A controlled one year study of a cognitive behavioural group therapy programme in the treatment of obsessive compulsive disorder
Cognitive Behavioural Therapy (CBT) is widely acknowledged to be an efficacious intervention in Obsessive Compulsive Disorder (OCD), although resources, in terms of therapists to carry out CBT on an individual basis, are scarce, and can incurr long waiting lists. A timely question therefore, was whether CBT could be effective in a group setting. An uncontrolled pilot study had generated promising data on the efficacy of group CBT in OCD and this controlled study aimed to assess any additional benefits of group CBT over and above those of group interaction per se. OCD patients were randomised to either (i) a group Cognitive Behavioural Therapy (CBT) programme or (ii) a group relaxation therapy (RT)programme. Individual relaxation therapy is well known to be a neutral treatment for OCD and this control condition, therefore, provided baseline data on the expected response rate in a group setting without active psychotherapy. 26 Patients were screened initially to comply with inclusion/exclusion criteria and randomised to each treatment arm. Treatment under both conditions comprised of 12 weekly two hour group sessions; in addition patients were seen individually once before treatment began (i. e. for screening) and post treatment (i. e. after 12 weeks for debriefing and feedback). Patients entering the study were assessed throughout by raters blinded to the patient's treatment. The principal outcome measure was the total score on the Yale- Brown Obsessive Compulsive Scale (Y-BOCS), and this was measured prior to treatment at baseline and then at fortnightly intervals. It was expected that group CBT patients would demonstrate significantly greater reduction in pre-treatment scores. The data was examined using a repeated measures analysis of variance: the repeated measures factor was `time' with 2 levels and the between subjects factor was `therapy group', also with two levels (CBT & RT). To adjust for differing baseline scores, a covariate `baseline' was also entered into the analysis. The main effect of time was found to be reliable (F(1,18) = 12.84, p< 0.01). The effect of group (F(1,17= 0.4, ns) and the group by time interaction (F(1,17)= 0.03, ns) were not statistically reliable. On the other hand there was a significant negative correlation between Duration of OCD and change in Y-BOCS scores in the CBT group (R= 0.72, p=0.006), whilst in the RT group there was a positive correlation (R=0.74, P= 0.096), approaching significance. This shows that the magnitude of change in Y BOCS scores was greater in those patients who had shorter duration of OCD in the CBT group; whilst in the RT group the patients with the longer Duration of OCD showed greater magnitude of change in Y-BOCS scores. Although the RT group results in this respect should be treated with some caution given the small number of completers (6) as compared to CBT completers (14). However this has raised an interesting factor and would be worth following up in future studies. Also, a secondary outcome measure, the Clinical Global Impression (CGI) scores (A 7 point scale measuring severity of illness) showed a reliable group by time interaction (F(1,16), = 10.5, p= 0.005) in that there was a greater decrease over time in CGI scores, in the CBT group than the RT group.