163

Psychology and Psychotherapy: Theory, Research and Practice (2015), 88, 163–177 © 2014 The British Psychological Society www.wileyonlinelibrary.com

Understanding process in group cognitive behaviour therapy for psychosis Tania Lecomte1*, Claude Leclerc2, Til Wykes3, Luc Nicole4 and Amal Abdel Baki5 1

Department of Psychology, University of Montreal, Quebec, Canada Department of Nursing, University of Quebec at Trois-Rivieres, Quebec, Canada 3 Department of Psychology, Institute of Psychiatry, University of London, UK 4 Programme PEP (Premiers Episodes Psychotiques), Montreal University Institute of Mental Health (IUSMM), Montreal, Quebec, Canada 5 Programme JAP (Jeunes Adultes Psychotiques), Montreal University Hospital Center (CHUM), Montreal, Quebec, Canada 2

Background. Group cognitive behaviour therapy for psychosis (GCBTp) has shown to be effective in diminishing symptoms, as well as in improving other psychosocial dimensions such as self-esteem. But little is known regarding the processes that generate these therapeutic improvements and might be harnessed to further improve its effectiveness. Objectives. The current study aimed at investigating these processes, particularly those linked to interpersonal relationships. Design. The participants were all assessed at baseline, were given 24 sessions of GCBTp over the course of 3 months and were assessed again at post-treatment as well as 6 months later (9 months from baseline). Method. Sixty-six individuals with early psychosis took part in a study of GCBTp where therapist alliance and group cohesion were assessed at three time points during the therapy, and punctual (each session) self-perceptions on symptoms and optimism were collected. Results. Improvements in symptoms (BPRS), self-esteem (SERS-SF) and in self-perceived therapeutic improvements (CHOICE) were linked to specific aspects of the alliance, group cohesion, as well as optimism. The variables retained were not always overall scores, suggesting the importance of the variables at key moments during the therapy. Conclusions. The results clearly demonstrate the importance of the alliance and group cohesion, together significantly explaining improvements measured at post-therapy or follow-up.

Practitioner points  This study has attempted to focus mostly on relational aspects, as well as on self-perceptions, in the context of a GCBTp for individuals with early psychosis.

*Correspondence should be addressed to Tania Lecomte, Department of Psychology, University of Montreal, C-358, 90 Vincent d’Indy Street, C.P. 6128, Succ. Centre-Ville, Montreal, QC, Canada H3C 3J7 (email: [email protected]). DOI:10.1111/papt.12039

164

Tania Lecomte et al.

 This study also showed that these therapeutic relationships are especially useful when they are more stable and at specific moments during the therapy, namely when more difficult psychological work is done.

Psychotherapy research in domains outside psychosis is more advanced. There has been a focus on identifying key therapy ingredients and even the integration of various approaches. Most research in psychotherapy for psychosis is at an early stage focusing mainly on feasibility and efficacy (Lecomte & Lecomte, 2012). This is understandable given that until recently people with psychosis were considered as untreatable via psychotherapy, mostly because they did not respond well to traditional psychodynamic therapy (Lecomte & Lecomte, 2012; Lysaker, Roe, & Kukla, 2012). However, cognitive behaviour therapy for psychosis (CBTp) developed over the last 20 years has now proven its efficacy for diminishing psychotic symptoms and distress, according to more than 5 meta-analyses and 30 randomized controlled trials (Addington & Lecomte, 2012; Mueser, Deavers, Penn, & Cassisi, 2013). Although the effect sizes vary according to the quality of the trials (Wykes, Steel, Everitt, & Tarrier, 2008), as often seen in psychotherapy research in general (Wampold, 2001), there is now a maturity in the field as guidelines adopt CBTp as a treatment of choice for individuals with schizophrenia (e.g., National Institute for Health and Care Excellence [NICE] guidelines: NICE, 2014, CG187; and PORT guidance: Dixon et al., 2010). The obvious next step is to try to isolate the key ingredients of therapy not only to allow treatment tailoring but also to increase CBTp’s therapeutic effects. Few studies have investigated specific client characteristics that predict higher improvement and even fewer have replicated the results. Predictors of positive CBTp outcomes also more often than not emerge from trials that showed no advantage of CBTp over the control condition, limiting their scientific weight (e.g., Bentall et al., 2002; Freeman et al., 2013; Jackson et al., 2005). Of the personal variables measured, cognitive flexibility, that is, the ability to consider alternatives for beliefs (Brabban, Tai, & Turkington, 2009; Garety et al., 1997; Naeem, Kingdon, & Turkington, 2008), insight into illness (Naeem et al., 2008), female gender (Brabban et al., 2009), fewer negative symptoms before therapy (Thomas, Rossell, Farhall, Shawyer, & Castle, 2011), personal attributions regarding the cause of the illness (Freeman et al., 2013), higher social functioning (Allott et al., 2011), better optimism for the future and believing they are responsible for change in their lives (Myhr et al., 2013) have been described as predicting better CBTp outcomes regarding symptoms and/or functioning. Personality traits, specifically conscientiousness, have also been linked to greater effects in group CBTp (GCBTp; Beauchamp, Lecomte, Lecomte, Leclerc, & Corbiere, 2013), especially regarding increased use of active coping strategies to cope with stress. In terms of essential components of CBTp, studies have been conducted to grasp what seemed essential for clinicians as well as for clients who have received CBTp. According to a Delphi study conducted with 28 CBTp therapists (Morrison & Barratt, 2010), at least 77 components were considered essential for CBTp to be efficacious, covering areas such as therapist attitudes and use of the CBT model and techniques. A recent qualitative study, looking at nine clients’ experience of CBTp and what they considered were the essential elements, revealed that specific techniques and principles of CBTp, such as normalization, seeking facts or alternative explanations, were seen as very helpful. The results also emphasized the relationship with the therapist, namely regarding trust, engagement, and partnership (Berry & Hayward, 2011). Similarly, participants interviewed following GCBTp mentioned specific CBT techniques (the stress-vulnerability-competence model, seeking alternatives, trying new coping skills) as well as relationship components (better

Understanding process in group CBT for psychosis

165

support network, sharing with others, learning from others, competent therapists) as being central to their positive experience of the therapy (Lecomte, Leclerc, & Wykes, 2012). Although techniques are essential and partial CBTp that does not include most CBT elements has been documented as not as effective as full CBTp (Dunn et al., 2012), the context of a working therapeutic relationship such as a good alliance, and group cohesion in group contexts, are considered essential for benefiting from those techniques (Castonguay, Constantino, & Holtforth, 2006; Lecomte & Lecomte, 2002). Studies in CBTp have demonstrated links between the strength of the therapeutic alliance and: (1) symptom reduction (Lecomte, Laferriere-Simard, & Leclerc, 2012; Solomon, Draine, & Delahey, 1995; Svensson & Hansson, 1999), (2) treatment adherence (Weiss, Smith, Hull, Piper, & Huppert, 2002), (3) higher attendance and participation in CBTp (Bentall et al., 2002; Frank & Gundersen, 1990; Johnson, Penn, Bauer, Meyer, & Evans, 2008; Lecomte, Leclerc, et al., 2012), (4) increased self-esteem (Lecomte, Laferriere-Simard, et al., 2012), and (5) better work performance (Davis & Lysaker, 2007). Most studies have used a single-point measure of alliance (e.g., EvansJones, Peters, & Barker, 2009; Johnson et al., 2008) as a predictor (typically the sixth session), but therapeutic alliances do evolve or change overtime. Greater fluctuations in the alliance as assessed by the client are linked to worse negative symptoms at posttherapy (Lecomte, Laferriere-Simard, et al., 2012), so a single-point estimate may be misleading. Other interpersonal factors may affect group rather than individual psychotherapy and one of these suggested factors is group cohesion (Norcross & Wampold, 2011). Group cohesion will develop or change in similar ways than the alliance will fluctuate with a therapist. These changes are also influenced by the content of sessions, the techniques used, and the personal changes that clients undergo. Given the important clinical effects recently documented in a trial on GCBTp for early psychosis (Lecomte et al., 2008), a process study looking into the static and dynamic predictors of outcomes in GCBTp was developed. The objectives of this study were to investigate relationship variables that might help explain some of the clinical improvements that have been documented in GCBTp for early psychosis, while taking into account the specific stages of the therapy. More specifically, the authors were interested in looking at predictors of change pertaining to symptoms, self-esteem, and selfperception of therapeutic improvement. Fluctuations over brief time periods have been documented for symptoms and self-esteem (Thewissen, Bentall, Lecomte, van Os, & MyinGermeys, 2008) as well as for the therapeutic alliance (Lecomte, Laferriere-Simard, et al., 2012) in individuals with psychosis; it is therefore important to consider multiple time point assessments to investigate the effects of such fluctuations. This exploratory study might therefore reveal predictors that might have been missed in cross-sectional studies.

Method Design This is an uncontrolled longitudinal study recruiting from a convenience sample within two early psychosis clinics. The participants were all assessed at baseline, were given 24 sessions of GCBTp over the course of 3 months, and were assessed again at post-treatment as well as 6 months later (9 months from baseline).

166

Tania Lecomte et al.

Participants A total of 66 individuals with early psychosis were recruited from 2008 until 2011 from the Louis-H Lafontaine Hospital and CHUM’s first-episode programmes in Montreal. The study was approved by both hospitals’ research and ethic’s boards. Informed consent was carefully observed, with the use of a true/false questionnaire to ensure the understanding of the study prior to signing the consent form. Inclusion criteria were as follows: presenting persistent or sporadic psychotic symptoms (even with prescribed antipsychotic medication), receiving services from a participating first-episode programme as an outpatient, and having the ability to read and write in French. Participants were on average 26 years old (SD = 6), mostly male (70%), Caucasian (73%) and had completed on average 12 years of education (SD = 3.2). The two first-episode clinics offer services to their clients for up to 5 years; as a result the median number of psychiatric hospitalizations was 2 (mode = 1, SD = 1.8). Chart review diagnoses revealed that 60% (n = 40) were diagnosed with schizophrenia, 18% (n = 12) with schizo-affective disorder, 15% (n = 10) with psychosis not otherwise specified, and 5% (n = 4) with depression with psychotic symptoms. Overall, 97% of the sample mentioned taking their medication as prescribed (N = 64), 35% (N = 22) for more than 12 months and 20% (n = 13) for less than 3 months.

Measures Client and relationship process variables The following process measures were administered during the therapy: the QuickLL (Lecomte, Spidel, & Leclerc, 2005): This brief questionnaire contains 14 items, rated on a 3-point scale (worst than usual, same as usual, better than usual), offering a snapshot of the participant at each session regarding his assessment of eight dimensions, namely the alliance with the therapists, the group cohesion, appreciation of the group therapy, optimism for the future, self-esteem (two items), current mood/anxiety (five items), ability to meet personal goals, and current presence of psychotic symptoms (two items), such as voices or distressful thoughts. A mean score and a fluctuation score (standard deviation) were calculated for each dimension summing each month of therapy. An improvement score (i.e., proportion of ‘better than usual’ compared to ‘worse than usual’) was also calculated for each month of therapy. A total score for the entire therapy was calculated as well. The brief version of the Working Alliance Inventory (WAI; Corbiere, Bisson, Lauzon, & Ricard, 2006; Horvath & Greenberg, 1989) was administered to therapists and participants once a month during therapy, for a total of three times. This version holds 12 items rating the Bond, Task, and Goal dimensions of the WAI for each month of therapy. We also calculated the total score for each month as well as the overall mean and fluctuations (standard deviations) across the therapy for each domain (bond, task, goals) and for the total score. As sessions were conducted by two co-therapists, the co-therapists’ ratings were averaged (i.e., both ratings of each participant, and each participant’s ratings of both co-therapists). The Cohesion Questionnaire (Piper, Marrache, Lacroix, Richardson, & Jones, 1983) was also administered three times, once a month, to participants and therapists. The participants rated the therapists as well as the other group members (together), and the therapists rated each participant. As with the alliance measure, we averaged the cotherapists’ ratings in order for each participant to have only one rating of the group

Understanding process in group CBT for psychosis

167

therapists (together) and for the co-therapists to have only one rating of each participant. The questionnaire assesses various aspects of group cohesion, namely regarding seeing positive qualities in group members or therapists, feeling a personal compatibility with group members or therapists, and appreciating the therapists’ leadership (or appreciating the participant’s role in the group). Each item is rated on a 7-point Likert scale. We calculated monthly scores for each domain and a total score, as well as overall mean and fluctuation score (standard deviation) across the therapy for each domain and for the total score.

Baseline, post-therapy and follow-up assessments.  Brief Psychiatric Rating Scale-Expanded version (BPRS-E; Lukoff, Nuechterlein, & Ventura, 1986) was administered as a semi-structured interview. As one goal of GCBTp is to diminish symptoms, BPRS-E measures overall psychiatric symptoms (24 items), positive psychotic symptoms (delusions and hallucinations: six items) and negative symptoms (e.g., motor retardation, emotional withdrawal: five items). As recommended by the UCLA BPRS fidelity gold standard (Ventura, Green, Shaner, & Liberman, 1993), consensus ratings had to be reached on a minimum of six interviews before interviewers could independently conduct BPRS interviews. As a form of interviewer blindness, different interviewers were used at each time point to ensure that the ratings were not influenced by the memory of a previous assessment or by knowledge regarding the study.  Self-esteem was measured by the Self-Esteem Rating Scale Short-Form (SERS-SF), which has 10 items assessing positive and 10 assessing negative self-esteem, rated on a 7-point Likert scale and has been validated with people with severe mental illness (Lecomte, Corbiere, & Laisne, 2006).  Clients’ perceptions of change during therapy was assessed using the Choice of Outcome (Greenwood et al., 2010). The CHoice of Outcome In Cbt for psychosEs (CHOICE) is a psychometrically reliable instrument measuring a single dimension regarding change: coping ability/severity of symptoms, as well as perceived satisfaction, and importance. Only the coping/severity of symptoms score was used. Each item is rated on a 10-point Likert scale.

Baseline only measures.  Sociodemographic information was obtained at baseline with the PSR toolkit (Arns, 1998), whereas diagnostic and medical information were obtained in chart reviews.

Therapy. Participants received 24 hourly sessions of GCBTp, twice a week, for 3 months. Groups consisted of six to eight participants – a total of 10 groups were conducted for this study. Two co-therapists led the groups: one from the first-episode clinic’s mental health team and one from the research team at each site (for a total of six participating therapists). The therapists all had previous experience working with individuals with psychosis, were either occupational therapists, psychologists or psychiatric nurses, and had received an intensive 2-day training on how to conduct the groups while following the GCBTp manual offered. All the sessions were audiotaped

168

Tania Lecomte et al.

for quality control. Supervision was offered every 2 weeks, but was occasionally offered more frequently when an issue arose in a session that needed to be discussed promptly. The GCBTp manual was developed by three of the authors (Tania Lecomte, Claude Leclerc, & Til Wykes) and integrates the principles CBTp, but adapted to a group format and tailored for individuals with early psychosis (see Lecomte, Leclerc, Wykes, & Lecomte, 2003 for more details). The GCBTp manual is built incrementally, starting with more neutral concepts and themes to more personal themes and more complex techniques and skills. In terms of process, the first month of therapy aims at instilling trust, feelings of safety and building group cohesion and therapeutic alliance. The second month aims at modifying distressing thoughts and changing behaviours by introducing various cognitive and behavioural techniques, and can be experienced as more difficult for some. The last month addresses coping strategies and strengths, but also the end of the therapy, including post-therapy goals. In terms of content, the manual is built in four parts, each containing six sessions: (1) stress: how it affects me, (2) testing hypotheses and looking for alternatives, (3) drugs, alcohol and how I feel, and (4) coping and competence. The manual follows a goal-oriented approach (instead of problem-based) while teaching essential CBTp techniques such as normalization (realizing that others have had similar experiences), the A-B-Cs of CBT (linking events to thoughts and emotional and behavioural consequences), Socratic questioning, checking the facts, finding alternatives, trying new coping strategies and modifying attributions.

Statistical analyses All analyses were carried out in SAS 9.3. Scores for the process measures were computed for the beginning (after 1 month), middle (after 2 months), or end of therapy (after 3 months), as well as score variations (SDs) and overall means across therapy. To confirm that the participants had improved on the selected outcome variables, repeated measures ANOVA as a linear mixed model estimated by restricted maximum likelihood (REML) with completed data were used to test the difference between baseline, post-therapy and follow-up for the six outcomes (BPRS negative, BPRS positive, BPRS total, CHOICE, SERSSF positive and negative scales). A random effect for the therapy group was used to take into account the dependency between the participants of a same group. A Bonferroni correction was used for the multiple comparisons. The results of the REML method were replicated with the missing data imputation last observation carried forward (LOCF) approach. Given the important proportion of missing data for both outcomes and independent variables, a comparison was made to verify if the participants who did not participate at the follow-up were different from those for whom we have values at followup. The same comparison was also done to compare the participants who took part in less than 50% of the sessions (12 sessions or less) for the QuickLL variables with those for whom we have values for more than 12 sessions. Evolution among T0, T1 and T2 was calculated by assessing the difference for each outcome. Univariate mixed linear regressions with missing data imputation were done for each outcome where changes overtime were found, with each of the 191 independent variables, and with a random effect for the therapy group. A subset of variables was created with those significant at the 5% level. A forward selection was then used on this subset of variables to obtain the final multiple regression model. R2 for mixed models were calculated to present the reduction in residual variance attributed to the fixed effects, using the formula 1 Vf/Vn, where Vf is the residual variance of the full model, whereas

Understanding process in group CBT for psychosis

169

Vn represents the residual variance of the null model containing only the random effect, with its variance fixed to its full model value.

Results Outcomes No significant difference was found between those lost to follow-up and those for whom we obtained data, except for self-esteem – those lost to follow-up had higher baseline positive self-esteem, t(58) = 2.60, p < .05 – therefore a covariate for the baseline score was added to the model to take into account this imbalance at T2. As for attendance, no significant differences were detected between those participating in more or less than 12 sessions, except for drug and alcohol consumers who significantly attended fewer sessions than non-consumers (v2 = 4.09, p < .05). The results were similar for the analyses using the linear mixed model ANOVA with REML and those using missing data imputation. Significant improvements over time were observed for symptoms, both overall and positive symptoms, and for positive self-esteem. Table 1 shows the means, ANOVA results, and significant differences following Bonferroni corrections. As can be noticed, overall symptoms as well as positive symptoms significantly improved between pre-therapy and follow-up, as well as between pretherapy and post-therapy, and post-therapy and follow-up for positive symptoms. No such improvements were measured for negative symptoms, although the baseline score was low. Data for self-esteem were only available at T0 and T2 (see Table 1), that is, at the baseline and at the 6 month follow-up. Significant improvements were found between those two time points, but only for the positive self-esteem scale. As for the CHOICE, significant perceived changes in coping and symptom severity were not found, but a trend between pre- and post-therapy could be seen.

Table 1. Means and standard deviations for outcomes at each time point and ANOVAs with Bonferroni correction

Variable

T0 (baseline)

Symptoms BPRS total Positive

1.76 (0.48) 1.97 (0.48)

T1 (post-therapy)

T2 (6 month follow-up)

ANOVA

Bonferroni T0 > T2 T0 > T1, T0 > T2

1.69 (0.45) 1.77 (0.83)

1.62 (0.44) 1.65 (0.79)

F(2, 122) = 5.09** F(2, 122) = 8.52***

Negative 1.64 (0.69) 1.71 (0.73) Self-esteem SERS Positive 46.98 (11.32) Missing Negative 29.54 (23.01) Perceived improvements CHOICE Coping/ 6.42 (1.91) 6.86 (1.58) symptom severity

1.64 (0.69)

F(2, 122) = 0.57

47.89 (10.65) 31.50 (21.85) 6.81 (1.73)

F(1, 80) = 8.70** F(1, 80) = 0.06

T0 < T2

F(2, 117) = 3.15*

Note. BPRS = Brief Psychiatric Rating Scale; SERS = Self-Esteem Rating Scale; CHOICE = CHoice of Outcome In Cbt for psychosEs. *p < .05; **p < .01; ***p < .001.

170

Tania Lecomte et al.

Fluctuations in alliance and cohesion Table 2 shows the alliance and cohesion scores at each month during the therapy. As can be observed, scores change over time. For instance, the alliance total score as rated by the participant significantly improved between the first and second month of therapy, t (29) = 2.06, p < .05, and between the first and third month of therapy, t(23) = 3.43, p < .01. The therapist’s assessment of the alliance also improved over time, although not as much, with the overall alliance being greater after the second month of therapy compared to the first, t(42) = 1.99, p = .52, and the third month being greater than the first, t(36) = 2.02, p = .51. Similarly, the group cohesion scores changed over time, with the overall appreciation of others in the group by the participant significantly improving between the first and third month of therapy, t(17) = 2.29, p < .05. The other total change scores did not reach significance. As can be seen in Table 2, some of the changes were not positive, with some therapists’ ratings regarding participants significance as a Table 2. WAI alliance and cohesion scores at each month of therapy and overall Variable

1st (session 8) M (SD)

2nd (session 16) M (SD)

Alliance assessed by participant WAI total 4.47 (1.22) 4.67 (1.42) WAI-goals 4.26 (1.40) 4.68 (1.53) WAI-tasks 4.16 (1.32) 4.42 (1.26) WAI-bond 4.98 (1.29) 5.14 (1.52) Alliance assessed by therapist WAI total 4.14 (0.66) 4.32 (0.68) WAI-goals 3.91 (0.77) 4.04 (0.77) WAI-tasks 3.79 (0.62) 4.06 (0.64) WAI-bond 4.71 (0.88) 4.86 (0.87) Cohesion assessed by participant regarding others in the group Cohesion-total 3.02 (0.76) 3.45 (0.96) Cohesion-compatibility 3.00 (1.09) 3.59 (1.17) Cohesion-positive 3.22 (0.68) 3.48 (0.82) qualities Cohesion-significance 2.70 (1.02) 3.19 (1.29) as group member Cohesion assessed by participant regarding therapists Cohesion-total 3.09 (0.93) 3.19 (1.04) Cohesion-compatibility 2.56 (1.23) 2.84 (1.27) Cohesion-positive 4.23 (1.12) 4.15 (1.20) qualities Cohesion-satisfaction 2.49 (1.14) 2.64 (1.19) with leaders role Cohesion assessed by therapists regarding participant Cohesion-total 2.95 (0.73) 2.98 (0.65) Cohesion-compatibility 2.11 (0.86) 2.11 (0.76) Cohesion-positive 3.17 (0.61) 3.30 (0.51) qualities Cohesion-significance 3.62 (1.22) 3.51 (1.21) as group member

3rd (session 24) M (SD)

Overall – M (SD)

4.86 (1.04) 4.73 (1.09) 4.55 (0.98) 5.30 (1.32)

4.77 (1.26) 4.68 (1.25) 4.55 (1.15) 5.23 (1.41)

4.37 (0.70) 4.11 (0.67) 4.16 (0.75) 4.84 (0.86)

4.18 (0.65) 3.89 (0.69) 3.89 (0.63) 4.75 (0.82)

3.32 (0.89) 3.48 (1.17) 3.17 (0.64)

3.19 (0.67) 3.23 (0.98) 3.24 (0.60)

3.31 (1.17)

3.03 (1.04)

3.23 (0.83) 2.94 (0.91) 4.01 (1.19)

3.28 (0.87) 2.89 (1.01) 4.28 (1.06)

2.73 (0.96)

2.71 (1.06)

2.95 (0.79) 2.13 (0.87) 3.34 (0.60)

2.97 (0.65) 2.17 (0.78) 3.27 (0.47)

2.88 (1.02)

3.50 (1.06)

Note. WAI = Working Alliance Inventory; M = mean; SD = standard deviation.

Understanding process in group CBT for psychosis

171

group member becoming worse over time (e.g., between the second and third month: t (27) = 2.98, p < .01).

Process involved in symptom outcomes Regarding total psychiatric symptoms as measured by the BPRS at post-therapy, four process scores emerged as significant predictors of fewer symptoms: participants’ assessment of the cohesion (total) during the therapy, the therapists’ assessment of the group cohesion-total at the end of therapy, and the alliance rated in the last month of therapy by the participant, and better alliance with the therapist (bond score) as perceived by the participant overall during the therapy (see Table 3). At the 6 month follow-up, the variables most strongly predicting improved symptoms were as follows: the therapists’ rating of shared tasks in the alliance at the end of therapy, b = 0.33, t(20) = 3.41, p < .01, and the mean cohesion score pertaining to the participant’s feeling of compatibility with other group members, b = 0.18, t(20) = 2.43, p < .05. As for positive symptoms, the participants’ assessment of the cohesion score of personal compatibility with the therapist at the second month of therapy, as well as the therapists’ assessment of cohesion regarding perceived positive qualities in the participant at the end of therapy, predicted improvements at post-therapy (see Table 3). At the 6 month follow-up, only the alliance as measured by therapists Table 3. Multiple mixed regression model results for symptoms, and perceptions of therapeutic improvements at post-therapy with missing data imputation

Outcome BPRS total symptoms

BPRS positive symptoms

CHOICE coping/ severity of symptoms

Process variables Mean cohesion-total Cohesion therapist total (end therapy) QuickLL alliance (end of therapy) Mean WAI alliance participant (bond) Cohesion-compatibility with therapist (mid therapy) Cohesion (therapist) positive qualities in participant (end of therapy) QuickLL optimism 1st month SD cohesion personal compatibility with other participants

t(df)

b

4.19 (18)*** 4.54 (18)***

0.42 0.27

3.94 (18)***

0.51

2.96 (18)**

0.18

2.92 (15)*

0.29

3.35 (15)**

0.64

3.76 (12)** 3.90 (12)**

3.11 1.87

Variance parameters for cohort/residual

R²a

0/0.06

0.52

0/0.30

0.21

0.63/1.53

0.27

Note. BPRS = Brief Psychiatric Rating Scale; WAI = Working Alliance Inventory; CHOICE = CHoice of Outcome In Cbt for psychosEs; SD = Standard Deviation. a 2 R is calculated with the residual variance of the full model and the null model with the cohort variance fixed to its value in the full model. *p < .05; **p < .01; ***p < .001.

172

Tania Lecomte et al.

regarding shared tasks at the end of therapy predicted improvement, b = (26) = 2.92, p < .01.

0.49, t

Process involved in self-esteem outcome Data were not collected at post-therapy. Only the standard deviation of the cohesion score regarding the participant perceiving positive qualities in the other participants was predictive, with a more stable score predicting more improvements in positive selfesteem at follow-up, b = 59.5, t(9) = 3.15, p < .05.

Process involved in perceived therapeutic improvements Two variables were the strongest predictors of perceived improvements at post-therapy assessed with the CHOICE: poor optimism at the beginning of therapy (after eight sessions) and variations across therapy in the participants’ perception of feeling compatible with other participants in the group. At the 6 month follow-up, only poor optimism at the beginning of therapy remained as a predictor of CHOICE improvements on coping/severity of symptoms, b = 2.06, t(33) = 2.88, p < .01.

Discussion The processes we uncovered as predictors of therapy outcome are not random and not all linked to a single-time point but can be understood in terms of what is happening during the sessions. Although meta-analyses (not specific to psychosis) typically report small effect sizes of the alliance on clinical outcomes (0.23–0.26; Horvath & Bedi, 2002; Martin, Garske, & Davis, 2000), symptomatic improvements have been documented as being linked to the alliance between a client and his therapist in GCBTp (Lecomte, LaferriereSimard, et al., 2012). Our results confirm these results, further specifying that improvement in total symptoms at post-therapy is linked to the participant’s evaluation of the alliance with the therapist during the last month of therapy, as well as specifically the bond created with the therapists throughout the therapy. A strong feeling of cohesion, both from the participant’s and the therapist’s perspective was also linked to improved overall symptoms. These results suggest that in a group context, it is important for the participants to feel safe and to relate to the other participants and to the therapists throughout the therapy in order to improve on their overall symptoms. Although total symptoms include anxiety and depressive symptoms, not only psychotic symptoms, cohesion was also important for positive symptoms in terms of compatibility with the therapists after 1 month of therapy, and for the therapist in terms of recognizing the participant’s positive qualities at the end of the therapy. Regarding the former result, 1 month after the beginning of therapy is where most of CBTp techniques aiming at cognitive and behaviour change are introduced, and where participants’ trust in the therapists and the group play an important role in enabling the person to seek alternatives for beliefs, and change their behaviours. As for the therapists’ cohesion and alliance with the participants, these have been found to be greatly influenced by participant’s attendance rates (Johnson et al., 2008; Lecomte, Leclerc, et al., 2012), with therapists feeling more engaged with participants who appear more motivated and come to most sessions. Symptoms are known to fluctuate overtime with life experiences and stress (Myin-Germeys & van Os, 2007), but seemed to continuously improve following GCBTp

Understanding process in group CBT for psychosis

173

in the context of a strong therapeutic relationship and group cohesion. From the client’s perspective, follow-up overall symptoms improvements were also linked to overall personal compatibility with the therapists, suggesting a lasting effect of a significant therapeutic relationship. The therapist’s perception of the alliance regarding shared tasks at the end of therapy was also linked to improved symptoms at follow-up. This suggests that the participants who fared better in the long run were those who related more to their therapists at the end of therapy, who in turn felt they agreed on specific skills and techniques the participants needed to continue using. It could also suggest that therapists invested more in participants who appeared to have gained more from the therapy, and who were therefore more likely to continue using the skills learned during the sessions. The improvements at the 6 month follow-up on positive self-esteem was linked to overall perceiving positive qualities in the other participants in the group, without much variations across the therapy. Recognizing positive qualities in others and in oneself are important aspects of self-esteem, namely in building strong senses of identity and belonging (Reasoner, 1992). The GCBTp offered in this study not only focused on selfesteem at different sessions throughout the therapy but also emphasized at the end of the therapy everything that had been accomplished by each, acknowledging each person’s strengths and qualities, as well as helping them realize the changes they had made and the goals they had met (Lecomte et al., 2003). As for realizing how the GCBTp had an impact on coping with symptoms and feeling more confident in overcoming obstacles and dealing with stress, one of the predictors of self-perceived therapeutic improvements (i.e., CHOICE) at post-therapy and at follow-up indicated that those who were more pessimistic about the future early on in the therapy felt they gained the most at the end of the therapy. This suggests that GCBTp is not only meant for those who are initially convinced of its value and that it can in fact offer hope. The other significant predictor was variations in feeling compatible with others in the group. Given the link with improved coping with symptoms, this could be interpreted as participants feeling more compatible as the group evolved and participants learned more about each other and realized they might have more in common than what they initially thought. Feelings of belonging and of being part of a group with similar attributes are particularly important for individuals following a first-psychotic episode when fear of social alienation is present. Feeling more compatible with the group members can also help people become more open to trying new coping strategies suggested by other group members for instance. This study has limitations. For one, although we used validated relationship measures, namely the alliance and group cohesion, the brief measure used at each session (QuickLL) has not been validated with a large sample yet. We also only included measures from the participant or the relationship with the therapists and have not specifically looked at therapist variables, namely competency. Given the importance of the results pertaining to the alliance and group cohesion, therapist competency in developing such therapeutic working relationships while delivering useful CBT tools to participants is essential and insufficiently studied in CBTp. The small sample size, especially at the 6 month follow-up with missing data imputations, as well as the missing data at various time points for the process measures during the therapy, are limitations. However, such an attrition rate is fairly typical of studies conducted with people followed in first-episode programmes (Spidel, Lecomte, & Leclerc, 2006) and is mostly due to people moving on with their lives and not taking the time to conduct assessments. Imputation by LOCF method has limitations associated to its assumptions that responses should be constant over time and that missing values are completely random. Furthermore, as many other

174

Tania Lecomte et al.

imputation techniques, it underestimates the standard errors and overestimates the statistical tests. Strengths of this study include that our results take into consideration cohort effects and interdependence of results within group therapy. The absence of significant improvements on negative symptoms, although documented in CBTp (Rector, Beck, & Stolar, 2005) and in GCBTp before (Lecomte, Leclerc, et al., 2012) did not allow for us to look at predictors for this variable. However, few participants scored high on negative symptoms in this sample. Finally, the absence of information regarding processes involved in the self-esteem outcome at post-therapy is unfortunate, especially given the strong impact of GCBTp on self-esteem at post-therapy previously documented (Lecomte et al., 2008). Researchers conducting CBTp studies have until recently spent a lot of energy on trying to demonstrate its efficacy. Now that CBTp is recognized as an evidenced-based intervention, a closer look at process variables explaining change is warranted. This study has attempted to focus mostly on relational aspects, as well as on self-perceptions, in the context of a GCBTp for individuals with early psychosis. The results clearly demonstrate the importance of the alliance and group cohesion together significantly explaining improvements measured at post-therapy or follow-up. This study also showed that these therapeutic relationships are especially useful when they are more stable and at specific moments during the therapy, namely when more difficult psychological work is done. More studies are warranted, not only on alliance and group cohesion, but also regarding the therapists’ needed skills and competencies to help clients achieve their therapeutic goals.

References Addington, J., & Lecomte, T. (2012). Cognitive behaviour therapy for schizophrenia. F1000 Medicine Reports, 4, 6–9. doi:10.3410/M4-6 Allott, K., Alvarez-Jimenez, M., Killackey, E. J., Bendall, S., McGorry, P. D., & Jackson, H. J. (2011). Patient predictors of symptom and functional outcome following cognitive behaviour therapy or befriending in first-episode psychosis. Schizophrenia Research, 132, 125–130. doi:10.1016/j. schres.2011.08.011 Arns, P. (1998). Canadian version of the PSR Toolkit. London, UK: Ontario Federation of Community Mental Health and Addiction Programs. Beauchamp, M. C., Lecomte, T., Lecomte, C., Leclerc, C., & Corbiere, M. (2013). Do personality traits matter when choosing a group therapy for early psychosis? Psychological Psychotherapy, 86, 19–32. doi:10.1111/j.2044-8341.2011.02052.x Bentall, R., Tarrier, N., Lewis, S., Haddock, G., Kinderman, P., Kingdon, D., & the SoCRATES team (2002). The therapeutic alliance in early psychosis. Acta Psychiatrica Scandinavica, 106(s413), 69–106. doi:10.1186/1744-859X-12-14 Berry, C., & Hayward, M. (2011). What can qualitative research tell us about service user perspectives of CBT for psychosis? A synthesis of current evidence. Behavioural and Cognitive Psychotherapy, 39, 487–494. doi:10.1017/S1352465811000154 Brabban, A., Tai, S., & Turkington, D. (2009). Predictors of outcome in brief cognitive behavior therapy for schizophrenia. Schizophrenia Bulletin, 35, 859–864. doi:10.1093/schbul/sbp065 Castonguay, L. G., Constantino, M. J., & Holtforth, M. G. (2006). The working alliance: Where are we and where should we go? Psychotherapy: Theory, Research, Practice, Training, 43, 271–279. doi:10.1037/0033-3204.43.3.271 Corbiere, M., Bisson, J., Lauzon, S., & Ricard, N. (2006). Factorial validation of a French short-form of the Working Alliance Inventory. International Journal of Methods in Psychiatric Research, 15, 36–45. doi:10.1002/mpr.27

Understanding process in group CBT for psychosis

175

Davis, L. W., & Lysaker, P. H. (2007). Therapeutic alliance and improvements in work performance over time in patients with schizophrenia. The Journal of Nervous and Mental Disease, 195, 353–357. doi:10.1097/01.nmd.0000261954.36030.a1 Dixon, L. B., Dickerson, F., Bellack, A. S., Bennett, M., Dickinson, D., Goldberg, R. W., . . . Kreyenbuhl, J. (2010). The 2009 schizophrenia PORT psychosocial treatment recommendations and summary statements. Schizophrenia Bulletin, 36, 48–70. doi:10.1093/schbul/sbp115 Dunn, G., Fowler, D., Rollinson, R., Garety, P., Freeman, D., Kuipers, E., & Bebbington, P. (2012). Effective elements of CBT for psychosis. Psychological Medicine, 42, 1057–1068. doi:10.1017/ S0033291711001954 Evans-Jones, C., Peters, E., & Barker, C. (2009). The therapeutic relationship in CBT for psychosis: Client, therapist and therapy factors. Behavioural and Cognitive Psychotherapy, 37, 527–540. doi:10.1017/S1352465809990269 Frank, A. F., & Gundersen, J. G. (1990). The role of the therapeutic alliance in the treatment of schizophrenia: Relationship to course and outcome. Archives of General Psychiatry, 47, 228– 236. doi:10.1001/archpsyc.1990.01810150028006 Freeman, D., Dunn, G., Garety, P., Weinman, J., Kuipers, E., Fowler, D., & Bebbington, P. (2013). Patients’ beliefs about the causes, persistence and control of psychotic experiences predict takeup of effective cognitive behaviour therapy for psychosis. Psychological Medicine, 43, 269–277. doi:10.1017/S0033291712001225 Garety, P., Fowler, D., Kuipers, E., Freeman, D., Dunn, G., Bebbington, P., & Jones, S. (1997). London–East Anglia randomised controlled trial of cognitive-behaviour therapy for psychosis. II. Predictors of outcome. British Journal of Psychiatry, 171, 420–426. doi:10.1192/bjp.171.5.420 Greenwood, K. E., Sweeney, A., Williams, S., Garety, P., Kuipers, E., Scott, J., & Peters, E. (2010). CHoice of Outcome In Cbt for psychosEs (CHOICE): The development of a new service user-led outcome measure of CBT for psychosis. Schizophrenia Bulletin, 36, 126–135. doi:10.1093/ schbul/sbp117 Horvath, A. O., & Bedi, R. P. (2002). The alliance. In J. C. Norcross (Ed.), Psychotherapy relationships that work: Therapist contributions and responsiveness to patients (pp. 37–69). New York, NY: Oxford. Horvath, A. O., & Greenberg, L. S. (1989). Development and validation of the Working Alliance Inventory. Journal of Counseling Psychology, 36, 223–233. doi:10.1037/0022-0167.36.2.223 Jackson, H., McGorry, P., Edwards, J., Hulbert, C., Henry, L., Harrigan, S., & Power, P. (2005). A controlled trial of cognitively oriented psychotherapy for early psychosis (COPE) with four-year follow-up readmission data. Psychological Medicine, 35, 1295–1306. doi:10.1017/ S0033291705004927 Johnson, D. P., Penn, D. L., Bauer, D. J., Meyer, P., & Evans, E. (2008). Predictors of the therapeutic alliance in group therapy for individuals with treatment-resistant auditory hallucinations. British Journal of Clinical Psychology, 47, 171–183. doi:10.1348/014466507X241604 Lecomte, T., Corbiere, M., & Laisne, F. (2006). Investigating self-esteem in individuals with schizophrenia: Relevance of the SERS. Psychiatry Research, 143, 99–108. doi:10.1016/j. psychres.2005.08.019 Lecomte, T., Laferriere-Simard, M.-C., & Leclerc, C. (2012). What does the alliance predict in group interventions for early psychosis? Journal of Contemporary Psychotherapy, 42, 55–64. doi:10. 1007/s10879-011-9184-2 Lecomte, T., Leclerc, C., Corbiere, C., Wykes, T., Wallace, C. J., & Spidel, A. (2008). Group cognitive behaviour therapy or social skills training for individuals with a first episode of psychosis? Results of a randomized controlled trial. Journal of Nervous and Mental Disease, 196, 866–875. doi:10. 1097/NMD.0b013e31818ee231 Lecomte, T., Leclerc, C., & Wykes, T. (2012). Group CBT for early psychosis – Are there still benefits one year later? International Journal of Group Psychotherapy, 62, 309–322. doi:10.1521/ijgp. 2012.62.2.309

176

Tania Lecomte et al.

Lecomte, T., Leclerc, C., Wykes, T., & Lecomte, J. (2003). Group CBT for clients with a first episode of psychosis. Journal of Cognitive Psychotherapy: An International Quaterly, 17, 375–384. doi:10.1891/jcop.17.4.375.52538 Lecomte, T., & Lecomte, C. (2002). Towards uncovering robust principles of change inherent to CBT for psychosis. American Journal of Orthopsychiatry, 72, 50–57. doi:10.1037/0002-9432. 72.1.50 Lecomte, T., & Lecomte, C. (2012). Are we there yet? Commentary on special issue on psychotherapy integration for individuals with psychosis. Journal of Psychotherapy Integration, 22, 375–381. doi:10.1037/a0029575 Lecomte, T., Spidel, A., & Leclerc, C. (2005). Expectancies and outcomes in inpatient CBT groups for psychosis – A pilot study. Paper presented at the Mental Health and the Justice System Across the Lifespan Conference, Vancouver, BC. Lukoff, D., Nuechterlein, K., & Ventura, J. (1986). Manual for the Expanded Brief Psychiatric Rating Scale. Schizophrenia Bulletin, 12, 578–602. doi:10.1093/schbul/12.4.578 Lysaker, P. H., Roe, D., & Kukla, M. (2012). Psychotherapy and rehabilitation for schizophrenia: Thoughts about their parallel development and potential integration. Journal of Psychotherapy Integration, 22, 344–355. doi:10.1037/a0029580 Martin, D. J., Garske, J. P., & Davis, M. K. (2000). Relation of the therapeutic alliance with outcome and other variables: A meta-analytic review. Journal of Consulting and Clinical Psychology, 68, 438–450. doi:10.1037/0022-006X.68.3.438 Morrison, A. P., & Barratt, S. (2010). What are the components of CBT for psychosis? A Delphi study. Schizophrenia Bulletin, 36, 136–142. doi:10.1093/schbul/sbp118 Mueser, K. T., Deavers, F., Penn, D. L., & Cassisi, J. E. (2013). Psychosocial treatments for schizophrenia. Annual Review of Clinical Psychology, 9, 465–497. doi:10.1146/annurevclinpsy-050212-185620 Myhr, G., Russel, J. J., Saint-Laurent, M., Tagalakis, V., Belisle, D., Khodari, F., & Pinard, G. (2013). Assessing suitability for short-term cognitive-behavioral therapy in psychiatric outpatients with psychosis: A comparison with depressed and anxious outpatients. Journal of Psychiatric Practice, 19, 29–41. doi:10.1097/01.pra.0000426325.49396.4c Myin-Germeys, I., & van Os, J. (2007). Stress-reactivity in psychosis: Evidence for an affective pathway to psychosis. Clinical Psychology Review, 27, 409–424. doi:10.1016/j.cpr.2006.09. 005 Naeem, F., Kingdon, D., & Turkington, D. (2008). Predictors of response to cognitive behavior therapy in the treatment of schizophrenia: A comparison of brief and standard interventions. Cognitive Therapy Research, 32, 651–656. doi:10.1007/s10608-008-9186-x National Institute for Health and Care Excellence (NICE) (2014). Psychosis and schizophrenia in adults (CG178). Retrieved from http://guidance.nice.org.uk/CG178 Norcross, J. C., & Wampold, B. E. (2011). Evidence-based therapy relationships: Research conclusions and clinical practices. Psychotherapy, 48, 98–102. doi:10.1037/a0022161 Piper, W. E., Marrache, M., Lacroix, R., Richardson, A. M., & Jones, B. D. (1983). Cohesion as a basic bond in groups. Human Relations, 36, 93–108. doi:10.1177/001872678303600201 Reasoner, R. W. (1992). Building self-esteem in the elementary schools: Teacher’s manual (2nd ed.) Palo Alto, CA: Consulting Psychologists Press. Rector, N. A., Beck, A. T., & Stolar, N. (2005). The negative symptoms of schizophrenia: A cognitive perspective. Canadian Journal of Psychiatry, 50, 247–257. doi:10.3410/f.13853959. 15292059 Solomon, P., Draine, J., & Delahey, M. A. (1995). The working alliance and consumer case management. Journal of Mental Health Administration, 22, 126–134. doi:10.1007/ BF02518753 Spidel, A., Lecomte, T., & Leclerc, C. (2006). Community implementation successes and challenges of a cognitive-behavior therapy group for individuals with a first episode of psychosis. Journal of Contemporary Psychotherapy, 36, 51–58. doi:10.1007/s10879-005-9006-5

Understanding process in group CBT for psychosis

177

Svensson, B., & Hansson, L. (1999). Therapeutic alliance in cognitive therapy for schizophrenic and other long-term mentally ill patients: Development and relationship to outcome in an in-patient treatment programme. Acta Psychiatrica Scandinavica, 99, 281–287. doi:10.1111/j.16000447.1999.tb07226.x Thewissen, V., Bentall, R. P., Lecomte, T., van Os, J., & Myin-Germeys, I. (2008). Fluctuations in selfesteem and paranoia in the context of daily life. Journal of Abnormal Psychology, 117, 143– 153. doi:10.1037/0021-843X.117.1.143 Thomas, N., Rossell, S., Farhall, J., Shawyer, F., & Castle, D. (2011). Cognitive behavioural therapy for auditory hallucinations: Effectiveness and predictors of outcome in a specialist clinic. Behaviour and Cognitive Psychotherapy, 39, 129–138. doi:10.1017/S1352465810000548 Ventura, J., Green, M. F., Shaner, A., & Liberman, R. P. (1993). Training and quality assurance with the Brief Psychiatric Rating Scale. International Journal of Methods in Psychiatric Research, 3, 221–244. Wampold, B. E. (2001). The great psychotherapy debate: Model, methods, and findings. Mahwah, NJ: Lawrence Erlbaum. Weiss, K. A., Smith, T. E., Hull, J. W., Piper, A. C., & Huppert, J. D. (2002). Predictors of risk of nonadherence in outpatients with schizophrenia and other psychotic disorders. Schizophrenia Bulletin, 28, 341–349. doi:10.1093/oxfordjournals.schbul.a006943 Wykes, T., Steel, C., Everitt, B., & Tarrier, N. (2008). Cognitive behavior therapy for schizophrenia: Effect sizes, clinical models, and methodological rigor. Schizophrenia Bulletin, 34, 523– 537.doi:10.1093/schbul/sbm114 Received 28 January 2014; revised version received 12 June 2014

Copyright of Psychology & Psychotherapy: Theory, Research & Practice is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.

Understanding process in group cognitive behaviour therapy for psychosis.

Group cognitive behaviour therapy for psychosis (GCBTp) has shown to be effective in diminishing symptoms, as well as in improving other psychosocial ...
165KB Sizes 3 Downloads 4 Views