Cognitive Behavioural Suicide Prevention for Male Prisoners: Case examples

Abstract

Suicide is a serious public health problem but a problem that is preventable. This complex and challenging problem is particularly prevalent amongst prisoners; associated with a five-fold increase in risk compared to the general community. Being in prison can lead people to experience fear, distrust, lack of control, isolation, and shame, which is often experienced as overwhelming and intolerable with some choosing suicide as a way to escape. Few effective psychological interventions exist to prevent suicide although cognitive behaviour therapies appear to offer some promise. Offering cognitive behaviour suicide prevention (CBSP) therapy to high risk prisoners may help to reduce the likelihood of preventable self-inflicted deaths. In this paper we present three cases drawn from a randomised controlled trial designed to investigate the feasibility of CBSP for male prisoners. Implications of the current findings for future research and clinical practice are considered.

Introduction

The prevention of suicidal behaviour is a high priority for healthcare providers (Department of Health[DH], 2012; Department of Health & Human Services, 2012) and yet it continues to be a complex and challenging problem. In the UK, more than 6000 people take their own lives each year reflecting an annual rate of suicide of 11 per 100,000; a rate which has remained largely unchanged for over 30 years (Office for National Statistics, 2013). In addition to completed suicides, a consideration of the prevalence of suicidal ideation is important. Approximately 1 in 6 people will experience suicidal ideation at some point in their lives, which will drive 1 in 20 into making an attempt (Bebbington et al., 2010). This equates to a person dying from suicide every 2 hours and an attempt being made every 6 minutes. Of course, not all individuals who engage in suicidal ideation or behaviour will eventually take their own lives, but all aspects located along the suicidal continuum are accompanied with significant, distressing, disruptive and undesirable psychological states worthy of therapeutic intervention (Tarrier et al, 2013).

The large body of epidemiological research into completed suicides has enabled the identification of key characteristics associated with an exaggerated risk. Typically, a high risk profile would be of a young male who is less ‘integrated’ within his community, so more likely to be single or divorced with no children and unemployed. He has an almost 90% likelihood of experiencing a diagnosable mental disorder, most likely depression, substance use, personality disorder and/or psychosis (Arsenault-Lapierre, Kim & Turecki, 2004).

The high risk profile can be seen to describe a substantial majority of the prisoner population, which have been shown to have a different demographic than the general population. Typically, male prisoners represent approximately 95% of the inmate population, with most establishments restricted to male-only prisoners. The age of prisoners upon reception tends to be between 18 to 35 years (Teplin, 1990; Andersen, Sestoft, Lillebæk, Gabrielsen, & Kramp, 1996; Bland, Newman, Thompson, & Dyck, 1998), with those aged below 18 years detained in Young Offender Institutions. The backgrounds of prisoners contain an exaggerated likelihood of childhood neglect, low levels of educational achievement, perhaps explaining the below average levels of intellectual and cognitive functioning reported for adult prisoners (Birmingham, Mason, & Grubin, 1996; Davidson, Humphreys, Johnstone & Owens, 1995). Almost half of prisoners say they have no academic qualifications, compared to 15% of the general population (Ministry of Justice[MoJ], 2012a).

Prisoners are a socially excluded population (Social Exclusion Unit, 2007), with unemployment, poor housing, financial difficulties, and loss of access to family and close support significantly more common than in the general population (Birmingham et al., 1996; Brooke, Taylor, Gunn, & Maden, 1996; Teplin, Abram & Mcclelland, 1996; National Offender Management Service, 2007). The ‘prisoner experience’ has been shown to be severely detrimental to the individual’s mental health and wellbeing (Birmingham, 2003). The social and health inequalities that are brought with the person as they enter custody, referred to as “imported vulnerability”, highlights the complexity of needs and challenges facing offender health and social care services responsible for meeting prisoners’ needs.

The rate of mental health problems in prisons is notoriously high. Up to 90% of prisoners have a diagnosable psychiatric disorder (DH, 2005; Royal College of Nursing, 2009) with 70% having two or more co-morbid diagnoses (DH, 2008). Prisoner groups typically have complex and long-standing mental health problems, such as psychosis, personality disorder, anxiety and depression, often co-morbid with substance and/or alcohol misuse (HM Inspectorate of Prisons, 2007). For instance, half of female and a quarter of male prisoners reported clinical levels of anxiety and depression, compared to 16% of the general population, and 25% of female and 15% of male prisoners reported symptoms of psychosis, compared to a rate of 4% in the community (MoJ, 2013a; Wiles, 2006).

Suicide behaviour is far more common within prisons compared to the community. Annual suicide rates of over 60 per 100,000 prisoners are 5 – 8 times that reported for the general population (MoJ, 2012b; Fazel, Grann, Kling, & Hawton, 2011) leading some to describe suicide as the leading cause of preventable death in prisons (Baillargeon et al., 2009). In addition to those suicide risk factors shared with the general population, prisoner populations experience additional risks due to the prison context. Overcrowding (Leese, Thomas, & Snow, 2006); extended periods of isolation (Bonner, 2006); interpersonal violence from other prisoners and subsequent traumatic stress responses (Blaauw, Arensman, Kraaij, Winkel, & Bout, 2002) have all been shown to heighten the risk of suicide behaviour. Coping with a prison environment that engenders fear, distrust, and a lack of control, can leave prisoners feeling overwhelmed and hopeless, leading some of them to choose suicide as a way to escape (Birmingham, 2003; Fazel et al, 2011). As such, prisoners have continued to be identified as a key high-risk group in the updated suicide prevention strategy for England and Wales (DH, 2002; 2012).

In the UK, prisoners have the right to expect an equivalent healthcare service as the general public receive (Home Office, 1990; 1991) with NHS mental health in-reach teams (MHIRTs) responsible for the delivery of mental healthcare for prisoners (HM Prison Service and NHS Executive, 1999). The demands placed upon many MHIRTS have exceeded their ability to supply good quality healthcare, especially to those at risk of suicidal behaviour (Brooker, Ricketts, Lemme, Dent-Brown, & Hibbert, 2005; HM Inspectorate of Prisons, 2007; Bradley, 2009). Prisons in England & Wales currently operate the Assessment, Care in Custody and Teamwork (ACCT) system, which aims to provide individualised care and support for prisoners at risk of suicide or self-harming behaviours (HM Prison Service, 2005). The ACCT system offers both crisis interventions as well as multi-disciplinary care to those with longer-term problems. ACCT can be aligned to the Care Programme Approach used within UK mental health services, with a focus beyond the surveillance and monitoring of prisoners to also include an individualised and interactive process to positively manage the risks presented by the prisoner. An ACCT is said to be ‘opened’ for a prisoner when a risk becomes known to staff, and remains open while the risk persists, during which time fortnightly reviews are undertaken by prison staff which healthcare staff and the prisoner should also contribute towards. When the level of risk is considered to be safely reduced, the ACCT is ‘closed’. Previous evaluations have reported this approach to supporting suicidal prisoners to be sufficiently sensitive in that the help provided is being delivered to high-risk individuals, however there remains considerable unmet need amongst the prisoner population with substantial proportions of suicidal prisoners failing to be identified as at-risk (Senior et al, 2007; Humber, Hayes, Senior, Fahy, & Shaw, 2011).

Empirical evidence for treatments shown to be effective in the prevention of suicide behaviour is limited although psychological treatments, particularly cognitive behavioural therapies, have attracted considerable interest with preliminary findings indicating significant promise of a preventative effect. In a review and meta-analysis of 25 studies of cognitive behavioural interventions for suicide behaviour, a highly significant overall effect was reported (Tarrier, Taylor, & Gooding, 2008). The review highlighted group CBT interventions were ineffective whereas individual sessions alone or when coupled with group sessions were highly effective. Importantly, CBT was only found to be effective when the therapy was directly focussed upon the prevention of suicidal behaviour, whereas suicide prevention viewed as a secondary gain within the treatment of another mental health problem, e.g., CBT for depression or psychosis, was ineffective. Since the review, this evidence base has continued to become more established. In a trial of 10 sessions of cognitive therapy following a recent suicide attempt, relative to participants receiving usual care, CBT recipients were 50% less likely to re-engage in suicide behaviour in the subsequent 18 months and achieved significant improvements on measures of depression and hopelessness and the rate of recovery for problem-solving skills (Brown et al., 2005; Ghahramanlou-Holloway, Bhar, Brown, Olsen, & Beck, 2012). Similarly, in a sample of 90 patients presenting to a local medical centre following suicidal behaviour, those randomised to receiving 12 sessions of CBT reported significantly reduced levels of suicidal ideation, improved problem solving ability and improved self-esteem, compared to the standard care group (Slee, Garnefski, van der Leeden, Arensman, & Spinhoven, 2008).

Most recently, a cognitive behavioural suicide prevention (CBSP) treatment was developed by the authors (Tarrier et al, 2013) which offers a structured, theoretically-based, psychological intervention designed to address and amend the specific psychological architecture responsible for suicidal behaviour. The CBSP treatment protocol was informed by a theoretically derived psychological model of suicide behaviour; the Schematic Appraisals Model of Suicide (SAMS; Johnson, Gooding, & Tarrier, 2008) which has been empirically validated in people experiencing suicidality, psychosis, and post-traumatic stress disorder (Johnson, Gooding, Wood, & Tarrier, 2010a; Johnson et al., 2010b; Panagioti, Gooding, & Tarrier, 2012a; Panagioti, Gooding, Taylor, & Tarrier, 2012; Pratt, Gooding, Johnson, Taylor, & Tarrier, 2010; Taylor, Gooding, Wood, & Tarrier, 2011; Taylor et al., 2010). The SAMS model specifies three key cognitive processes, namely information processing biases, a suicide schema and appraisals of the self, situation and coping (Tarrier et al, 2013). CBSP was developed to target these specific cognitions. In an initial evaluation of the CBSP intervention in a randomised controlled trial of 50 suicidal patients experiencing psychosis, relative to a control group receiving treatment as usual, the treatment group, who had received up to 24 sessions of CBSP, was shown to be significantly superior on measures of suicide probability, suicidal ideation and hopelessness (Tarrier et al, 2014).

To date, the development of CBT approaches for the prevention of suicide has not been extended into working with high-risk prisoners, despite the exaggerated rates of suicide in this high-risk group. Indeed, prisoners’ access to psychological interventions for mental health problems has generally been found to be largely absent (DH, 2009). This may seem surprising considering the prevalent use of cognitive behavioural programmes for the reduction of re-conviction and recidivism of offending behaviours. Such programmes have proven to be particularly successful in several evaluations and meta-analytical studies (McGuire, 1995; Lipsey, Landenberger, & Wilson, 2007; McDougall, Perry, Clarbour, Bowles, & Worthy, 2009) with programmes shown to be most effective when they have been well-designed, targeted, and systematically delivered (McGuire, 1995; 2002). Drawing upon this supportive evidence of the feasibility and acceptability of CBT for the prevention of criminal behaviour to prisoners, and also the preliminary support for CBT for suicidal behaviour (albeit outside of offender groups), there is reason to be optimistic that a cognitive-behavioural suicide prevention treatment could be feasibly delivered within the context of a prison setting and offer considerable clinical benefit to the prisoner patient.

With CBT holding the potential to be an effective treatment for the prevention of suicide behaviour amongst prisoners at risk of suicide, in this paper, we provide information on how a cognitive behavioural suicide prevention (CBSP) therapy was implemented with three male prisoners at risk of suicide. The three cases were selected from a randomised controlled trial designed to evaluate the feasibility of delivering CBSP to suicidal prisoners and to examine the impact of CBSP upon participants. Of specific interest in this paper is an examination of the pattern of changes in relation to suicidal thoughts and ideation, risk of future suicide behaviour and related psychological distress, such as hopelessness, associated with this intervention. Additionally, consideration will be made of how the CBSP therapy was modified to suit the demands and requirements of the custodial setting.

Methods

Participants

Participants were recruited from a male high-security prison in the England with capacity to house up to 1200 prisoners. All participants were identified under the Assessment, Care in Custody and Teamwork (ACCT; MoJ, 2013b) system to be at potential risk of suicidal behaviour within the past month, and aged 18 years and over.

Measures

The assessments used in this study were selected to focus upon the suicide behaviour continuum (attempts, plans, ideation) and established psychological correlates of suicide behaviour, i.e. hopelessness and depressive symptomatology. Hence, the primary outcome measure was the total number of episodes of suicidal behaviour within the past 6 months, which was recorded at the start of therapy and at a 6 month follow-up¹. Since, actual suicide attempts were anticipated to be too rare to be a reliable indicator, psychometric measures assessing suicidal ideation and risk were also administered.

The Beck Scale for Suicide Ideation (BSS; Beck, & Steer, 1991) is a 21-item self-report instrument that is widely used for assessing the intensity of the individuals’ specific attitudes, behaviours, and plans to complete suicide during the past week. Only the first 19 items are used within this study to assess current ideation, since the final two items record the number of previous suicide attempts and the seriousness of the intent to die associated with the last attempt. The BSS has demonstrated alpha reliability coefficients ranging from 0.84 to 0.93 in psychiatric samples (Beck, Brown, & Steer, 1997; Beck, Kovacs, & Weissman, 1979; Beck, Steer, & Ranieri, 1988). Palmer and Connelly (2005) reported a mean (SD) score on the BSS of 6.38 (9.20) for prisoners with a history of suicidal behaviour.

The Suicide Probability Scale (SPS; Cull & Gill, 1988) is a 36-item self-report measure designed to assist in evaluating future / potential suicide risk. A four-point likert scale is used to assess risk by exploring participants’ subjective experiences and past behaviours. The SPS has demonstrated acceptable test-retest reliability (r=0.92) and internal consistency (r=0.94) in clinical samples (Cull & Gill, 1988) and high levels of specificity and sensitivity amongst offender samples (Perry, Marandos, Coulton, & Johnson, 2010). Threshold scores for the SPS are as follows: no-low suicide risk (0-67), mild–moderate suicide risk (68–79) and high suicide risk (≥ 80) (Cull and Gill, 1988). Mean (SD) scores on the SPS for prisoners with a history of suicidal behaviour have been reported to be 46.97(20.94) (Naud & Daigle, 2010).

In addition to the direct measures of suicidality, two further assessments were administered to measure hopelessness and depression. Hopelessness was measured using the Beck Hopelessness Scale (BHS; Beck & Steer, 1993) which is a 20-item, self-report inventory for measuring negative expectancy of the immediate and long-term future, with higher scores indicative of a greater degree of hopelessness. Threshold scores for the BHS are: normal range (0–3), mild hopelessness (4-8), moderate (9-14) and severe (>14) (Beck and Steer, 1993). The BHS has previously been used with offender and general populations and has shown to be a reliable instrument (α=0.93) (Dunham, 1982) with mean (SD) scores ranging from 5.83(5.50) - 10.13 (4.81) for prisoners with a history of suicidal behaviour (Eidhin, Sheehy, O’Sullivan, & McLeavy, 2002; Palmer & Connelly, 2005).

Depressive symptom severity was measured using the revised version of the Beck Depression Inventory (BDI-II; Beck, Steer, & Brown, 1996), The BDI-II is a 21-item multiple-choice self-report inventory which participants rate how they were feeling for the past fortnight on a four point scale. The items relate to depressive symptoms, cognitions, and physical symptoms. Responses are summed to provide an overall score ranging from 0 to 63, with higher scores indicating greater severity. Threshold scores for the BDI-II are: no depression (0-9), mild depression (10-19), moderate depression (20-29) and severe depression (≥ 30). Mean (SD) scores on the BDI-II have reported to range from 21.66(10.03) – 27.42 (12.55) for prisoners with a history of suicidal behaviour (Eidhin, Sheehy, O’Sullivan, & McLeavy, 2002; Palmer & Connelly, 2005). The BDI-II has demonstrated high internal consistency (α=0.93) and test-retest reliability (0.93) (Beck et al., 1996).

Procedures

The host prison maintained a register of all prisoners currently identified under the ACCT system. Each individual on this register was provided with information about the study and invited to take part. With informed consent obtained, those that met the study criteria were invited to a briefing session where the study details were explained followed by an opportunity to ask questions. Measures were completed by participants at baseline (0 months), on completion of therapy (4 months) and at follow-up (6 months).

The study was conducted in accordance with research ethical approval provided by the Research Ethics Committee for Wales (REC Ref: 11/WA/0002), which specialises in research involving prisoners or prisons.

CBSP Intervention

Overview

The CBSP intervention was designed to restructure the specific psychological architecture responsible for suicidal behaviour. The delivery of CBSP was modified in line with the specific pragmatic and contextual restrictions and demands of a prison setting. The change methods within CBSP were based on established cognitive-behavioural techniques that were modified to specifically target the psychological mechanisms underlying suicide behaviour. Delivery of the intervention took place over a 4 month period with each participant being offered up to 20 individual sessions lasting 30 to 60 minutes, typically delivered on a once or twice a week basis. Therapy sessions had to be timed to fit in line with the demands of the prison regime, hence a 2-hour morning and 2½-hour afternoon window was made available for the therapist and participant. Participants were authorised by the prison to miss their usual work responsibilities in order to attend therapy sessions.

Initial sessions focused on engagement and assessment of the participants’ presenting problems, previous experiences of suicidal ideation and behaviour, and formulation of key areas for intervention. During intermediate sessions, the participant was supported in developing a set of helpful skills and strategies to improve coping and enhance resilience towards suicide behaviour. The final phase of the treatment was the development of a ‘maintaining well-being’ plan or therapy blueprint which served as a summary of work completed. Each of these phases will now be described in more detail.

Engagement and assessment

A comprehensive assessment was undertaken which aimed to identify the participant’s current level of risk in order to inform risk management, to gather information about factors which may contribute to the participant’s vulnerability to suicide, and to identify factors that may be harnessed to reduce vulnerability and improve resilience. In addition to the participant’s self-report, and with their informed consent, the therapist gathered information from those significant others who were also involved in the participant’s ongoing care and support, which tended to comprise prison and probation staff, healthcare professionals, and family members where contact had been maintained.

Drawing together the information gathered during the assessment phase, a personalised case formulation was then collaboratively developed. The formulation sought to incorporate relevant material from the participant’s previous life experiences, core beliefs about suicide and what suicide means to them, key appraisals of the individual’s situation, self-perceptions, and the subsequent emotional, behavioural and cognitive responses to suicidal behaviour.

Once the participant and therapist had discussed the formulation, and made refinements where necessary, the goals for therapy were then discussed. Identifying goals for change was a difficult challenge for some participants who were particularly pessimistic about their future and the likelihood of anything worthwhile coming from their involvement in the treatment (Britton, Williams, & Conner, 2008). In such circumstances, the therapist presented the treatment as a ‘no-lose’ situation, in which the participant simply had to be willing to try the therapy for a few weeks before making any firmer commitment to continue. Collaboratively agreed goals then informed the development of the treatment plan and the prioritising of the subsequent intervention modules.

Treatment

Following the assessment and conceptualisation of the participant’s suicidal experiences, a treatment plan was developed which targeted up to five key suicidal processes. Throughout treatment, a range of cognitive and behavioural techniques were introduced in accord with the target processes. The therapist found it helpful to initially model the use of each technique to ensure the participant had a sufficient understanding to allow them to practice the technique as a homework task.

The five modules within the CBSP treatment programme (Tarrier et al, 2013) were:

1. Attentional control training

2. Appraisal restructuring

3. Problem-solving training

4. Behavioural Activation

5. Schema focussed techniques

Attentional control

According to the SAMS model (Johnson et al, 2008), in times of a suicidal crisis, individuals tend to become ‘locked-in’ to a pattern of suicide ideation that they believe to be difficult to control. With the suicide schema activated, the individual’s attention becomes overly focused upon identifying potential threats in their environment, and even non-threatening stimuli becomes interpreted as dangerous. Hence, the first stage of CBSP is to reduce the valence of threat-focussed attention and other information processing biases that maintain the activation of the suicide schema. CBSP uses an attention training technique (Wells, 2009) to help reduce the participant’s excessive tendency to focus upon threat- and suicide-related stimuli (both internal and external).

Through the use of the attention training technique, participants learned how to overcome ‘attentional fixation’ by switching their attention onto non-self-relevant aspects of their environment, with the initial focus of the technique on sounds from various spatial locations. The technique was then extended to include attention to imagery, with initial practice focussed on visual objects within the participant’s immediate environment, such as a mug on a shelf, before practising focussing attention upon neutrally valenced internal images, e.g., waiting in the dinner queue. Through continued practice, participants were able to develop a broader sense of their experiences and the world around them, despite the presence of distressing thoughts, and developed a greater sense of control over whether or not to engage with the interfering cognitions.

Attentional control was then further strengthened using Broad Minded Affective Coping (BMAC; Tarrier, 2010; Panagioti, Gooding & Tarrier, 2012b; Johnson, Gooding, Wood, Fair, & Tarrier, 2013). In the BMAC technique, the participant was initially asked to think about a memory related to a positive event from their past, typically a family event, birthday party, wedding day. To begin the technique, the participant was asked to relax through a brief breathing exercise, and then begin to bring to mind an image of their preferred positive memory. Whilst maintaining the image in their attention, the participant was invited to fully engage with the memory across all their senses. So, the participant was prompted to move around the image and focus in on the visual details (e.g. objects, people, background) and then observe any sounds from various locations within the scene. The participant was then prompted to bring to life the recalled memory through any related sensations of touch, smell and/or taste. Through recalling sensory details and sustaining the memory in their attention, the participant was then instructed to recall positive feelings they experienced at the time of the memory, and then to bring these emotions back to life in the present-moment – to re-experience the positive affect in the here-and-now. Subsequent questioning by the therapist helped the participant to identify any positive meaning attached to this emotion and the implications for the participant’s sense of control over their attention and associated feelings.

Appraisal restructuring

According to the SAMS model, a number of key appraisals are maintained by the suicidal participant, which need to be identified and challenged through traditional cognitive methods (Beck, 1979; Wenzel, Brown, & Beck, 2009). Specific appraisals often concerned the participant’s current situation in prison, past events often related to offending and/or drug-related behaviours, and negative predictions for the future. Furthermore, participant distress also seemed to stem from a fragmented sense of self, low perceived personal agency and little confidence in their ability to effect positive changes in the future. Initially, psychoeducation work focused on improving the participant’s awareness of common thinking biases, e.g., selective abstraction / negative mental filter, magnification / catastrophising, and arbitrary inference / jumping to conclusions (Williams & Garland, 2002) before the participant began to monitor occurrences of any such unhelpful thinking styles. The accuracy and likelihood of the participant’s appraisals was then considered by encouraging them to evaluate the evidence for and against the thought. Where possible, behavioural experiments were used to encourage the participant to seek out and identify disconfirmatory evidence directly. Any new evidence was then considered to challenge the accuracy and usefulness of any unhelpful beliefs.

Problem-solving skills training

Since deficits in interpersonal problem-solving are associated with suicidality (Pollock & Williams, 1998), a structured technique to developing skills in improving solving such problems was considered. The CBSP intervention (Tarrier et al, 2013) proposed the following approach to problem-solving training:

1. List clearly the problem(s) to be resolved.

2. Select a problem and clearly and simply define it.

3. Brainstorm as many solutions as possible.

4. List the advantages and disadvantages of each possible solution.

5. Select and implement a solution.

6. Evaluate the effectiveness of the selected solution. If ineffective, select and implement an alternative solution, and repeat.

It was initially helpful for participants to learn the steps of this technique on a hypothetical, everyday problem, e.g., “You have a headache and would like paracetamol”. In this example, the therapist would use a Socratic questioning style to support the participant to arrange hypothetical access to non-prescription medication through the usual prison application procedure.

When confidence in the use of the technique was developed, the participant was then encouraged to apply the technique to their own previous problems related to suicidal distress. This approach thus helped participants to learn how the technique may have helped them to consider alternative responses to suicide behaviours. To extend the use of this technique further, the participant considered potential problems likely to affect them in the future and then worked through the above steps again to develop plans of how to resolve the problem should those problems actually occur. Self-reflections on the participant’s use of the problem-solving techniques were then used to inform the development of more positive appraisals of coping in future situations, such as “When I think problems through, I am able to find solutions” and “I am learning how to cope better when I’m facing a problem”

Behavioural activation

An effective behavioural technique when working with a depressed and/or suicidal participant is behavioural activation (Jacobson, Martell, & Dimidjian, 2001; Dimidjian, et al., 2006). The sense of inertia that a state of hopelessness, defeat or entrapment often left a participant with was challenged using regular self-monitoring of the participant’s activities of daily living (Beck & Greenberg, 1974). Activities found to be associated with an increased sense of pleasure and/or achievement were timetabled into the participant’s daily or weekly schedule. The resulting increase in time spent accessing pleasurable and/or achievement activities had the potential to elevate the participant’s engagement with their external environment, such as time spent with peers, family and friends, and ensure routine access to positive, alternative schema that would strengthen the participant’s resilience, sense of personal agency and control over their life (Tarrier et al, 2013; Wenzel, Brown, & Beck, 2009). This improved sense of meaning in the participant’s life and increased connection with their social network served to undermine the potential ‘no rescue’ appraisals, such as “I’m alone and no one will help me” (Johnson et al., 2008; Williams & Pollock, 2000). In prison, the participant’s access to pleasurable activities was considerably compromised, which required the therapist to support the participant to focus on realistically achievable activities, e.g. watching TV, listening to music, reading a book, visiting the gym, talking to a friend, and walking with peers in the exercise yard. Despite such limitations, the basic premise of the technique – engaging more in pleasurable and/or achievement activities serves to improve mood – remained a realistic goal for this phase of therapy.

Schema focussed work

The final component of a participant’s treatment plan was schema focussed work. The aim for this phase was to deactivate, inhibit and/or change the suicide-related schema through the adoption of new and appropriate schematic beliefs about the participant’s circumstances, self and future (Tarrier et al, 2013). Reduced activation of the suicide schema was intended to be achieved through the enhancement of more positive schemas with associated links with more adaptive problem solving responses. Methods were used which helped to promote positive self-worth, e.g. the participant was supported to draw up a list of ten positive qualities about themselves, each of these qualities were then rated as to how much the participant believes they were are actually true. The participant was then invited to recollect specific examples of when they demonstrated each quality, with detailed memories then used as examples for BMAC practices between sessions. This practice was intended to emphasise and bolster the participant’s memories of positive experiences. Re-ratings of the participant’s belief in each of the positive qualities were then used to emphasise to the participant that their beliefs could change depending upon what evidence they focused their attention on. Furthermore, the participant was supported to develop an implementable plan of preferred goals and ambitions for their future, especially for when released from prison and back with their family. This served to build hope for the future and was intended to provoke a reason for living despite current distress and hardship.

Maintaining well-being plans

Typically in the final 2-3 sessions, the participant developed a plan for how they intended to maintain any gains in their sense of well-being achieved during the course of therapy. The focus of the plan was to summarise key lessons learnt from the treatment, to identify new coping techniques and strategies now available to the participant, and additional sources of support within the participant’s community (peers, staff, family). Key high risk situations of potential future suicidal crises were identified, and the participant was encouraged to imagine how they would use any new skills learnt during therapy and/or external resources to overcome or, at least manage, the distress associated with the crisis.

Case Examples

The following three cases were selected from the therapy group within the trial to highlight our experiences of working with this clinical group. Each case study has been selected to represent the various experiences, benefits and challenges faced by the participants, to challenge the principles and techniques delineated and to provoke constructive discussion of the obstacles that had to be overcome by the participant and therapist when undertaking this work. Some identifying details of each case have been altered to preserve confidentiality.

Table 1. Summary of outcome measures.

	Arthur	Mark	John
Suicidal Behaviour in last 6 months
Baseline	1	0	1
Follow-up	0	0	0
BSS
Baseline	0	0	1
End Therapy	0	0	0
Follow-up	0	0	--
SPS
Baseline	64	90	82
End Therapy	38	60	41
Follow-up	47	53	--
BHS
Baseline	6	5	7
End Therapy	6	2	4
Follow-up	3	3	--
BDI
Baseline	20	30	38
End Therapy	6	9	5
Follow-up	6	7	--
Treatment Retention
Number of sessions attended	17	19	15
Number of CNAs	2	5	4
Number of DNAs	0	1	0

Key: BSS = Beck Scale for Suicidal Ideation. SPS = Suicide Probability Scale. BHS = Beck Hopelessness Scale. BDI = Beck Depression Inventory. CNA = Could Not Attend Session. DNA = Did Not Attend Session.

Arthur

Overview

Arthur was a 40 year-old white British male who had received a five year sentence for robbery. Having served more than three years of this sentence in prison, Arthur had previously been released on parole. However, a few months prior to the start of the treatment period, Arthur had been recalled back into prison for a second time due to another breach of his parole conditions. Arthur was expecting to serve the remaining nine months of his sentence in prison. Outside of prison, Arthur had a wife and a six year-old daughter who paid regular visits. Arthur attended therapy over a period of four months, during which time he chose to attend 17 sessions offered to him, and was unable to attend a further two sessions due to prison regime restrictions. Arthur did not receive any other mental healthcare input during the course of therapy.

Life history

Following the “disappearance” of his mother when he was aged five years, Arthur spent the remainder of his childhood in various care homes with occasional visits from his father. From the age of 10 years, along with his peers, Arthur began to engage in petty crimes, such as shoplifting and stealing from cars, to help him raise enough money to pay for a train ticket to visit his father. Arthur was 12 years old when his father died, after which point Arthur’s use of street drugs and alcohol significantly increased, alongside his involvement in criminal behaviour. By the age of 15, Arthur had received his first prison sentence for an acquisitive offence and, in the subsequent 25 years, he had been imprisoned on more than 20 further occasions. Arthur had spent more of his adult life inside prison than out.

During childhood, Arthur received significant physical and emotional abuse from the care home staff, including violent beatings and being placed in isolation for several days at a time. Arthur attributed this abuse to him being physically weaker than the staff and therefore unable to “hold them off”. Repeated episodes of abuse led him to develop the beliefs that “I am a weak and vulnerable person” and “people in authority are not to be trusted”. In assessment sessions, it became apparent that Arthur continued to harbour an extreme dislike of people who abused their position of power over vulnerable others. Arthur demonstrated this conviction by defending ‘weaker’ prisoners from others intending to exploit them even though such behaviour was likely to attract adjudications and additional punishments from the prison.

Arthur had engaged in two previous suicide attempts. His first attempt occurred eight years ago when Arthur was living in the community. At the time of the attempt, his wife was pregnant with twins and Arthur was heavily using drugs and alcohol and frequently engaging in criminal behaviour such as violence, burglaries and selling drugs. Tragically, both of the twins were born prematurely and died at birth. Arthur recognised his behaviour may have caused his wife significant stress during the pregnancy, which may have contributed to the twins’ stillbirths. This reflection led to self-critical thoughts that he was to blame for the twins’ deaths and that he was a terrible parent for doing so. The associated feelings of shame and sadness, labelled as “mental hurt” by Arthur, became increasingly unbearable. Arthur appraised this experience as a punishment for his previous actions, and that he deserved to suffer. In the midst of this episode, Arthur began to consider suicide as the solution to rid himself of the emotional pain he was experiencing and the “punishment I deserve”. He also began to consider suicide as an end to his worries about the disappointment he had caused his wife and the fear she would leave him. Arthur took an overdose of paracetamol and was later admitted to a mental health ward for a period of assessment and treatment.

Arthur’s second suicide attempt occurred three weeks prior to the start of therapy. Having recently been recalled into prison for the second time (due to a breach of his parole conditions), Arthur had again been removed from his family home. During the first week or so back in prison, Arthur began to reflect on the increasing amounts of time he was spending in prison and away from his family. During his evenings, Arthur began to ruminate on his continued absence from his wife and daughter and the “failure” he perceived himself to be in his roles as a husband and father. Arthur’s recall into prison also coincided with the anniversary of the twins’ deaths, which served as a reminder of his previous regrets and disappointments as a father. The ‘mental hurt’ returned to Arthur who swiftly fell into a cycle of depressive thinking and self-criticism. Based on a prediction that he would not be released from prison again, any time soon, Arthur felt increasingly hopeless and this triggered his existing belief that suicide was the only way out of intolerable situations. Arthur stockpiled paracetamol from other prisoners and attempted an overdose, which resulted in an admission to the prison healthcare centre.

Treatment process

From the initial sessions onwards, Arthur expressed considerable motivation to engage in the therapy. Since physically recovering from the recent suicide attempt, Arthur had became more aware of his protective beliefs against suicide, for instance, expressing his feeling that “suicide is a selfish way out” and that “my daughter needs her Dad around”. As such, Arthur’s therapy goals were to develop more helpful ways of coping with the ‘mental hurt’ without the use of drugs or alcohol.

Subsequent sessions were used to train Arthur in the attention training technique to improve his control over the focus of his attention. Arthur learned to deliberately move his attention to various sources of sounds, images and, eventually, thoughts. Arthur stated this technique was especially helpful when he was beginning to feel stressed or sad for missing his family. Instead of following the racing thoughts that he had previously experienced, Arthur chose a 12 minute attention training practice, which helped him to regain control of his focus and also helped him to physically relax. Due to the perceived helpfulness of the attention training, additional sessions were used to extend the practice to the Broad-Minded Affective Coping (BMAC) technique. Arthur developed several BMAC practices that enabled him to recall, in detail, images and related feelings associated with memories of holidays spent with his wife and daughter. Arthur found the BMAC practices helped him to bring back some of the pleasant feelings and give his current mood a ‘boost’. Complimenting the regular use of these BMAC practices, Arthur identified a small number of activities and tasks that he expected would bring him a lift in his mood (e.g., phone call to family, attending the education group). By monitoring his self-ratings of mood before and after engaging in such tasks, Arthur identified key ‘lifting activities’ which he then planned into his forthcoming week, thus improving his sense of personal agency over his mood. Finally, Arthur was supported to challenge some of the unhelpful beliefs he held about himself (“I am a weak and vulnerable person”, “I must hit out first, else I will be hit”) and his role as a husband and father (“I have failed my family”, “My family would be better off without me”). Completion of thought records and consideration of evidence for and against each belief served to help Arthur establish a more helpful and balanced conclusion associated with less shame and sadness. A ‘Maintaining Progress’ plan was developed in the final therapy sessions which summarised key learning points and offered a reminder of future sources of help, including Arthur’s personal officer, mental healthcare staff and the chaplaincy.

Treatment outcome

The formal assessments completed at baseline, end of therapy and follow-up (Table 1) show that Arthur experienced no suicidal behaviour during the six months follow-up. Arthur’s baseline score for suicidal ideation (BSS=0) was perhaps surprisingly low considering this assessment was completed just a few weeks following his most recent suicide attempt. BSS scores remained at the minimum level across all time-points. Noticeably, the measure of potential suicide risk in the future decreased from baseline (SPS=64) to the end of therapy (SPS=38), with the decrease maintained, although to a lesser extent, at follow-up (SPS=47). The end of therapy and follow-up SPS scores fell well within the no/low risk range (Cull & Gill, 1988). Arthur’s BDI scores fell from a moderate level of depression at baseline (BDI=20) into the non-clinical range by end of therapy and follow-up (BDI=6). Hopelessness scores also fell from baseline (BHS=6) by follow-up (BHS=3), which was consistent with Arthur’s own perception that the future was beginning to look more manageable and appealing to him.

Mark

Overview

Mark was a 22 year old white British male with a history of depression and anxiety problems and was currently prescribed anti-depressant medication from the prison doctor and was receiving regular contact with the Mental Health In-Reach Team in the prison. Mark had been imprisoned for seven months at the time of his referral to the study and was fearful of a potential seven year sentence due to the violent nature of his index offence. Mark’s court case was expected to be heard in nine months’ time. Mark attended therapy over a period of four months, during which time he attended 19 sessions, was unable to attend a further five sessions, due to legal visits, and chose not to attend one session.

Life history

Mark described his childhood as chaotic and unpredictable. He had a younger brother but no recollection of his father who had left the family home when Mark was two years old and his mother was pregnant with his younger brother. Mark’s mother was a long-term user of illicit drugs for most of his upbringing. Mark reported frequent episodes of his mother being “out of it”, leaving him to cook, clean and care for his brother. At the age of 14 years, Mark’s mother left the family home and Mark assumed the role of carer for his younger brother. Mark recalled feeling “abandoned” by his mother and then responsible for his brother’s well-being, which led him to develop a core belief of “I am unlovable” and a conditional assumption of “To be happy, I have to be with someone I love”. Following his mother’s departure, Mark began to engage in gang-related criminal behaviour to earn a living. Mark received several Anti-Social Behaviour Orders (ASBOs) before his first prison sentence at the age of 18 years for a robbery offence, serving two years in prison. Within 4 months of his release, Mark was remanded back into custody for a new offence of aggravated robbery.

At the time of referral into the study, Mark had been in a relationship with his girlfriend for approximately one year and she had recently had their first child. Mark received regular visits from his girlfriend but was becoming increasingly worried that she would leave him if he continued to spend time in prison and away from their home. During Mark’s previous prison sentence, an ex-girlfriend ended her relationship with Mark stating she wanted to be with someone else (outside of prison).

During the current prison term, Mark reported he had harmed himself on five previous occasions, each time by cutting his upper arms or chest. Initial sessions focused on the collaborative development of a formulation of Mark’s most recent suicide behaviour which occurred three weeks prior to the start of therapy. Mark reported he had received news from his girlfriend that his daughter was ill and had needed to be admitted to hospital. Mark had received this news in a brief phone call with his girlfriend that was ended abruptly as Mark had run out of credit on the prison telephone. Mark was left with a number of questions and worries about his daughter’s ill-health but with no access to any immediate answers since his phone account could not be credited again until the following week. Upon returning to his cell after the phone call, Mark described how he began to “lose control of my mind” with his anxious thoughts spiralling into greater catastrophes linked to the belief of “I am unlovable”. Within a couple of days, Mark had become convinced that his girlfriend would leave him for someone who could provide care and support to her and their daughter. With such an imagined future considered indubitable, Mark predicted he would be abandoned again, which would lead to terrible feelings of loneliness. Becoming increasingly hopeless activated beliefs that “there’s no point in doing anything to stop this” and “everything is ruined and it’s all my fault”. At this time, Mark withdrew from his daily activities and chose to spend more of his time in his cell away from others. Mark’s feelings of despair and defeat grew into a realisation that ending his life was the only solution to his situation. Mark secured a blade from another prisoner on the wing and cut himself on the upper chest during the night time. The next morning, Mark was assessed by the healthcare staff and considered to be not requiring further medical assistance since the injury was reported as “superficial”. Hearing this description, Mark appraised his suicide attempt as yet another example of his continuing failures (“I can’t even kill myself properly!”).

Treatment process

Mark considered his use of self-harm to be unhelpful in the long-term and served to maintain his self-critical beliefs and underlying low self-esteem. The goals for therapy were (i) to improve Mark’s response to problematic social/interpersonal situations and (ii) to enhance Mark’s ability to manage intense emotions and stress.

Mark expressed a willingness to engage in practically focussed techniques, and so a series of problem-solving training sessions began the intervention phase of the therapy. Mark was instructed in how to use a systematic approach to problem-solving. Initial examples of everyday problems were used to enable Mark to familiarise himself and rehearse the new skills (e.g. wanting to read the newspaper but out of credit). Later sessions drew upon more recent scenarios that had caused Mark to become distressed, such as being refused access to the gym. Through the completion of structured worksheets between sessions, Mark internalised the steps of the process. The application of this technique was then extended to hypothetical scenarios that were similar to Mark’s previous suicidal crises (e.g. girlfriend ‘dumps’ me over the phone). Future or potential high risk situations were also worked through to support the integration of the new skills into Mark’s repertoire of coping strategies. The next phase of therapy enabled Mark to challenge some of the catastrophic predictions and worries he often experienced when in a highly distressed state. Through the completion of thought diary records, both in session and as homework tasks, Mark developed a systematic approach to “thinking about my thinking” which encouraged him to access more rational alternatives to the worst case scenario often predicted. The final phase of the therapy work was focussed on improving Mark’s self-esteem. Enhanced access to more positive schema was achieved through supporting Mark to identify a range of positive qualities he had previously demonstrated to significant others, with an accessible memory associated with each quality. Mark identified positive traits as being “generous, caring and helpful to others” through recalling the difficult period when sole carer for his younger brother and occasionally having to go without food so his brother could enjoy a meal. BMAC practices allowed Mark to repeatedly access these memories and associated feelings which were followed by reflection with the therapist on what such feelings mean about Mark and the positive relationships he had with his family and friends. Through continued practice, Mark self-reported an increasing confidence in being able to manage his mood by using the range of BMAC practices when he noticed a downward change to his mood. A relapse prevention plan was developed in the final sessions.

Treatment outcome

The formal assessments completed at baseline, end of therapy and follow-up (Table 1) indicted Mark had not engaged in any suicidal behaviour in the six months before baseline or during the follow-up. Since this data was obtained from the official records of the host prison, it remains unclear why the self-reported episodes of self-harm were not recorded on the prison system, although the judgement of staff that the behaviour was “superficial” may offer some explanation. Similarly, Mark’s suicidal ideation scores remained at the minimum throughout the study (BSS=0). The probability of potential suicide decreased from the high suicide risk range at baseline (SPS=90) to low risk by the end of therapy (SPS=60) and follow-up (SPS=53). Similarly, scores for depression fell from the severe range at baseline (BDI=30) to the non-clinical range by end of therapy (BDI=9) and follow-up (BDI=7). Scores for hopelessness also reduced during the course of therapy, from BHS=5 at baseline to BHS=3 at follow-up.

John

Overview

John was a 54-year old white British male who had been remanded into custody charged with a series of historical sexual offences. At the time of his referral, John had been detained in custody for nine months awaiting trial. John attended therapy over a period of four months, during which time he attended 19 sessions, with two further sessions offered but John chose not to attend.

Life history

John was one of nine children who perceived his mother to be loving and caring but a father whom John described as “distant” and “angry”. John later found out that his father had been a heavy drinker for most of his upbringing. Throughout his childhood, John was physically and verbally abused by his father, who often returned home intoxicated. John was also a frequent victim of bullying at school with very few friends. As an eight year old, John recalled an experience of being “molested” by a family friend during a visit to the local cinema with his father. John attempted to tell his parents of this incident, but felt ignored by them. John developed a core belief of himself as “vulnerable and weak” and of others as “dangerous and powerful”. At the age of 14 years, John began to cut himself on his arms and legs as he found that cutting helped provide relief from his intense and overwhelming distress since the physical pain served as a distraction from the emotional pain.

Aged 21 years, John married his wife and started a family, whilst working at the local factory as the workshop foreman. John recalled this period of his life as being “hard work but happy”. After 20 years of marriage, John’s wife left the family home to live with a new partner, taking the two teenage children with her. This had a devastating effect on John who fell into a deep spiral of self-criticism and blame. John felt he was being punished for being an inadequate husband. In response, John would drink in local pubs and ‘pick fights’ with fellow drinkers hoping he would receive a beating as his punishment. During the year following the divorce, John lost his long-term job at the factory due to his declining attendance, lost most of his friends who no longer wanted to drink with him, and spent more and more of his time isolating himself from others. He then made his first suicide attempt by overdosing on paracetamol resulting in hospital treatment and a two week admission to the local mental health ward.

In the subsequent five years, John continued to struggle to make friends preferring to keep to himself and away from busy social situations, only leaving the house if it felt absolutely necessary. John became aware of on-line communities where he could ‘socialise’ with like-minded individuals. He was introduced to pornographic material on the internet and then became involved in the distribution of explicit images. John was aware of his wrongdoing and increasingly distressed by the feelings of guilt and shame associated with this behaviour. John’s extreme ambivalence about continuing to engage in this offending behaviour caused him emotional turmoil as he recognised the pleasure of engaging in the activity and the sense of belonging to a community, and yet the strong disgust of himself often triggered by thoughts about his children’s appraisals of his current actions. Feeling trapped in a situation with no escape, John made his second suicide attempt. Following his recovery, John recognised his need for help and came to the decision to make a confession at the local police station. He subsequently received a three year prison sentence.

Having served 18 months in prison, John was released with strict parole conditions and a subject of the Sex Offender Register. John joined his local church who offered him the sense of community and belonging he had desperately longed for since his marriage. John felt accepted by this community and began to feel valued again. Two years later, John was charged with historical sexual offences and remanded into custody. Upon reception into prison, John felt as though his life had been wrenched back into the depressed pit of disgust that he had felt trapped in around the time of his first prison sentence. Once again, John’s thinking became dominated by self-criticisms of “I’ve done terrible things”, “I am a monster” and “I deserve to be punished”. This rumination triggered John’s suicide schema leading to assumptions such as “If I’m not here, then I don’t have to deal with these problems” which emphasised suicide as a preferred way out. In the six months prior to his referral, John engaged in a ‘medically serious’ episode of suicide behaviour where he had severely cut his throat. An admission to prison healthcare and medical treatment was required.

Treatment process

John presented to the initial assessment sessions as motivated to learn more about why he was experiencing such difficult feelings and expressed a firm willingness to develop greater independence in his own ability to cope with his feelings away from his current use of self-harm. Assessment sessions allowed John to tell his story, which seemed to offer catharsis to him since few significant others in his life were able or willing to listen to John’s narrative. During the assessment and formulation sessions, John began to develop a sense of hope for his future to be better than he had previously predicted, although he continued to harbour considerable doubts about the likelihood of him ever achieving a worthwhile life again. Before addressing these hopeless appraisals of his future, it felt important to enable John to achieve a ‘quick win’ in therapy and thus bolster the therapeutic engagement in the work ahead.

The attention training technique was prioritised as an intervention, since John expressed an interest in gaining control over his ruminative thinking when alone in his cell at night. John recognised that the rumination preceded emotional difficulties and by positively responding to the rumination, subsequent distress could be minimised. John engaged well in the attention training practices which he saw as akin to “attentional muscle strengthening” with each brief practice seen as a “mind gym workout”. Continued practice soon resulted in John’s ability to use the BMAC technique as a new response to counter rumination. By choosing to focus on positive memories of spending time with his children and siblings, John was able to re-experience associated feelings that his current context may not have provoked. This technique also provided John with some preliminary evidence against his appraisal that he needed others to help him cope with his emotions. Subsequent work then focussed on monitoring John’s regular activities, despite the limited range available to prisoners. This enabled John to appreciate the impact of his behaviour upon his mood. A schedule of pleasurable or rewarding tasks was established for John to draw upon when experiencing a drop in mood. Through further reflection upon this work, John appreciated yet further evidence that he was becoming increasingly able to manage his own feelings. The final intervention work attempted to address John’s hopelessness about his future and the low level of self-esteem that he had experienced since returning to prison. With support, John was able to list several positive qualities he considered himself to possess along with a few associated examples of himself demonstrating each quality. A series of BMAC practices were then derived with each practice tailored to focus on previous examples of John demonstrating a positive quality. John recalled a happy memory of playing darts with two of his younger brothers, where he allowed them to beat him, which was used as an example of love, kindness and caring for others. This memory was repeatedly held in John’s attention as he improved his present-moment access to the associated positive feelings. Through repeated BMAC practices to the positive memories, John began to develop a more balanced appreciation of his strengths and weaknesses. Finally, a ‘maintaining progress’ plan was developed which documented key learning points and technique/practice reminders for John to refer back to following the completion of therapy.

Treatment outcome

The formal assessments were completed at baseline and end of therapy (Table 1). Follow-up data were unavailable since John was unexpectedly transferred out of the host prison a week before the follow-up assessment was due to take place. Since his suicide attempt shortly prior to the referral to the study, John had not repeated any suicide behaviour in the six months during the follow-up. John’s suicidal ideation scores hardly changed throughout the study (baseline BSS=1 to end of therapy BSS = 0). The probability of potential suicide decreased by half from the high suicide risk range at baseline (SPS=82) to the low risk range by end of therapy (SPS=41). BDI scores fell from a severe level of depression at baseline (BDI=38) into the non-clinical range by end of therapy (BDI=5). Hopelessness scores also fell from baseline (BHS=7) by the end of therapy (BHS=4)

Discussion

This paper offers an introduction to the application of Cognitive Behavioural Suicide Prevention (CBSP; Tarrier et al, 2013) to individuals identified to be at risk of suicide whilst detained in prison. The primary aim of this paper was to demonstrate how CBSP can be modified to fit within the prison environment and to meet the needs of suicidal prisoners. Although this is a small and preliminary study, the findings demonstrate reasons to feel confident that CBSP may offer achievable benefit to prisoners experiencing suicidality.

In terms of outcome, no participants had engaged in suicidal behaviour in the 6 months prior to follow-up, compared to a single episode each for Arthur and John in the 6 months prior to baseline. The rare occurrence of recorded suicidal behaviour throughout the course of therapy for each client demonstrated the importance of complementing this observable outcome with a range of assessments from across the suicide continuum. The administration of clinical measures at pre-, post- and follow-up assessments allows for a consideration of the severity of suicidality and distress experienced by the three case examples compared to the broader prisoner population context. Specifically, the baseline SPS scores for the three cases (Arthur=64, Mark=90, John=82) were all notably above previously reported mean scores for prisoner samples with a history of suicidal behaviour (mean[SD]=46.97[20.94]) (Naud & Diagle, 2010). Also, the baseline BDI-II scores for the two of the three case examples (Mark=30, John=38) were somewhat higher than the mean scores reported previously (mean[SD]=21.66[10.03]-27.42[12.55]) (Eidhin, Sheehy, O’Sullivan, & McLeavy, 2002; Palmer & Connelly, 2005).

From the questionnaire scores, it can be seen that considerable improvement was achieved by each case. Potential suicide risk, as measured by the SPS, was severe for two cases at baseline, and in each case, there were substantial reductions by the end of therapy. Although Arthur’s baseline SPS score was low at baseline, this score reduced further by the end of therapy and remained in the no risk range during follow-up. A similar pattern was shown for depressive symptoms, as measured by the BDI. At baseline, Mark and John both received a BDI score indicating severe depression, which was greatly reduced by the end of therapy, and maintained at follow-up for Mark (John was not available for follow-up assessment). Arthur’s baseline BDI score indicated moderate depression, which also reduced into the non-clinical range by the end of therapy and during follow-up. Scores for hopelessness, measured using the BHS, also demonstrated a reduction across participants with Arthur, Mark and John all achieving scores within the mild range at baseline and falling into the nil range after receiving therapy.

The completion of questionnaire measures of suicidality and related constructs also helped participants to reflect upon changes that had occurred during therapy. For instance, Arthur wanted to be sure that he had become more able to cope with the ‘mental hurt’ that had previously lead to suicidal thoughts. Whilst Arthur’s subjective report was that he felt more confident within himself by the end of therapy, it was the pre-post difference on the questionnaire scores that provided Arthur with the more objective proof that he needed.

In addition to the questionnaires completed at pre-, post- and follow up assessments, measures could also have been administered on a sessional basis with the cases. Regular monitoring of key symptoms, thoughts and experiences may have allowed for a more detailed understanding of the impact of the intervention techniques throughout the treatment process and also enabled the therapist to ‘track trajectories of change’. Fuller indicators of the process of the intervention are required to develop a more complete understanding of the mediators and moderators of change (Emsley, Dunn & White, 2010; Kazdin, 2007).

Although a number of modifications were made to the CBSP treatment, these changes did not have a significant impact upon the form and delivery of the intervention that had previously been implemented with people experiencing suicidality outside of a prison setting (Tarrier et al, 2014). For all cases in the current study, assessments were informed by the underlying cognitive behavioural model (SAMS; Johnson et al, 2008) and the interventions used were drawn from traditional techniques that have been evaluated for use with a non-forensic and/or non-suicidal population. Nevertheless, the challenge of engaging suicidal prisoners into a psychological treatment and motivating them to maintain their attendance throughout delivery of the treatment can be intimidating. The adaptations to CBSP for the current study related to the restrictions placed upon the delivery of therapy by the prison regime and environment, and to the heightened need for the therapist to emphasise engagement and develop trust with the participant. No changes were required to the theoretical underpinnings of the intervention.

The target problems for each of the participants were often a perceived inability to tolerate distress, hopeless or pessimistic predictions about the future, together with considerable self-critical thinking. An idiosyncratic formulation of the specific complexities and difficulties of each case was necessary to help the therapist and participant enter a collaborative process for developing the most acceptable and appropriate treatment plan. Initial engagement in CBSP was offered to each case as a ‘no-lose’ opportunity, with minimal commitment required from participants during initial sessions. The therapist adopted a non-directive approach during the first few sessions, which coupled with information provision about the structure of therapy, contents of the therapy programme, and role expectations of the therapist and participant, enabled each case to become familiar with what a course of therapy may offer to them and also presented a series of opportunities for the participant to gradually increase their level of comfortable self-disclosure. Despite the understandable reluctance from participants during the initial sessions, all three cases went on to experience high levels of therapeutic engagement attending almost all of the 20 sessions available and with minimal numbers of session refusals. Nevertheless, a number of cancelled sessions occurred for each participant. Due to the demands of an ongoing court case, the unexpected timing of legal visits prevented Mark from attending five therapy sessions as he had to be “held back” by prison staff. Although Mark recognised the importance of his attendance at such legal visits, the unknown timings of these visits presented considerable inconvenience to his continued commitment to the therapy process.

Patients with recent experiences of suicidal ideation or behaviour can find it very difficult to engage in a conversation about these experiences, and, thus, prove to be difficult to engage in treatment with recorded levels of drop outs and refusals of up to 60% (Motto, 1989; Lizardi & Stanley, 2010; Rudd, 2006). Arthur, Mark and John all expressed an initial reluctance to talk about personal difficulties and previous struggles, with a commonly held view that seeking help from others was a sign of weakness which was out of keeping with their gender. A strongly held belief shared by all three men was that it was not masculine to ask for help and that they should have been able to cope on their own like “real men” (O’Brien, Hunt, & Hart, 2005; Ridge, Emslie, & White, 2011). Discussing this point-of-view, in terms of the pros and cons for the individual male struggling to cope, served to generate a more balanced and rational consideration of help-seeking The male gender of the therapist (DP) may also have helped to ‘allow’ this conversation. To foster engagement, all participants were grateful for the opportunity to discuss the impact of suicidal behaviour upon their lives to date, how likely they felt they were to attempt suicide in the future and how confident they felt in being able to cope with feelings that may lead them to consider suicide as a solution. This preliminary assessment of the client’s reasons to undertake a course of therapy served to motivate them and identify a clear set of goals for the treatment. At the same time, it was important to recognise the client’s potential ambivalence about living and dying, especially if there seemed to be little hope for a meaningful life. This ambivalence was most apparent with John who felt that a life without his family and having to adhere to the requirements of the Sex Offenders Register would offer him little, if any, value. A brief conversation at the outset of therapy attempted to resolve some of this ambivalence by identifying the client’s reasons to live, although further sessions that allow for a greater emphasis on motivational enhancement prior to the commencement of the more constructive intervention techniques should be considered in future treatments (Britton, Patrick, Wenzel, & Williams, 2011). Initial treatment sessions could also place a focus upon the development of a safety plan (e.g. Brown & Stanley, 2012) to help identify the client’s early warning signs of a suicidal crisis, internal coping strategies, and potential sources of support. Prisoner patient safety plans would have to be sensitive to the environmental demands and constraints, with the therapist also mindful of the client’s motivations and potential reluctance to engage in treatment.

All 3 cases were typical prisoners in that they came from socially excluded and hard-to-reach groups within society, with minimal usage of statutory services (Social Exclusion Unit, 2007). Perhaps it is necessary and adaptive for such groups to be distrustful of others, especially if the ‘other’ is perceived to be coming from a position of authority. As such the therapist prioritised the development of a sufficient level of trust, as has previously been highlighted when commencing psychotherapeutic work with prisoners (Rappaport, 1971; Morgan, Winterowd, & Ferrell, 1999). Alongside the development of trust at the interpersonal level, between the participant and the therapist, there was also a need for the therapist to encourage the participant to develop trust at an organisational level. For example, Arthur disclosed a history of feeling exploited by persons in positions of authority and had subsequently developed a strong reluctance towards any help from “official services”. Whilst the therapist could not avoid being in a position of authority and employed by a statutory service, an emphasis was held on the development and maintenance of trust between Arthur and the therapist as an individual, more so than what the therapist represented to Arthur. It swiftly became apparent that all three cases had no previous experience of psychological therapy and the privileged position of being able (and encouraged) to talk openly about current problems with a warm, empathic listener adopting a non-judgemental approach. As is the case for all good psychotherapy (Miller, Duncan, & Hubble, 2008), the importance of therapist-client alliance should not be underestimated (Davidson & Scott, 2009; Messer & Wampold, 2002), especially in such a socially disenfranchised group.

Another key issue to address was the ethical dilemma associated with offering confidentiality to prisoner patients (Brodsky, 1980; Morgan, Winterowd, & Ferrell, 1999). It was critical for the therapist to describe and discuss confidentiality, and its limits, during the initial session with each of the participants. None of the three participants had previously engaged in mental healthcare services, and so it was important to ensure that each participant had a firm understanding of the confidentiality offered to them, to undermine any potential belief that the therapist will be reporting the participants’ disclosures to the prison staff or to solicitors, and to resolve any ambiguities about the limits of confidentiality related to risks of harm to themselves, others or the security of the prison (e.g. escape plans, contraband, etc).

A feature of CBSP that all three participants reported to be helpful was the predictable nature of the intervention, with each session commencing with the agreement of an agenda listing priority topics for the session. Also, the intervention package was presented as a series of ‘treatment modules’ and so the client could monitor their progress through the therapy. Upon the completion of each module, the therapist and client would reflect on key learning points from the completed module before considering which module should be prioritised for the next session, bearing in mind the client’s therapy goals. By offering such explicit structure within and across sessions, participants were able to recognise progress achieved whilst also being familiar with what they would be involved in next.

Participants appeared to gain the most benefit from the specific CBSP techniques when the practical skills were made directly relevant to participants’ day-to-day lives. For instance, John found his use of the BMAC technique to be more helpful when his practices were scheduled into those times in his day and week where we would expect to feel more isolated and defeated, such as following contact with his family and during the day at weekends when he was able to spend little time out of his cell. Similarly, Mark spent several sessions practicing his problem-solving skills for likely problems in his future, which helped him to feel more confident in coping with such scenarios if and when they actually occurred.

A challenge experienced by the clinician responsible for the delivery of therapy was in relation to the apparently limited time available within a 4-month therapy window. It was difficult to schedule in all 20 sessions to participants during the allotted 4 months. The main restriction being due to the limited time ‘out of cell’ that was made available for prisoners to attend therapy sessions; typically 2 hours in the morning and 2½ hours in the afternoon. Of course, the therapist would endeavour to swiftly rearrange any sessions that had to be cancelled due to the demands of the prison regime (e.g. lockdowns, staff shortages, limited access to therapy spaces), although this proved difficult to achieve within the restricted availabilities of the prisoners. For instance, Mark had 5 sessions cancelled by the prison staff since he had been called up to a visit from his solicitor. As such, the limited time available to deliver the intervention lead to the experience of a time pressure upon the therapist (and the client) to ‘fit in’ as much of the intervention within this time frame as possible. The extension of the 4 month therapy window was considered, although the contrasting rationale for keeping to the original time frame was to ensure all participants would remain within the host prison throughout the duration of their treatment. A longer therapy window may have increased the likelihood of an unexpected transfer out of the host prison and thus prevent a participant from having the opportunity to complete the course of therapy. It is recognised that the host prison clearly has a responsibility to ensure the safety and well-being of its prisoners and staff, and prisoner movements between establishments offer a useful tactic to achieve this aim. Future investigations of the delivery of CBSP in prisons may choose to revisit this decision and consider extending the therapy window to 6 months.

The current study represents an investigation of the feasible delivery and acceptability of CBSP to individuals at risk of suicide within a prison setting, with preliminary results suggesting this approach may offer benefit. Clearly, this case series of three participants cannot directly assess the effectiveness of the CBSP approach, although the results do suggest that this intervention may offer a preventative effect. The results achieved by participants in the current study appear in line with those previously reported in larger scale trials of CBT interventions for individuals with a recent history of suicide attempts (Brown et al, 2005; Slee et al, 2008). The willingness and commitment to maintain high levels of engagement and attendance throughout therapy, despite the considerable and challenging demands of the prison environment, demonstrate the recipients’ demand for this new intervention approach. A formal evaluation of the efficacy of CBSP in a controlled study is warranted and if further research confirms the value of the approach, then an argument would be presented for CBSP to become an important addition to the treatment as usual currently available to those identified to be at risk of suicide whilst detained in prison.