barriers to successful implementation of prison research include the necessity to work within stringent security regulations and the prevailing organizational culture and attitudes of prisoners and staff. 27 Given the uniqueness of the prison setting and the lack of contextual knowledge accessible to mainstream researchers, it is especially important to involve individuals with "lived experience" of imprisonment as service user advisors or consultants to bring a "real-world" perspective to the design and implementation of prison research. The purpose of this study was to investigate and understand SURG member's experiences of being involved in the PROSPeR study.

| The service user reference group
Recruitment of ex-offenders with experience of being suicidal during imprisonment was achieved by phone or email approaches to organizations likely to be in contact with ex-offenders (e.g probation services, prison charities, user groups). Those willing to help were then asked to circulate a poster inviting interested ex-offenders to a meeting at the university where information was provided about the study and SURG. Seven people attended the meeting following which they were invited to meet research staff individually to ascertain their possession of the required "lived experience" and ability to attend meetings regularly. This resulted in five people wishing to join the SURG; however, one member dropped out after two years for reasons unknown.
Throughout the three-year study, four SURG members attended monthly research meetings at the university to guide the research team throughout all aspects of the research process. The SURG was founded on the collaborative model of involvement advocated by INVOLVE 16 whereby contributions of the researchers and SURG members held equal value despite emanating from different perspectives.
Further details concerning the recruitment strategies and operational structures and processes we developed are described elsewhere. 28

| Design
In-depth individual qualitative interviews, following Interpretative Phenomenological Analysis (IPA) methodology, 29 were conducted.
We aspired to elicit the particular subjective experiences of each SURG member in accordance with IPA's ideographic focus. 30

| Recruitment
Ethical approval was obtained from the university's Research Ethics Committee (reference number 11473). Recruitment and data collection were conducted by CM who was independent from the PROSPeR research team. Information about the study was offered by CM at a SURG meeting.

| Eligibility
Inclusion criteria specified current membership of the PROSPeR SURG and ability to provide informed consent. There were no specified exclusion criteria.

| Participants
To preserve the anonymity of this unique sample, we have limited the provision of demographic data, and participants are identified by a pseudonym. All four SURG members gave written consent to participate. The sample had an equal gender split, mixed ethnicity and age range of 40-60 years old. Past imprisonment ranged from 1 to 28 episodes equating to almost 50 years collectively.

| Procedure
Interviews were carried out at the university at times convenient to participants. A semi-structured topic guide (available upon request) posed questions to initiate conversations about participants' experiences as SURG members. Questions investigated participants' motivations and expectations of joining the SURG, views about how they were involved and how involvement had impacted on them personally. Questions also probed participants' views about relationships with fellow SURG members and the research team, along with any challenges or disappointments encountered. Interviews lasted approximately one hour (range 50-66 minutes) following which they were transcribed verbatim by the interviewer who removed all identifying information.

| Analysis
As advocated for IPA, each transcript was analysed individually. 31 Several readings of each transcript preceded "initial noting" of descriptive comments which were then coded to form "emergent themes." Clusters of related themes were then abstracted into superordinate and subordinate themes. This was repeated for each transcript prior to searching for similarities, idiosyncrasies and patterns across the entire data corpus. The first and second authors each carried out these procedures independently before meeting together to review interpretations. Final interpretations were agreed with all research team members.

| RESULTS
There were two superordinate themes: (i) "Working Together" captured participants' perceptions of factors attributable to their successful working relationships with researchers; (ii) "Journey of Change" depicted SURG members' personal journeys as they embraced both the challenges and opportunities encountered. Each superordinate theme comprised three subordinate themes (See Figure 1).

| Being listened to
Participants' past experiences within other groups was not always positive. This influenced their expectations and concerns about involvement with PROSPeR. Being able to talk freely and feel "listened to" was a key concern. Sam reflected back to his thoughts during initial SURG meetings: Expectations of involvement with the study are likely to have been influenced by participants' past experiences of imprisonment, where some felt that their views were neither sought nor respected. Being "listened to" was perceived as a welcome, albeit unexpected, surprise.   The tone and energy of Ted's language indicated his pleasure but equally his incredulity that his past criminal career had paradoxically led him into a different career within academia.

| Changing self-perceptions
Self-perceptions gradually changed as participants became aware that they possessed a valued specialist body of knowledge and expertise gained from their past experiences of imprisonment from which they now had something to offer society:

| DISCUSSION
This study investigated the experiences of a small group of community-dwelling ex-offenders in their role as word deleted "fo- pants' focus on talking about their changed self-identity largely concerned movement from offender to non-offender status.
Participants described how involvement in the study precipitated an internal process of reflection whereby they challenged and rejected their criminal identity to assume a new "changed person" identity.
This was an unexpected outcome of their involvement with the study both for the researchers and the ex-offenders. Goffman's 32 seminal works on stigma postulated how labelling of deviants leads to "spoiled identity." Subsequently, Lemert 33 described how societal "labelling" of criminals often leads to the labelled individual's acceptance and internalization of self-stigma based upon the perceived views of others.
Conversely, "delabelling" 34 as a precursor of "changed person" status is professed to carry greater impact when validated by "certification" from respected officials. 35 This may explain the high acclaim with which participants held their university honorary staff contracts.
Training is recommended to empower service users to make a meaningful contribution to research 36,37 . Therefore, SURG members were given honorary contracts to enable access to university educational resources. However beyond providing training, the honorary staff identity card was valued by SURG members as a "badge of honour" marking their acceptance and status by a respected official institution. It was also used as "proof" to convince sceptical others that old criminal lifestyles were redundant.
People with criminal backgrounds are traditionally considered "hard to engage" in formal processes, 38  misdemeanours, appeared highly influential to participants' sustained engagement in the study. Indeed, engagement in new prosocial roles marked by strong affiliation with an institution has been found to be influential in maintenance of desistance. 41 Being involved with PROSPeR appeared to provide a dynamic change force epitomizing a "turning point" 34 influential to supporting participants' desistance.
Although the value of developing a positive environment to promote service user involvement has been described for forensic 42 and general health populations, 37,43 our experiences suggest that this may assume greater importance for ex-offender groups.
Motivation to become a service user consultant in research is often driven by a passion to "make a difference" 44,45 or to "give something back", 43 and these ideals were clearly evident for SURG members. Participants' experiences mirrored a process of post-traumatic growth whereby positive personal "growth" results from traumatic events. 46 Participants described how they reconceptualized past negative experiences of imprisonment in terms "making good out of bad" recognizing that it had enabled them to help others and to also help themselves. Positive psychology approaches that foster opportunities for post-traumatic growth within offender rehabilitation programmes have been described. 47 Interestingly, whilst the narratives from Ted, It is, therefore, not insignificant that our ex-offender SURG members revealed that an unintended outcome of their involvement with PROSPeR involved a move away from crime.
Finally, a further consideration is the personal benefits to exoffender service users of involvement in studies such as PROSPeR.
Ex-offenders are likely to experience challenges such as reintegration into conventional society, financial hardship and restrictions on employment opportunities due to being imprisoned. Our work showed that our SURG members perceived a range of benefits from being part of this group which nurtured their enthusiasm and desire for remaining with, and actively contributing to the project.

| Strengths
This study has contributed to the knowledge base in the underresearched area of service user involvement in prison mental health.
Our study offers valuable insights for other researchers of effective ways of involving ex-offender service users and other "hard-to-reach" populations. Our findings suggest that good practice in service user involvement may extend impact beyond the intended outputs of improving the quality of the research to also impacting on the lives of service users that can lead to transformational personal change.
Considerable attention to ensuring scientific rigour was invested in the current study: (i) interviews were conducted by a researcher who was external from the PROSPeR study thereby minimizing potential social desirability bias; (ii) to ensure high-quality interview data, and with reference to Yardley's principles, 51

| Limitations
Although our sample comprised only four participants, this represented the entire membership of the SURG, who between them had over 50 years experience of imprisonment. A prevailing feature of IPA is a focus on ideography with small homogenous samples. 29 Although our sample was small, the level of rich detail of the experiences of this unique population of ex-offenders is striking. In keeping with the aims of qualitative research, we do not make any claims for empirical generalizability 29 of our findings. We do however suggest the potential for "theoretical transferability" of an appropriate level of our findings based on assimilation of readers' own professional knowledge and experience. 29,59 The organizational constraints pertaining to the PROSPeR study were immense, and we believe the SURG members were highly influential to its success. There are challenges to recruiting ex-offender service users who are willing to get involved in research and we believe this may be the first and possibly only such group currently in existence. As such our sample size reflects, the very small pool of ex-offenders contributing to research currently. Further studies with ex-offender service user consultants are required to elicit a wider range of experiences.

| CONCLUSIONS
This study has shown that it is possible to engage and work effectively with ex-offender service user consultants and outlines impor-