Intensive cognitive-behaviour group therapy for ...

Intensive cognitive-behaviour group therapy for diagnostically heterogeneous groups of patients with psychiatric disorder

Aust NZ J Psychiatry Downloaded from informahealthcare.com by University of Western Australia on 03/01/11 For personal use only.

J. J. Manning, G . R. Hooke, D. A. Tannenbaum, T. H. Blythe, T. M. Clarke

This uncontrolled study evaluates the efficacy of a combined treatment of medications (for the majority of patients) and a closed group, intensive (twoweek) cognitive-behaviour therapy programme for heterogeneous groups of psychiatric patients. Five hundred and thirty-one patients at a private psychiatric clinic were included in the study. Self-report measures of depression, anxiety, self-esteem and “locus of control” were administered before and after treatment and at intervals up to one year later. Statistically and clinically significant improvements were found in all measures and these improvements were maintained up to one year. The results provide support for the efficacy of the treatment. Australian and New Zealand Journal of Psychiatry 1994; 28:667-674 Cognitive-behaviour therapy in a group format is increasingly recognised as an effective treatment for a variety of clinical disorders including Major Depression [ 11, chronic pain [2], Bulimia [3], Panic Disorder [4] and Agoraphobia [5]. A group therapy format has the appeal of cost-effectiveness in many hospital/clinic settings. With few exceptions 151, group cognitive therapy has been evaluated when applied to a closed group for 6-20 sessions, usually held on a weekly basis. Again with few exceptions [6], groups have been diagnostically homogeneous: for example, Depression or Bulimia.

Mounts Bay Clinic, 35 Richardson Street, West Perth, Western Australia J. J . Manning BSc(Hons), MPsych G. R. Hooke BAppSc (OT). GradDip (Hlth Sci) D. A. Tannenbaum MB, BCh. FRANZCP T. H. Blythe MBBS, AIT(Pharm), BSc, MSc, FPS, FRANZCP T. M. Clarke MBBS. FRANZCP Correspond with Mr Geoff Hooke

The aim of the present study was to evaluate the efficacy of a large-scale programme with an intensive group format (60 hours of therapy over two weeks) in a day hospital setting using heterogeneous groups where participants have a mix of psychiatric disorders or presenting problems. Follow-up data were collected to 12 months after the group programme. A further aim was to investigate the efficacy of this form of treatment for different categories of psychiatric diagnoses. Cognitive-behaviour therapy seeks to correct unrealistic or irrational thinking through cognitive restructuring and behavioural experiments. While the success of cognitive-behaviour therapy has perhaps been most extensively investigated in the treatment of depression, its application was never intended to be confined to depression [7,8]. The relevance of the cognitive-behavioural model can be appreciated in any situation where an individual’s interpretations of external events and internal thoughts or sensations can lead to differing behavioural reactions and intensities of affect. Beck et al 171 described dysfunctional patterns of thinking associated with the range of anxiety disor-


668

INTtNSlVE COGNITIVE-BEHAVIOUR GROUP THERAPY

ders. Generalised Anxiety Disorder, for example, often involves exaggerated fears of physical illness or injury, inability to cope, and fears of rejection, depreciation or domination. Recent cognitive-behavioural models of Panic Disorder 19.101 described the central cognitive component of catastrophic misinterpretation of bodily sensations. Emmelkamp [ 1 I ] and Steketee and Foa 1121 reviewed cognitive deficits or maladaptive beliefs that contribute to Obsessive-Compulsive Disorder, including unrealistic threat appraisal and a strong belief in perfcctionisrn. Fairburn [ I3 I outlined a cognitive-behavioural treatment for Bulimia aimed at modifying dysfunctional beliefs and values relating to body shape and weight, and the associated “perfectionistic tendencies and reliance on external criteria to gauge self-worth” (p. 178). Finally. Beck and Freeman I 141 recently described cognitive schema thought to underlie the Personality Disorders, together with outlines for cognitive-behavioural therapy. Given the identification of dysfunctional patterns of thinking in the presentation of a variety of clinical disorders, it seems logical to assume that a general cognitivc-behaviour therapy group approach will be useful in the treatment o f a wide range of disorders. Many hospital or outpatient facilities face practical difficultics in arranging homogeneous groups with sufficient numbers of patients. Heterogeneous groups represent a potential solution, with the advantage of offering places to patients at an optimal time for change rather than employing long waiting lists.

Method Subjects The sample consisted of 53 I patients referred to the Day Hospital programme run over two weeks by six consultant psychiatrists in a private psychiatric clinic (with both in- and out-patient facilities). Patients were referred to the programme over a period of two years and two months. DSM-I11 diagnoses were established by the psychiatrisis before referral to the programme. All patients received concurrent treatment from their psychiatrist. One hundred and forty-three patients (approximately 275%of the sample) were inpatients at the clinic at some time within the month beforc or month after participation in the programme. Group size was limited t o X participants with a mean of 6.4 (mode =

7). Group composition was not coordinated with respect to demographic variables such as age or sex, or to diagnostic categories.

Design Repeated measures on self-report questionnaires were taken at various intervals before the programme and up to 12 months post-group follow-up. Owing to ethical considerations of patients’ clinical needs, no control groups were employed. Patients completed questionnaires at the initial assessment by their psychiatrist, although this was not possible in all cases (for example, some acute inpatient admissions). Questionnaires were completed again at pre-group, postgroup, and at 6-week and 3-, 6- and 12-month f o I I ow - up i n t erv al s. Patients al so com p I e ted programme evaluation questionnaires post-group.

Therapists Two therapists were assigned to each two-week group, making a total of four therapists conducting two overlapping groups (a new group starting each week). Therapists held formal qualifications in either clinical psychology or occupational therapy, with training and experience in cognitive-behaviour therapy. The mean level of experience in the adult psychiatric area was 6 years. In most cases a male and a female therapist were assigned to each group. Daily meetings were held between the two therapists to coordinate the programme and discuss patients’ progress. To maximise uniformity of treatment, a therapist manual was employed which provided detailed outlines of each session of the programme.

Measures Five self-report questionnaires were administered pre-group through to 12 months follow-up: the Beck Depression Inventory (BDI) 11.51, the State and Trait versions of the Spielberger State-Trait Anxiety Inventory (STAI) 1161, the Rosenberg Self-Esteem Scale (RSES) [ 171 and a Locus of Control Scale [ 181. The BDI is a 21-item self-report scale designed to measure level of depression among clinical and nonclinical populations and is widely used in research on depression. Ranges on the BDI represent the following categories of mood: (0- 10) normal mood; ( 1 1 - 16)mild mood disturbance; ( 17-20) borderline clinical depres-


J.MANNING, G.HOOKE, D.TANNENBAUM, T.BLYTHE, T.CLARKE

sion; (21-30) moderate clinical depression; (3 1-40) severe clinical depression; and (over40) extreme c h i cal depression [ 15 ]. The STAI is a self-report inventory that measures both state and trait anxiety (20 items in each scale). The State scale is considered to be an indicator of transitory anxiety while the Trait scale measures more stable individual differences in anxiety proneness. The STAI has been widely used in both clinical and research settings. STAI score ranges represent the following levels of anxiety: (less than 40) normal; (41-50) mild; (5 1-60) moderate; (61-70) severe; and (71-80) extreme [ 161. The RSES is a measure of general self-concept consisting of 10 items using a 4-point Likert-type response format. A high score reflects a high level of self-esteem with a maximum positive score of 40. The Locus of Control Scale assesses an individual’s sense of control over themselves and their lives. A low score indicates a greater sense of control. Scores range from 0 to 85. The programme evaluation questionnaire assesses patients’ perceptions of, and satisfaction with, the programme, using a 1-9 point scale for items such as: “How has your ability to cope with your problems changed over the past two weeks?” Responses range through: very much worse ( l ) , worse (3), no change (5), better (7), to very much better (9).

Description of programme The two-week Day Hospital programme was conducted over a period of ten working days for a total of 60 hours. Each day consisted of 4 sessions, each of 90 minutes duration. Two follow-up reinforcement sessions (also 90 minutes) were arranged 6 weeks and 3 months after the end of the programme. All patients were reviewed by their psychiatrist at completion of the programme, and many received further treatment. In aminority of cases, patients were involved in further therapy with either one of their two group therapists. This level of therapy may be individual, couple or family in nature. The programme is structured to include several major components: - cognitive therapy (with self-monitoring) - behavioural assignments to challenge thoughts and beliefs - realistic goal setting - assertion skills training

669

- self-esteem work to challenge cognitive distortions about self - two interactive lectures on the nature of anxiety and depression - stress management - emphasising importance of balanced lifestyle - relaxation training - one “supporters” session where each participant attends with a significant person (e.g. spouse, friend, family member). This structure allows sufficient time and flexibility for the discussion of both group and individual issues. Each patient is supplied with a standard manual (approximately 150 pages) which covers these components.

Results Table 1 shows the general demographic characteristics of the sample. The modal patient was a married female in her late 30s. The average age was 39.6 years (SD=10.5, minimum=17.8, maximum=69.9). For the purpose of this study, DSM-111 diagnoses were grouped into six categories: Affective Disorders, Anxiety Disorders, Adjustment Disorders, Psychosis, V Codes and Others. This categorisation was agreed upon by the consultant psychiatrists. Affective Disorders included both unipolar and bipolar depression, and depression with psychotic features. Anxiety Disorders included panic disorder, agoraphobia, phobias, generalised anxiety disorder and obsessive-compulsive disorders. The “Others” category included personality disorders, addictions, sexual disorders and eating disorders. The percentages of patients in each category were: Affective Disorders 45.0%; Anxiety Disorders 26.2%; Adjustment Disorders 17.4%; V Codes 7.5%; Other 3.4% and Psychosis 0.6%. Twelve patients (2.2% of sample) did not complete the programme. The mean number of days attended by these patients was 5.1 (mode = 5). Diagnostic categories were: Depressive Disorder (6 patients), Anxiety Disorder (5) and Adjustment Disorder ( 1). Mean pre-group questionnaire scores for the 12 were: BDI - 19.2 (SD=12.7); State Anxiety - 46.3 (SD=I 1.6); Trait Anxiety - 53.7 (SD=10.2); Locus of Control - 31.5 (SD=I 1.1); and Self Esteem - 25.9 (SD=7.4). All these mean scores are more favourable than mean scores for the majority of patients who did complete the programme. Post-group data for an addi-

670

INTENSIVE COGNITIVE-BEHAVIOUR GROUP THERAPY

Table I . Demographic informationf o r 531 patients completing the Cognitive-Behaviour Therapy Programme


80

Sex Male Female Age (at onset) under 25

26-35 36-45 46-55 over 56

n

%

162 369

30.5 69.5

w

Beck

50

II

a Trait

0 30

~

66 135 195 89 46

12.4 25.4 36.7 16.8 8.7

Pre-grp

Post-grp

6 weeks

T.~I Aamlnlamlon

Marital status Single Marriedlde facto Divorcedkeparated Widow(er) Unknown

16

21.5 64.2 9.8 1.5 3.0

Employmentstatus Full time Part time Home duties Unemployed Unknown

360 20 105 14 32

67.8 3.8 19.8 2.6 6.0

Occupation Managers and administrators Professionals Para-professionals Tradespersons Clerks (including secretarial) Personal service and sales Plant and machine operators Labourers and related workers Inadequately described and NS Student Housewife Unknown

49 69 56 29 76 43 8 10 38 32 105 16

9.0 13.0 10.5 5.5 14.5 8.1 1.5 1.9 7.2 6.0 19.8 3.0

114 341 52

a

tional I5 patients who completed the programme were missing. At the time of participation in the group programme, 366 patients were taking medication (mainly antidepressants) and 141 patients were taking no medication; this information was missing for 24 patients. Outcome data from the entire sample were calculated in several ways. First, all questionnaire scores were averaged for each assessment time (pre- and post-group, through to 12 months follow-up). The

state

3 months

~~

6 months

nm.

Figure 1:Percentage of patients' scores within normal levels for BDI and State and Trait Scales of the STAI,from pre-group to 6 months follow-up. mean score for each measure showed an improvement from pre- to post-group that was maintained over the following year. Pre group sample size was 531 patients. The response rate at twelve months dropped to approximately 42.2 %. For statistical comparisons, the second analysis included only those patients with complete responses on each measure from pre-group to six months follow-up. Means, standard deviations (SD) and numbers of patients (n) are shown in Table 2. Repeated measures analyses of variance (ANOVA) with Dunn's t-test post hoc comparisons showed significant differences from pre- to post-group on all measures (ANOVA main effects; BDI: F=172, df=4,784; State Anxiety: F=144, df=4,768; Trait Anxiety: F=152, df=4,772; RSES: F=75, df=4,764; Locus of Control: F=135, df=4,772, all p values