Abstract
Introduction
In addition to the symptoms of their disorder, people suffering from severe, chronic mental illness, such as schizophrenia and bipolar affective disorder, present with an impairment of functioning in a range of different domains of their everyday life. Personal hygiene and care, the initiation and maintenance of interpersonal and social relationships, and the capacity to learn and enjoy life are all generally impaired. As a result, the patients’ autonomy is often severely affected (Goldman et al., 1981). Moreover, most patients require repeated and often long psychiatric hospitalization (Zissi, 2002).
Medical treatment of these patients is based on the use of antipsychotic medication, the effectiveness of which has been well established in terms of symptom remission, relapse prevention, and readmission to hospital. In addition to medication, psychosocial interventions have been developed, such as counselling and supportive therapy, individual, group and family therapy, occupational therapy and vocational rehabilitation. These psychosocial strategies include specific psychological treatment approaches (such as family psychoeducation, cognitive behaviour therapy for ‘voices’, and so on) and more general psychosocial interventions. Both approaches aim at helping patients develop their skills, improve their relationships with others and deal with their symptoms (Breier and Strauss, 1983; Cohen and Berk, 1985; Corrigan and Storzbach, 1993; Dobson et al., 1995; Falloon and Talbot, 1981; Kanas and Barr, 1984; Kuipers et al., 1992).
Among psychological treatment techniques, various forms of group therapy have attracted attention both on the grounds of cost-effectiveness and therapeutic efficacy. Therapists save time and it has been suggested that desired changes are effected faster, last longer and interpersonal relationships are promoted to a greater extent than those in other types of therapy (Kanas, 1996; Sandell et al., 2000).
Four main group therapy approaches have been described, based on principles of the therapeutic model used: psychodynamic, psychoeducational, cognitive-behavioural, and supportive/interpersonal.
Psychoeducational therapy, with elements from other theoretical models, originates from the work of Lazell (1921) and Marsh (1933). Lazell approached war veterans with a diagnosis of schizophrenia by using psychodynamic principles and psychoeducational techniques. Marsh described a therapy programme for psychotic patients at the Worcester State Hospital in Massachussetts, USA, which included group activities involving brief speeches in simple terms, watching short films, advising, problem solving, role playing and homework. This approach aimed at (a) increasing knowledge about mental illness, (b) improving skills that help the patient cope with their symptoms and problems, (c) managing tension and anxiety, (d) recognizing symptoms of relapse and (e) complying with pharmaceutical treatment.
Specifically, cognitive-behavioural psychotherapy for psychotic patients was based on intervention strategies that had been previously developed for anxiety (Beck et al., 1985) and depression (Beck et al., 1979). However, it was orientated towards psychotic symptoms and the biopsychosocial structure of the illness (Zubin and Spring, 1977). Cognitive processes research in psychosis suggests that hallucinations and delusions may be extreme versions of the cognitive processes of estimation and belief formation, with a tendency for people with psychosis to overestimate coincidences (Maher, 1988) and use false ideas (Bentall et al., 1994). It has been reported that 5% of the general population report quasi-hallucinatory experiences (Tien, 1992), but most of them realize their internal origin (Johns and van Os, 2001; van Os, 2003).
Psychodynamic group therapy arose from the application of psychoanalytic concepts to groups of patients and was first reported by Schilder (1939), Semrad (1948), Standish and Semrad (1951), Lawton (1951) and Pinney (1956). Schilder used psychodynamic techniques in the therapy of a mixed group of neurotic and psychotic outpatients. After World War II, Semrad and his colleagues at the Boston State Hospital started using psychoanalytic techniques (such as free associations, transference and exploration of the unconscious) in the therapy of psychotic patients in groups. The aim of the psychodynamic approach was to decrease conflicts, improve the function of the ego (by increasing insight into one’s problems and malfunctioning behaviours) and enhance personal autonomy.
Supportive/interpersonal group therapy was first applied by Powdermaker and Frank (1953; Frank, 1955). Yalom (1975, 1983) is the current, main representative of this approach. The discussion of interpersonal problems, the encouragement of interaction between members of the group, and the interpretation of this interaction in the here and now, contribute to decreasing social isolation and improving one’s ability to relate to others.
Empirical research on the issue of group therapy has increasingly attracted attention recently (Huxley et al., 2000; Jones, 2004). The aim of this paper is to review the studies that have been published on group therapy with patients with schizophrenia or bipolar affective disorder, over two decades. The scope and limitations of these studies are presented and topics for future research are suggested.
Method
An electronic search was conducted through Medline and PsychINFO to identify articles relevant to group therapy of adults with schizophrenia and bipolar affective disorder. Articles that were published in the English language between January 1986 and May 2006 were searched and studies with a control group and of at least 20 participants were included. Keywords used were psychotic-disorders-therapy, bipolar disorder, psychotherapy, group therapy, psychosocial deprivation, psychoeducation, psychosocial treatments, social skills training, schizophrenia, schizoaffective disorders and review. This procedure was supplemented with a manual search of the bibliography included in the articles found, and of relevant reviews, in order to gain access to additional references. Narratives and case presentations were not included. Studies with vaguely described interventions or with inadequate outcome information were excluded. The methodological quality of the studies was assessed according to the following variables: diagnostic criteria used, demographic representation of indicators, study design (e.g. sample size, control groups, direct vs. indirect assessment), assessor’s characteristics (e.g. bias and blindness), and assessment tools (e.g. outcome measures). The design of a study was considered ‘controlled’ when treatments were comparatively evaluated.
The focus was placed on outcomes of group therapies that evaluated symptoms (positive or negative), social functioning, relapses, coping with relapse, rehospitalizations, mood stability, expressed emotion, compliance, medication management skills, symptom management skills, control of voices, grooming, recreation, vocational and cognitive skills, cognitive insight, stigma, trust, and satisfaction. Results were reported as positive only when supported by statistical significance.
Results
Psychoeducational group therapy for adults with schizophrenia and bipolar affective disorder
Schizophrenia
Five studies of psychoeducation-based group interventions were identified (Table 1). One of the studies (Goldman and Quinn, 1988) evaluated brief psychoeducational programmes for inpatients. The other four (Herz et al., 2000; Hornung et al., 1999; Pitschel-Walz et al., 2006; Shin and Lukens, 2002) had a more extensive programme, mostly for outpatients.
Psychoeducational group therapy of schizophrenic patients
See Appendix for abbreviations
Psychoeducational group therapy of bipolar patients
See Appendix for abbreviations
All studies reported improvement in at least one parameter. Goldman and Quinn (1988) demonstrated statistically significant improvement in both positive and negative symptoms and in insight. Herz et al. (2000), Hornung et al. (1999) and Pitschel-Walz et al. (2006), with follow-up lasting five, one and a half, and two years respectively, demonstrated statistically significant improvement concerning readmissions. Shin and Lukens (2002) reported statistically significant improvement with regards to positive symptoms, stigmatization and coping skills.
Bipolar affective disorder
Four studies of psychoeducational interventions were found (Colom et al., 2003; Peet and Harvey, 1991; van Gent and Zwart, 1991, 1993). All referred to outpatients and all studies reported improvement in at least one parameter.
Colom et al. (2003) included a large sample and the treatment had an extensive duration. This study showed a decrease in the frequency of manic-depressive episodes but longer periods for the attainment of symptom remission, compared to the control group. Follow-up after 24 months showed (a) a decrease in manic-depressive episodes, (b) a longer period of symptom remission, (c) fewer readmissions and a shorter duration of hospitalization, and (d) higher lithium levels.
The brief intervention of Peet and Harvey (1991) concluded that there is improvement concerning knowledge about the disorder and compliance with medication.
Van Gent and Zwart (1993) conducted a brief intervention study and found only subjective reports of benefit. A similar intervention was used for relatives of patients with bipolar affective disorder, in which they reported an increase in the understanding of the disorder, immediately after the intervention as well as six months later.
Cognitive-behavioural group therapy
Schizophrenia
Eight studies of group therapy for in- or outpatients with schizophrenia were found (Table 3).
Four of these studies followed the Programme of Social and Independent Living Skills Training developed by Liberman (1986, 1988) at the UCLA Centre for Research on Severe Mental Illnesses (Eckman et al., 1992; Liberman et al., 1998; Marder et al., 1996; Wallace et al., 1992). This was a well-structured training programme that promoted social functioning, management of symptoms, interpersonal interaction (e.g. participating in a conversation), management of pharmaceutical treatment, self-care, and work research. It used a variety of activities, problem-solving techniques and different types of tasks. The duration of the therapy varied from six months (Liberman et al., 1998) to two years (Marder et al., 1996). Two of these studies included a follow-up assessment (Liberman et al., 1998; Wallace et al., 1992).
Cognitive-behavioural group therapy for patients with schizophrenia
See Appendix for abbreviations
All the above studies presented evidence of significant improvement in at least one parameter: Eckman et al. (1992) in management skills concerning symptoms and medication; Liberman et al. (1998) in symptoms, social functioning and quality of life; Marder et al. (1996) in social skills; and Wallace et al. (1992) in self-care skills and management of medication.
The other four studies (Andres et al., 2000; Granholm et al., 2005; Hayes et al., 1995; McCay et al., 2006) used cognitive-behavioural techniques, such as training, modelling, rehearsal and reinforcement, set homework based on teaching and focused on specific aspects of the patient’s behaviour (e.g. eye contact, body language, speech). Two of these studies included follow-up (Andres et al., 2000; McCay et al., 2006)
All studies reported improvement in at least one parameter: Andres et al. (2000) in coping skills only in the follow-up; Hayes et al. (1995) in symptoms and quality of life; Granholm et al. (2005) in improvement of social skills and insight; and McCay et al. (2006) in the perception of the self, quality of life and symptoms.
Bipolar affective disorder
No published study of group therapy for adults with bipolar affective disorder using a cognitive-behavioural approach, with a control group, and with at least 20 participants, was found.
Supportive group therapy
Only one study was found using this model. This study of people with schizophrenia (Kanas et al., 1989) used an intervention of brief duration for outpatients with follow-up. It reported no improvement after therapy, but at follow-up improvement in both symptoms and social skills was reported.
Comparisons and combinations of therapeutic modalities
Nine studies comparing different modalities of group therapy were found (Andres et al., 2000; Bechdolf et al., 2004, 2005; Colom et al., 2003; Eckman et al., 1992; Hayes et al., 1995; Liberman et al., 1998; Marder et al., 1996; Pollack and Cramer, 1999). The most common comparison was between a specific type of group therapy (i.e. social skills training or psychoeducation) and a non-specific type of intervention (i.e. supportive group therapy). The specific types of group therapy were typically found to be more effective.
Bechdolf et al. (2004, 2005) compared the cognitive-behavioural model to that of psychoeducation and found a superiority of the first in terms of readmission and follow-up in six months, but this benefit was lost at the two-year follow-up. Liberman et al. (1998) reported a better outcome of the social skills group compared to the vocational skills group. Pollack and Cramer (1999), looking at bipolar patients, considered a comparison between groups exposed to either self-management or interaction skills training. Although neither of the two types of approach had been clearly preferred by the participants, their answers were useful in determining necessary changes of the techniques used.
No study was found comparing group therapy of schizophrenic or bipolar patients with some other type of psychosocial intervention, for example, family therapy or art therapy.
Seven studies were found regarding the combination of different psychosocial interventions including group therapy. Six of them (Table 4) combined two psychoeducational groups (patients and families). The seventh (Drury et al., 1996) reported positive results considering the symptoms of the disorder after a combination of individual and group cognitive-behavioural therapy. Buchkremer et al. (1997) reported improvement in social functioning, whereas three other studies (Herz et al., 2000; Hornung et al., 1999; Pitschel-Walzet al., 2006) reported improvement regarding readmission.
Combinations of psychotherapeutic interventions for patients with schizophrenia
See Appendix for abbreviations
Discussion
In terms of group therapy of patients with schizophrenia, the majority of the studies (10 out of 16) involved cognitive-behavioural therapy and focused on the development of social and vocational skills improvement, while five used a psychoeducational approach. Although the 10 studies mentioned above concerned the same type of approach, significant differences were found between them in terms of the duration of the intervention, follow-up period, sample size, use of a control group, use of evaluation tools and standardized outcome measures. For instance, duration varied from one study to another, from three months (Andres et al., 2000; McCay et al., 2006) to 18 and 24 months (Eckman et al., 1992 and Marder et al., 1996, respectively). Four studies included reassessment at follow-up (Andres et al., 2000; Liberman et al., 1998; McCay et al., 2006; Wallace et al., 1992). In all the studies, improvement, in at least one parameter, was significantly higher than that found in the control group. Improvement most frequently seen in particular skills and less often in clinical symptoms.
A point for discussion and a concern about the implementation of social skills training programmes in schizophrenia centres on the issue of whether the skills learnt at the sessions are generalized and retained in more real, naturalistic interactions in patients’ everyday life (Liberman and Wallace, 1985). Moreover, Gomes-Schwartz (1979) claims that patients with schizophrenia cannot respond to behavioural approaches in the same way that other psychiatric patients do. This raises a question concerning the validity and feasibility of behavioural techniques that are not especially designed for schizophrenia. Also, when studies differ significantly concerning cognitive-behavioural techniques used and the number of sessions, it is not clear whether the conclusions made are of equal credibility. Liberman (1985) claim that psychotic patients in different levels of their functionality need different skills training approaches. In general, social skills generalization could be evidenced by follow-up of patients but few studies provide this (only four in this paper: Andres et al., 2000; Hayes et al., 1995; Liberman et al., 1998; Wallace et al., 1992). However, meta-analyses (Benton and Schroeder, 1990; Dilk and Bond, 1996) of a relatively small number of homogeneous follow-up, case-control studies produced some evidence that training in social skills during hospitalization can improve patients’ social functioning many months after the intervention.
Differences in terms of duration and sample size are also observed in psychoeducation. One of the studies was conducted within 0.75 month (Goldman and Quinn, 1988) but another (Hornung et al., 1999) was conducted over 9.5 months, had a combined cognitive approach and used a large sample. Nevertheless, all studies reported improvement in at least one parameter.
A psychoeducational approach is based on learning processes and the management of symptoms and problems arising from the patient’s lack of knowledge and understanding of the illness. Long-term interventions or a follow-up programme using this approach would have more positive results, especially in people with cognitive deficits such as those found in people with schizophrenia, as they would better understand their illness.
Moreover, positive outcome could be enhanced if the family attended a parallel psychoeducational programme (Buchkremer et al., 1997; Grace et al., 1996; Herz et al., 2000; Hogarty et al., 1991; Hornung et al., 1996, 1999; McFarlane et al., 1995, 2003; Merinder et al., 1999; Murray-Swank and Dixon, 2004; Pitschel-Walz et al., 2001, 2006; Rummel-Kluge et al., 2006).
Increasing the frequency of the sessions would be another possible way to help inpatients more. Many services offer inpatients one session a week, but as the period of hospitalization is reduced, two sessions a week could reduce the number of dropouts from the service, thus helping more patients (Rummel-Kluge et al., 2006).
No studies of psychodynamic group therapy were identified for patients with psychosis, with a control group and with at least 20 participants. This was despite the long historic course of psychodynamic group therapy. This may be due to the fact that psychodynamic interventions are more timeconsuming and require planning for longer-term studies.
Moreover, in psychodynamic therapies one should use specific experience scales to evaluate the level of the patient’s involvement in the psychotherapeutic process – i.e. the expression of the patient’s personal, phenomenological view and whether this is used productively in the therapeutic session. In a high degree of involvement, emotions can be examined and points of reference can be found for problem resolution and self-understanding. Only a few standardized scales are available in this area and this is a topic that needs further development. One example of such an experience scale is the Experiencing Scale of Klein et al. (1970).
Four of the five studies that were found on group therapy of bipolar patients concerned the psychoeducation model (Colom et al., 2003; Peet and Harvey, 1991; van Gent and Zwart, 1991, 1993); three studies included reassessment at follow-up. Again, as in research concerning schizophrenics, there are major differences in sample size and duration of intervention, which affects the results.
Generally, there are significant differences between the studies in terms of sampling, duration of therapy, the evaluation tools and the measurement of outcome. These differences create many difficulties for the meta-analysis of a sufficient number of studies. In most studies there was insufficient information regarding the clinical state of the patients prior to the intervention (acute phase, chronic, residual, treatment resistant) and there was typically no reference regarding the training and competence of the therapists used. Aside from the above methodological problems, combination, comparative or cost-effectiveness studies are scarce.
The following important research questions cannot be answered on the basis of the available evidence and should be the subject of future research: whether (and under which conditions) group therapy is preferable to other psychological interventions and which are the relative advantages or shortcomings of each type of group therapy. In other words, what works for which type of psychotic patients? More evidence is required from well-structured, homogenous studies in order to document the effectiveness of each intervention and the predictors of successful outcome.
Conclusion
It can be concluded that the observed improvement after participation in programmes of group therapy referred mainly to an increase in social skills, reinforcement of self-care, improved compliance with medical treatment, and better coping strategies. Thus, gains in socialization and appropriate interpersonal interaction can be expected. Less improvement was observed with regards to positive symptom reduction. Group therapy, in various modalities, can be a clinically efficacious and cost-effective adjunct to medical treatment for adults suffering from the two major psychotic conditions.
