Reply

In their letter Biondi and Picardi [1] provide some data that also suggest that attribution of improvement to medication is a risk factor for relapse of panic disorder after medication is tapered off. The main limitation of their findings is that they only evaluated patients who received ‘integrated treatment’: a combination of medication and cognitive-behavioral therapy. They have also demonstrated superiority of combined treatment over pharmacotherapy alone, especially with respect to the number of patients that relapse after cessation of therapy [2]. Our group demonstrated superiority of combined pharmacotherapy (antidepressants) and psychotherapy (exposure in vivo) in a meta-analysis of both short-term and long-term efficacy studies [3, 4]. However, our research on the locus of control orientation did not include patients who received combination treatment [5]. Therefore, the effects of pharmacotherapy alone versus a combination treatment probably have to be studied in one design before strong conclusions can be drawn. Also, the study of Basoglu et al. [6], to which they also refer, does not give data with respect to this. There are some additional problems that we would like to address: it is unclear if different drugs (e.g. antidepressants or benzodiazepines) foster different attributions. Since it takes only a short time before the effects of benzodiazepines (e.g. alprazolam) can be experienced, it may be that patients attribute improvement much more to the effects of treatment with this type of drug than in the case of medications like antidepressants, which work only after a couple of weeks. Another factor that may have confounded the findings of Biondi and Picardi [1] is the duration of drug treatment: from their data it is unclear how long the medications were taken. Since antidepressants have to be taken for long periods of time, a change in attributions may also take place months after the medication was started. Therefore, it would be of interest to know whether relapse could be related to other factors, e.g. the duration of medical treatment or the type of medication, in the study of Biondi and Picardi [1]. The aforementioned study by Basoglu et al. [6] included only alprazolam, whereas Biondi and Picardi [2] mainly used antidepressants. Differences between both groups of drugs therefore may have been missed so far. Finally, the rationale that is used for the medication in a combination treatment is probably a very strong predictor for the chance of relapse as well, and can easily be modified in the preferred direction (‘rely on yourself, medications are only supportive’). Both papers lack information with respect to the rationale that was provided at the beginning of a treatment with both medication and cognitivebehavioral therapy [2, 6].