Social problem-solving interventions in medium secure settings for women

Abstract

Problem-solving interventions are a feature of overall medium secure treatment programmes. However, despite the relevance of such treatment to personality disorder there are few descriptions of such interventions for women. Beneficial effects for women who completed social problem-solving group treatment were evident on a number of psychometric assessments. A treatment non-completion rate of one-third raises questions of both acceptability and timing of cognitive behavioural interventions.

Introduction

The need to provide gender-specific and gender-sensitive treatment for women in secure settings has been repeatedly affirmed.^1,2 Women in medium secure settings are more likely than men to have a primary diagnosis of personality disorder (often borderline type), to be more likely to engage in self-harming behaviour and to have an index offence of arson.³ A study of the first 70 admissions to a women's medium secure unit⁴ found that the majority had a primary diagnosis of personality disorder together with a secondary diagnosis of substance misuse or mental illness. There is evidence that such dual diagnosis persons have problem-solving deficits;⁵ and that such deficits are linked to hostility and aggression,⁶ substance misuse,⁷ personality disorder⁸ and some criminal behaviour.⁹

Nezu et al. ¹⁰ define social problem-solving as ‘the self directed cognitive-affective behavioural process by which an individual attempts to identify or discover solutions to specific problems encountered in everyday living’ (p. 11). Treatment approaches based on the work of Nezu et al. ¹⁰ include the ‘Stop and Think’ social problem-solving intervention¹¹ which has been shown to affect positive change in young prisoners^13–16 and personality disordered offenders.⁹ Social problem-solving therapy has been effectively applied to problems of anxiety, depression, schizophrenia, suicidality, substance abuse and anger,¹⁰ and to patients who deliberately self-harm.^11,12 A meta-analysis of 58 studies of the effects of cognitive behaviour therapy on offenders confirmed that the largest reduction in recidivism was associated with programmes that included interpersonal problem-solving and anger control.¹³

A recent survey of treatment provision in medium secure facilities^15,16 found that problem-solving (e.g. reasoning and rehabilitation, social problem-solving and enhanced thinking) was a feature of several; however, there are few reports in the literature about such treatments for women. Long et al. ¹⁷ describe a ‘best practice service for women in a medium secure setting that employs a manualized gender-specific group treatment programme for dual diagnosis patients with personality disorder and mental illness. One of the core group treatments is social problem-solving adapted from the five-step problem-solving training manual¹⁸ and from the ‘Stop and Think’ problem-solving therapy for personality disordered offenders'.¹⁹ The first aim of this paper is to examine the effectiveness of the social problem-solving group interventions. The second aim is to examine the differences between completers and non-completers of group treatment and the clinical implications of these.

Method

A total of 24 women diagnosed as dual diagnosis with stable psychiatric symptoms agreed to participate in group treatment – their mean age was 32.6 years (SD 9.22 years, range 19–49 years) and their mean IQ of 85.5 (SD 16.99) was in the low average range. The primary diagnosis was personality disorder (emotionally unstable or with mixed features) (n = 18 [75%]), schizophrenia, schizoaffective and delusional disorder (n = 5 [20.8%]) and bipolar disorder (n = 1 [4.16%]). All of the women were detained under Section 3 of the Mental Health Act 1983 and had a history of disturbed behaviour, particularly self-harm, and emotional dysregulation. The majority of participants had been admitted from other medium secure units (n = 7, 29.1%), from prison (n = 7, 29.1%) and from maximum secure units (n = 6, 25%); index offences ranged from major violence (n = 10, 41.6%) to arson (n = 6, 25%) and assault, grievous bodily harm (GBH) and actual bodily harm (ABH) (n = 5, 20.8%). Three patients had no index offence.

Setting

Two medium secure wards within the Women's Service of an independent, registered charitable trust hospital (St Andrew's Healthcare, Northampton, UK).

Measures

Pre- and postgroup measures were:

Social Problem-Solving Inventory – Revised (SPSI-R:S).²⁰ This 25 item self-report questionnaire has five subscales: Positive Problem Orientation (PPO); Negative Problem Orientation (NPO); Rational Problem-Solving (RPS); Impulsivity/Carelessness Style (ICS) and Avoidance Scale (AS). Subscale scores combine to give a total Social Problem-Solving Score. Constructive problem-solving is indicated by higher scores on PPO, RPS and SPS; dysfunctional problem-solving is indicated by higher scores on NPO, ICS and AS. Information on reliability and validity is documented by D'Zurilla et al. ²⁰ The SPSI-R has high internal consistency (ranging in four normative samples from 0.85 to 0.96) and test–retest reliability is high (ranging from 0.68 to 0.91 across the 5 subscales). In terms of concurrent validity the SPSI-R correlates significantly with another problem-solving measure: Problem-Solving Inventory.²¹ The measure has been widely used in assessing the treatment outcomes in personality-disordered offenders.^11,12

Beck Depression Inventory (BDI-II).²² The BDI-II is an indicator of the presence and degree of depressive symptoms. Scores can be classified as minimal (0.13); mild (14–19); moderate (20–28); and severe (29–63). The scale has good test–retest reliability and construct validity. It has become one of the most widely accepted instruments for assessing the severity of depression in diagnosed patients.²² Information about the psychometric characteristics of the BDI-II has been collected on over 500 psychiatric patients.

Generalized Self-Efficacy Scale (GSES).²³ The GSES is a 10-item self-administered scale that assesses the strengths of an individual's belief in his or her own ability to respond to difficult situations and to deal with any associated obstacles or setbacks. Test and re-test reliability for women was found to be 0.63 and in terms of predictive validity the measure correlated positively with measures of self-esteem (0.40) and optimism (0.56).²⁴ The GSES has been used widely in research on women in secure settings (e.g.²⁵).

Camberwell Assessment of Need Forensic Version²⁶ (CANFOR) Item 14: company. CANFOR assesses 25 areas of need and includes questions on offending and on patient agreement with prescribed treatment. The validity and reliability of CANFOR were investigated in medium and high security psychiatric hospitals.²⁶

Health of the Nation Outcome Scale – Secure (HoNOS – Secure²⁷): Item 9: problems with relationships. The Health of the Nation Outcome Scales for users of secure psychiatric services, prison and forensic community services (HoNOS – Secure) is a global indicator of outcome. A score of one or more on any one of the HoNOS – Security scales A–G indicates that a risk management intervention is needed, while a score of two or more on scales 1 to 12 indicates that a care or treatment intervention is needed. HoNOS – Secure is a reliable outcome measure when used in routine clinical practice – test–retest reliability was found to be 0.65²⁸ and information on the validity of a range of HoNOS family tools is given in Pirkis et al. ²⁹ HoNOS – Secure is sensitive to change in women treated in secure psychiatric settings.³⁰

Intervention

The social problem-solving group employs a closed group format and consists of ten 90-minute group sessions. It is preceded by an individual orientation session and followed by a postgroup individual session. Time-limited (30 minutes) individual sessions occur on an ‘as needed’ basis between group meetings to help place group skills in a personal context and to address homework issues. Sessions were delivered weekly by an accredited cognitive behavioural therapist and a co-facilitator.

The individual orientation session involves meeting with potential group participants to discuss aspects of the group programme. The group programme along with its purpose and aims is familiar as regular review and treatment planning takes place from the point of admission for each woman admitted to the service. The meeting covers practical considerations related to times, dates and names of group facilitators. It also includes detailed discussion of the content of the programmes, and the potential benefits of attendance as related to needs assessment. It enables the potential participant to ask any immediate questions and raise any concerns about attendance. Pre-group assessments are usually also completed at this meeting.

Sessions covered an introduction to problem-solving, styles of problem-solving and their effectiveness and the ADAPT (Attitude; Define; Alternatives; Predict and Try) problem-solving steps¹⁸ with subsequent sessions devoted to further practice and review and application. The content of group sessions are:

Session 1 – Introduction to the programme

Session 2 – Introduction to social problem-solving

Session 3 – Styles of problem-solving and their effectiveness

Session 4 – Step 1 – Attitude (choosing the most helpful attitude for solving problems)

Session 5 – Step 2 – Define (the problem)

Session 6 – Step 3 – Alternatives (brainstorming possible solutions)

Session 7 – Step 4 – Predict (evaluating and choosing solutions)

Session 8 – Step 5 – Try out (implementing solutions)

Session 9 – Further practice

Session 10 – Review and application

The post-group individual session aims to (1) capture the learning and behaviour changes that have taken place over the duration of the group programme and (2) identify how the group participants can continue to benefit from the learning and extend this to other areas of life. An action plan is completed.

Concurrent with attendance at population-specific manualized cognitive behavioural therapy (CBT) groups¹⁷ patients attend a variety of sessions led by psychology, occupational therapy and nursing staff. These are needs-led and include self-care, daily living skills, education, physical health and exercise, and recreational skills.

Ethical approval

The current study formed part of a wider evaluation of St Andrew's Healthcare Women's Service with approval from a National Health Service research ethics committee (LNRI 06/02501/91).

Statistical analysis

Preliminary analyses were conducted to check the data for any outliers or errors and found no violation of normality, linearity and homoscedasticity. The frequencies of each variable were checked for minimums and maximums. No errors or outliers were found. Comparisons between treatment completers and non-completers were analysed using independent t-tests. Pre- and postintervention treatment scores were analysed using paired t-tests. Six separate analyses were conducted using the SPSI-R:S thus increasing the probability of a Type I error. A Bonferroni correction was therefore used resulting in an operational alpha level of 0.008 (0.5/6). Significance levels were set at 0.05 for other tests. Effect sizes were calculated for each paired t-test comparison using eta squared. Following Cohen³¹ an eta squared of 0.01 is a small effect, eta squared of 0.06 is a moderate effect and eta squared of 0.14 is a large effect. For each individual sample and test the effect size was calculated using a common index, the d statistic,³² where d of 0.20 is a small effect, d of 0.50 a moderate effect and d of 0.80 or greater a large effect.

Results

Patients who had attended more than 70% of sessions (i.e. 8 or more group sessions) were classified as treatment completers: 15 (62.5%) women were completers. Nine were the non-completers. Comparisons of the completers and non-completers revealed that there were no differences between the two groups in terms of age, marital status, index offence, diagnosis, IQ, source of admission or in terms of their prescores on psychometrics (SPSI-R:S; BDI; GSES; CANFOR item 14; HoNOS – Secure). All patients were on medication throughout the study but there was no difference in prn use between completers and non-completers.

The analysis revealed several changes pre- and postintervention for completers (see Table 1).

Table 1

Pre- and post-group psychometric assessment for treatment completers (n = 15)

	Pre-measure		Post-measure
Measure	Mean	SD	Mean	SD	Paired samples t-test	Number who improved (n, %)	Effect size (eta squared)
SPSI-R:S
Total SPS	10.26	(4.36)	13.33	(3.19)	t(14) = 3.76, P < 0.08)	10 (66%)	0.21^†
Positive problem orientation	8.92	4.75	10.25	4.66	NS	4 (26%)
Negative problem orientation	10.42	(5.85)	6.75	4.41	t(14) = 2.79, P < 0.08	9 (60%)	0.16^†
Rational problem-solving	8.35	6.28	11.78	4.66	t(14) = 2.24, P < 0.08	10 (66%)	0.13
Impulsivity/carelessness	7.39	5.18	4.26	2.97	t(14) = 2.81, P < 0.01	11 (73%)	0.16^†
Avoidance style	8.76	5.48	5.58	4.88	t(14) = 2.62, P < 0.01	12 (80%)	0.15^†
Generalized self-efficacy	20.23	2.94	36.78	7.04	t(14) = 3.23, P < 0.05	10 (66%)	0.18^†
BDI	16.06	7.89	6.23	4.34	t(14) = 2.89, P < 0.05	12 (80%)	0.17^†
CANFOR (Item 4: company)	1.50	1	0.58	0.98	t(14) = 3.23, P < 0.05	8 (53%)	0.18^†
HoNOS – Secure (Item 9: problems with relationships)	2.18	0.96	1.87	0.83	t(14) = 2.06, P < 0.05	9 (60%)	0.12

SPSI-R:S, Social Problem Solving Inventory – Revised; BDI, Beck Depression Inventory; CANFOR, Camberwell Assessment of Need Forensic Version; HoNOS, Health of the Nation Outcome Scale

^†Large effect size

The SPSI-R:S subscales showed a significant change in Total Social Problem Solving (t(14) = 3.76; P < 0.008), and RPS (t(14) = 2.24, P < 0.05), and a decrease in NPO (t(14) = 2.79, P < 0.008), Impulsivity/Carelessness (t(14) = 2.81, P < 0.008) and Avoidance Style (t(14) = 2.79; P < 0.01). Thus Total SPS NPO and Impulsivity/Carelessness were significant following a Bonferroni correction. GSES (t(14) = 3.23, P < 0.05), BDI-II (t(14) = 2.89, P < 0.05), the ‘company’ item of CANFOR (t(14) = 3.23, P < 0.05) and HoNOS – Secure item number 9: ‘Problems with Relationships’ (t(14) = 2.06, P < 0.05) also changed significantly post-treatment. Table 1 also shows in two-thirds of completers (n = 10; 66%) improved on total SPS RPS, Impulsivity/Carelessness, Avoidance Style, Self-efficacy (GSES) and Depression (BDI). No patients got worse. Non-completers did not complete a second SPSI-R:S. However, they did not show any significant pre–post changes on the other measures. Differences postintervention for completers versus non-completers (Table 2).

Table 2

Post-group psychometric assessment: completers versus non-completers (n = 24)

Measure	Completers (n = 15)		Non-completers (n = 9)		t-value	Cohen's d	Effect size (r)
	M	SD	M	SD
Generalized self-efficacy	36.78	7.04	22.16	3.11	t(22) = 3.36, P < 0.05)	2.68	0.80^†
BDI-II	6.23	4.34	17.13	7.45	t(22) = 3.78, P < 0.05	−1.78	−0.66
CANFOR (Item 4: company)	0.58	0.98	1.60	1	t(22) = 2.87, P < 0.05	−1.03	−0.45
HoNOS – Secure (Item 9: problems with relationships)	1.87	0.83	2.20	0.89	t(22) = 2.98, P < 0.05	−0.38	−0.18

BDI, Beck Depression Inventory; CANFOR, Camberwell Assessment of Need Forensic Version; HoNOS, Health of the Nation Outcome Scale

Note: negative effect sizes should be interpreted as a scale direction rather than a perceived lack of performance

^†Large effect size

Large effect sizes were noted for self-efficacy (GSES) and moderate effect size for depression (BDI).

Discussion

The current study suggests that for women in secure settings a social problem-solving group can effect positive changes in social problem-solving abilities and adds to an emergent literature on the value of this approach with individuals whose RPS abilities are impeded by negative orientation, impulsivity and avoidance.⁹ In contrast to previous studies,^13,14 the current evaluation used a stricter measure of treatment completion (70% of treatment sessions) and also concomitantly monitored mood/depression suggesting that mood improvements were associated with improved social problem-solving.^15,16 Current findings accord with research linking problem-solving deficits to depression and with studies that have used problem-solving training as a treatment for depression.³²

Baseline scores for SDSI-R:S subscales were comparable with those of vulnerable adult prisoners^13,14 and with scores for mentally ill³³ and personality disordered offenders.⁹

Women completing treatment in the current sample showed a significant improvement on overall problem-solving: in particular they became less negatively oriented toward problem-solving (NPO), and less impulsive/careless in their approach to problem-solving (ICS). They are also reported (to a lesser extent) being less avoidant (AS) and employing a more rational approach to social problem-solving (RPS). Changes with a large effect size were evident on the SPSI-R subscales of total problem solving, NPO, ICS and AS. This is of particular importance in a population with a dominant diagnosis of borderline personality disorder where impulsivity and emotional dysregulation is a defining characteristic.³³ Findings were paralleled by an increase in self-efficacy (GSES) and by lower ratings of depression (BDI-II) with changes showing a large and a moderate effect size respectively. While effectiveness research of this sort³⁴ inevitably sacrifices a degree of methodological rigour in the interests of ecological representativeness the future use of multisite investigations will undoubtedly enhance the generalization of favourable outcomes.

Limitations of this study include a reliance on self-report measure, and a small, possibly non-representative sample of women in medium secure care. Although no other manualized group treatments were concurrent with this study intervention it is difficult to distinguish effects of a problem-solving intervention in the current setting from other interventions and medication.¹⁷ A further possible confounding variable is the likelihood that patients who fail to complete a particular treatment show a generally lower level of treatment engagement.³⁴

The problem of treatment adherence remains a problem of significance in medium secure settings³⁵ and non-completion of group treatment rate of one-third (37.5%) raises questions of both acceptability of treatment³⁶ and of the timing of cognitive behavioural interventions.³⁷ Nonetheless, data derived from the current study indicate that women undergoing treatment in medium secure settings can derive benefit from a social problem-solving group that seeks to address some of the core needs of female patients admitted to such settings.^15,16

References

Department of Health. Womens Health: Into the Mainstream. London: Department of Health, 2002

Department of Health. Mainstreaming Gender and Women's Mental Health. Implementation guidance. London: Department of Health, 2003

Coid

, Kahtan

, Gault

, Jarman

. Women admitted to secure forensic psychiatry service I: comparison of women and men. J Forensic Psychol 2000;11:275–95

Long

, Hall

, Craig

, Mochty

, Hollin

. Women referred for medium-secure inpatient care: a population study over a six-year period. J Psychiatry Intensive Care 2011;7:17–26

Carey

, Carey

. Social problem-solving in dual diagnosis patients. J Psychopathol Behav Assessment 1990;12:247–54

Keltikangas-Jarvinen

. Social problem-solving and the development of aggression. In: McMurran

, McGuire

eds. Social Problem-Solving and Offending. Chichester, UK: Wiley, 2008:31–50

Herrick

, Elliott

. Social problem-solving abilities and personality disorder characteristics among dual diagnosed persons in substance abuse treatment. J Clin Psychol 2001;51:75–92

McMurran

, Duggan

, Christopher

, Huband

. The relationship between personality disorders and social problem solving in adults. Pers Indiv Diff 2007;42:145–55

McMurran

, Fyffe

, McCarthy

, Duggan

, Latham

. ‘Stop and Think’ social problem-solving therapy with personality disordered offenders. Crim Behav Ment Health 2001;11:273–85

10.

Nezu

, D'Zurilla

, Zwick

, Nezu

. Problem-solving therapy for adults. In: Chang

, D'Zurilla

, Sanna

eds. Social Problem-Solving: Theory, Research and Training. Washington, DC: American Psychological Association, 2007:171–81

11.

McMurran

, Egan

, Duggan

. ‘Stop and Think’ social problem–solving therapy with personality disordered offenders. In: McMurran

, McGuire

eds. Social Problem Solving and Offending. Chicester UK: Wiley, 2005:207–20

12.

Townsend

, Hawton

, Artman

, . The efficacy of problem-solving treatment after deliberate self-harm: meta-analysis of randomized controlled trials with respect to depression, hopelessness and improvement in problems. Psychol Med 2001;31:979–88

13.

Landenberger

, Lipsy

. The positive effects of cognitive-behavioural programmes for offenders: a meta-analysis of factors associated with effective treatment. J Exp Criminol 2005;1:451–76

14.

Hayward

, McMurran

, Sellen

. Social problem-solving in vulnerable adult prisoners: profile and intervention. J Foren Psychiatry Psychol 2008;19:243–8

15.

Biggar

, Power

. A controlled, problem-solving, group-based intervention with vulnerable incarcerated young offenders. Int J Offender Ther 2002;46:678–98

16.

Davies

, Oldfield

. Treatment need and provision in medium secure care. Br J Foren Practice 2009;11:24–30

17.

Long

, Fulton

, Hollin

. The development of a ‘Best Practice’ service for women in a medium-secure psychiatric setting: treatment components and evaluation. Clin Psychol Psychother 2008;15:304–19

18.

Nezu

, Nezu

, D'Zurilla

. Solving Life's Problems: A 5-Step Guide to Enhanced Well Being. New York: Springer Publishing Company, 2007

19.

McMurran

, Richardson

, Egan

, Ahmadi

. Social problem-solving in mentally disordered offenders: a brief report. Crim Behav Ment Health 1999;9:315–22

20.

D'Zurilla

, Nezu

, Maydeu-Olivares

. Manual for Social Problem Solving Inventory-Revised. Nawaada, NY: Multihealth Systems, 2002

21.

Heppner

, Petersen

. The development and implications of a personal problem-solving inventory. J Couns Psychol 1982;29:165–75

22.

Beck

, Steer

, Brown

. The Beck Depression Inventory. 2nd edn. San Antonio TX: The Psychological Corporation, 1996

23.

Jerusalem

, Schwarzer

. Self-efficacy as a resource factor in stress appraisal process. In: Schwarzer

ed. Self-efficacy: Thought Control of Action. Washington, DC: Hemisphere; 1992:P191–215

24.

Schwarzer

. Measurement of Perceived Self-Efficacy: Psychometric Scales for Cross-Cultural Research. Berlin: Freie Universitat, 1993

25.

Long

, Fulton

, Fitzgerald

, Hollin

. Group substance abuse treatment for women in secure services. Ment Health Subst Use: Dual Diagnosis 2010;3:227–37

26.

Thomas

, Harty

, Parrott

, McCrane

, Slade

, Thorneycroft

. CANFOR: Camberwell Assessment of Need Forensic Version. A Needs Assessment for Forensic Mental Health Service Users. London: Gaskell, 2003

27.

Sugarman

, Walker

. Health of the Nation Outcome Scales for Users of Secure Services. London: Royal College of Psychiatrists Research Unit, 2004

28.

Dickens

, Sugarman

, Walker

. HoNOS – Secure: a reliable outcome measure for users of secure and forensic mental health services. J Foren Psychiatry Psychol 2007;18:507–14

29.

Pirkis

, Burgess

, Kirk

, Dodson

, Coombs

, Williamson

. A review of the psychometric properties of the Health of the Nation Outcome Scales (HoNOS) family of measures. Health Qual Life Outcomes 2005;3:76

30.

Long

, Dickens

, Sugarman

, Craig

, Mochty

, Hollin

. Tracking risk profiles and outcome in a medium secure service for women: use of the HoNOS – Secure. Int J Forensic Ment Health 2010;19:215–55

31.

Cohen

. Statistical Power Analysis for the Behavioural Sciences. 2nd edn. Hillsdale NJ: Erlbaum, 1988

32.

D'Zurilla

, Nezu

. Problem-Solving Therapy: a Positive Approach to Clinical Intervention. New York: Springer, 2007

33.

Koons

, Robbins

, Tweed

, Lynch

, Gonzalez

, Moore

. Efficacy of dialectical behaviour therapy in women veterans with personality disorder. Behav Ther 2001;32:371–90

34.

Day

, Casey

, Ward

, Howells

, Wess

. Transitions to Better Lives. Offender Readiness and Rehabilitation. Ulfculme Cullompton Devon: Willan Publishing, 2010:141–57

35.

Gudjonsson

, Young

, Yates

. Motivating mentally disordered offenders to change: instruments for measuring patients perception and motivation. J Foren Psychiatry Psychol 2007;18:74–89

36.

McMurran

, McCulloch

. Why don't offenders complete treatment? Prisoner's reasons for non-completion of a cognitive skills programme. Psychol Crime Law 2007;13:245–54

37.

Morris

, Moore

. An evaluation of group work as an intervention to reduce the impact of substance misuse for offender patients in a high secure hospital. J Foren Psychi Psych 2009;20:559–76