The psychological and pharmacological treatment of panic disorder and agoraphobia in primary care
Following a review of treatment outcome study methodology, a comparative study of psychological versus pharmacological treatments was conducted; subsidiary studies investigated aspects of treatment outcome in more detail. 193 patients with DSM III-R panic disorder with or without agoraphobia were randomly allocated to; fluvoxamine, placebo, fluvoxamine + CBT (cognitive behaviour therapy), placebo + CBT, or CBT alone. Patients received no concurrent treatments and were treated to the same schedule, with therapist contact balanced across groups. Treatments were conducted in the primary care setting. Outcome at treatment end-point and 6 month follow-up, assessed in terms of both statistical and clinical significance, showed patients receiving active treatments improved significantly, with improvement better preserved over follow-up in the groups receiving CBT. The CBT alone and fluvoxamine + CBT groups showed the most consistent gains, the latter group showing gains earliest in treatment. Outcome was also investigated using brief global ratings of symptom severity, change in symptoms following treatment, general wellbeing and social disruption, completed by psychologist, referring GPs, and patients. Using these measures all active treatments showed statistical advantage over placebo with the groups employing CBT showing the most robust and consistent response. Overall there were no significant differences in drop-out rates between groups although the drop-out rate for patients receiving CBT alone was higher than that for placebo + CBT. Agreement with main outcome measures was demonstrated for psychologist and patient ratings, but not for GP ratings. An investigation of panic attack variables as treatment outcome measures indicated that these did not function as discriminative treatment outcome measures with all treatment groups showing significant reductions in panic attack variables over treatment with few significant differences between treatment groups on any variable throughout treatment. An investigation of prognostic indicators of treatment outcome indicated good prediction of post treatment response using pre-treatment measures of anxiety level, frequency of panic attacks, extroversion and treatment group. Predictions of outcome at 6 month follow-up were less robust. Results are discussed in terms of their relevance to wider clinical practice.