Exposure and response prevention (ERP) is the behavioural treatment of choice for obsessive-compulsive disorder (OCD). ^{Reference Van Oppen, Steketee, Pato and Zohar1} Many early studies showed 70–75% improvement after 15 sessions of ERP, ^{Reference Marks, Hodgson and Rachman2,Reference Emmelkamp and Kraaen3} but later studies showed positive gains from cognitive approaches aimed at reducing obsessions and rituals. Several cognitive models have been proposed for OCD, including those of Salkovskis ^{Reference Salkovskis4} and Rachman. ^{Reference Rachman5} Rachman suggests that obsessions are caused by catastrophic misinterpretation of the significance of one's thoughts as originating in normal intrusive thoughts which are interpreted in relation to responsibility beliefs, controllability of thoughts and beliefs concerning danger and threat. Danger ideation reduction therapy, for individuals with OCD with washing/contamination concerns, has been shown to be as useful as ERP. ^{Reference Jones and Menzies6–Reference Jones and Menzies8}

Despite increasing interest in the role of cognitive therapy in the treatment of OCD, there is limited evidence to suggest that cognitive therapy is superior to ERP. In a study of three groups, one receiving cognitive therapy alone, one receiving ERP alone and one receiving ERP with cognitive therapy, there were no reported differences between the groups. ^{Reference Krochmalik, Jones and Menzies9} Another study showed no difference between rational emotive therapy (analysing irrational thoughts) and exposure in vivo. ^{Reference van Oppen, de Haan, van Balkom, Spinhoven, Hoogduin and van Dyck10}

Patients with OCD respond well to individual therapy combining behavioural and cognitive approaches (cognitive–behavioural therapy, CBT) when delivered by a trained therapist. ^{Reference Emmelkemp and Beens11} Unfortunately, the supply of trained CBT therapists is limited, leading to long waiting lists for this essential psychotherapy. Despite this, few studies have examined the effectiveness of group therapy for OCD. Krone et al ^{Reference van Oppen, van Balkom, de Haan and van Dyck12} used psychoeducation about cognitive therapy and ERP approaches with 36 patients in weekly groups for 7 weeks. Yale–Brown Obsessive–Compulsive Scale (Y-BOCS) scores changed significantly from moderately severe before treatment to below clinical levels at 3-month follow-up. A further study compared three groups: group one received ERP, group two received individual ERP and group three received individual sessions of progressive muscle relaxation. This study demonstrated that, after 2 h once-weekly sessions, both individual and group ERP showed significant post-treatment improvement in OCD symptoms, depression and anxiety scores. ^{Reference Krone, Himle and Nesse13} Bouvard et al ^{Reference Fals-Stewart, Marks and Shafer14} reported that 19 patients treated with six sessions of cognitive therapy followed by six sessions of ERP had a significant improvement following the cognitive therapy, but the addition of the ERP did not improve the patients further. The study was uncontrolled and non-randomised, but the clinical impression was that cognitive therapy helped to maintain the results in the long term.

Method

As a result of the limited data available on group CBT for patients with OCD, we undertook this naturalistic study in the Department of Cognitive Behavioural Therapy, St Patrick's University Hospital, Dublin, Ireland. St Patrick's is a 300-bed facility affiliated to the University of Dublin, Trinity College, and a national referral centre for the management of mental illness. The CBT department has a particular interest in individuals with OCD.

Twenty-seven patients with OCD consecutively referred for CBT were invited to participate in group CBT as an alternative to individual treatment. Three patients declined, preferring to remain on the waiting list for individual therapy. Three of the 24 patients dropped out of the study during the course of the programme. All were out-patients and each met the DSM–IV criteria for OCD. ^{Reference Bouvard15} They were referred by general practitioners and reviewed by consultant psychiatrists. The therapists worked under the supervision of a consultant psychiatrist.

We divided the participants into four groups (five, four, eight and seven patients). Each patient underwent an initial screen for diagnosis using DSM–IV criteria and suitability for CBT (e.g. motivation for change, behavioural goals, willingness to participate in a group). Ten men and fourteen women participated.

Seventeen patients were established on a selective serotonin reuptake inhibitor prior to participation in the group. The remaining seven patients were drug-free. The patients did not have any changes to their pharmacological treatment or participate in an alternative psychotherapeutic intervention during the course of the group therapy.

All participants completed pre- and post-treatment questionnaires. The rating scales used were the Yale–Brown Obsessive–Compulsive Scale (Y-BOCS), ¹⁶ the Maudsley Obsessive–Compulsive Inventory (MOCI), ^{Reference Goodman, Price, Rasmussen, Mazure, Delgado and Heniger17} the Life Adjustment Questionnaire and Visual Analogue Problems and Goals Measure, ^{Reference Sternberger and Burns18} the Beck Depression Inventory (BDI), ^{Reference Marks and Matthew19} and the Beck Anxiety Inventory (BAI). ^{Reference Beck, Steer and Garbin20}

Each group completed ten 90-minute CBT sessions, facilitated by a trained CBT therapist (YT or CK); the first and last group were facilitated by both therapists. A consultant psychiatrist (JVL) co-facilitated one session per group. Generally, each session began with feedback from the previous session, followed by a review of the previous week, a teaching part on a planned topic, discussion within the group, a key learning point and homework. The group format is presented in Table 1.

Table 1 Summary of intervention^a,

Session	Planb
1	Introductions, purpose of group, ground rules
	Definition of OCD, CBT rationale for treatment, handouts, ratings/measures done
	Key learning points, homework for week
2	Goal-setting
	Maintenance of OCD, exposure and response prevention
	Key learning point and homework
3	Continued teaching on exposure and response prevention
	Key learning point and homework
4	Obsessive thoughts diary
	Work on responsibilities and appraisals
	Key learning point and homework
5	Specific work on thoughts; meaning, gathering evidence, use of thought diary
	Set behavioural experiment
6	Cognitive strategies: probability measures, ‘responsibility pie’ Reference Krochmalik, Jones and Menzies9
	Set homework
7	Review of goals
	Overview of CBT strategies so far
	Key learning point and homework
8	Case formulation, developing own by looking at early experiences and beliefs and rules that maintain OCD
	Key learning point and homework
9	Review of skills learned
	Blueprint of recovery
	Relapse prevention
10	Preparation for discharge
	Feedback regarding group
	Follow-up

The group initially focused on exploring the fear and avoidance model of OCD and how this contributes to maintaining OCD cycles. It then moved into evolving a formulation based on Salkovskis' ^{Reference Salkovskis4} model. The group then focused on misinterpretation of thoughts and the rose of neutralisation, i.e. engaging in overt or covert impulsive rituals which maintain the illness. The group then moved on to look at appraisals that maintain OCD and challenged the meaning clients attribute to having the thoughts in the first place. We also explored patients' tendencies to overestimate dangers and their responsibility attached to events. Through the use of thought diaries we encouraged patients to watch out for misinterpretations of the significance of their thoughts and to identify and challenge them when possible. We then requested participants to identify a key learning point at the end of each session and from this the ERP homework task was collaboratively derived. Each participant was encouraged to see how their experience fits into Salkovskis' model. At the end of therapy, participants were encouraged to design a blueprint for recovery summarising what they have learnt during the group sessions and what they need to focus on to consolidate the progress made in managing their OCD.

Statistical analysis was carried out using SPSS version 12 for Windows. Paired student's t-tests were used to compare symptom levels, using an intention-to-treat analysis (due to three non-completers), pre-treatment and post-treatment. Pearson's product moment correlation examined relationships between clinical characteristics and symptom variables.

Ethical approval for this study was obtained from the research ethics committee, St Patrick's University Hospital.

Results

Measures of illness severity pre- and post-treatment

There was a significant reduction in clinical symptom ratings on completion of group CBT for individuals with OCD (Table 1).

There were no significant correlations between age, age at onset, illness duration and clinical variables (e.g. severity of illness) in relation to treatment response (Table 2).

Table 2 Rating scales: baseline scores and outcomes

	Pre-treatment score Mean (range)	Post-treatment score Mean (range)	Significance of changea
Yale—Brown Obsessive—Compulsive Scale	24.7 (12-38)	17.1 (6-28)	t=8.4
Maudsley Obsessive—Compulsive Inventory	17 (5-26)	14.4 (5-20)	t=3.3
Beck Depression Inventory	21.9 (3-45)	13.8 (0-31)	t=5.88
Beck Anxiety Inventory	20.8 (7-43)	15.7 (3-34)	t=4.9

Discussion

Without access to effective CBT/ERP, many individuals with OCD will remain symptomatic. This disorder is not amenable to other forms of psychotherapy and it has a very low rate of placebo response. ^{Reference Beck, Epstein and Brown21} This fact alone distinguishes it therapeutically from many other psychiatric disorders.

Although it is generally accepted that this patient group needs access to CBT/ERP, ^{Reference de Veaugh-Geiss, Landau and Katz22} many patients are denied this form of treatment because of the scarcity of trained CBT therapists. It has been suggested that CBT could be applied in groups, with results equivalent to those of individual treatment, particularly when the number of sessions is comparable with that usually provided for individual patients in controlled research (i.e. 12–20 sessions). ^{Reference Van Oppen, Steketee, Pato and Zohar1} Individuals with OCD frequently postpone requests for therapeutic help for many years and when they do present to services they may be faced with a long waiting list for effective psychotherapy. Evidence confirming the effectiveness of group-based CBT/ERP is therefore welcome as an aid to increase the availability of therapy to patients without a necessary increase in resources.

We have previously reported that individuals with OCD frequently refuse to participate in individual behavioural therapy. ^{Reference Braga, Cordioli, Niederauer and Manfro23} However, the low level of drop-out from the CBT groups supports the impression that group therapy is an acceptable way of delivering CBT.