What is already known on this topic

The efficacy of psychological treatments for panic disorder — without or with agoraphobia — in comparison with waitlist, psychological placebo or pharmacological placebo controls is well established.1 Also pharmacotherapy has proved to be efficacious compared with placebo.2 However, meta-analyses directly comparing psychological and pharmacological treatments tend to be outdated.3 The aim of this paper was therefore to assess the relative efficacy and acceptability of psychological and pharmacological treatments for panic disorder in adults.

Methods of the study

The authors searched the Cochrane Common Mental Disorders Group Specialised Register on 11 September 2015. Among psychological treatments, the focus was on cognitive-behavioural, psychodynamic and supportive therapies. The main pharmacotherapies comprised tricyclic antidepressants, other antidepressants, selective serotonin reuptake inhibitors, serotonin-noradrenaline reuptake inhibitors and benzodiazepines. The methodological quality of the studies was assessed according to the risk of bias. Dichotomous data were calculated as risk ratios (RRs) and their 95% CIs using a random-effects model, and continuous data as standardised mean differences.

What this paper adds

• This review provides an up-to-date evaluation of the relative short-term (within 6 months after start of treatment) efficacy of psychological therapies and pharmacotherapy.

• No indications of superior efficacy of one or the other treatment among the 16 included studies were found. In the largest comparison, there was no evidence of difference between psychological therapies and either antidepressants alone or antidepressants and benzodiazepines in terms of short-term response (RR 0.95, 95% CI 0.76 to 1.18; 12 studies; 800 participants) (low-quality evidence).

• The authors also did not find differences in acceptability measured by the number of dropouts for any reason. In the largest comparison, there was no evidence of difference between psychological therapies and either antidepressants alone or antidepressants and benzodiazepines in terms of dropout for any reason (RR 1.08, 95% CI 0.77 to 1.51; 13 studies; 909 participants) (very low-quality evidence).

Limitations

• The overall quality of the evidence was rated as very low to moderate. This means that further research is likely or very likely to have an important impact on our confidence in the estimate of effect and may change the estimate.

• The finding that medication seemed to be as acceptable as psychological interventions may be due to self-selection bias. That is, those sceptical of medication may have refused to participate in the studies, given the prospect of being randomised to medication, or withdraw after being so randomised. In one study 21% refused to take medication when allocated to that condition.4 Information about refusers was seldom reported in the individual studies and not provided in the review.

• Information about adverse effects was not provided in the review, although it was reported in some of the individual studies.4

• There were also some unfortunate errors in the Cochrane review. They claim that none of the studies had information about long-term (more than 6 months after start) outcome. However, in at least one of the studies,5 patients were followed up at least 9 months after discontinuation of medication. Moreover, the review authors reported a proportion of less than 30% women in the study of Clark et al, but the correct number is actually 78%.4

What next in research

• The long-term response, quality-of-life outcome and adverse effects of psychological therapies and pharmacotherapy should be studied.

• In particular, studies should focus on whether medication has a healing effect, that is, whether the medication can be withdrawn without being associated with relapse.

• The influence of accompanying agoraphobia should be studied.

Do these results change your practices and why?

No. The review leads to no recommendation for the choice of intervention. Thus, decisions to use pharmacotherapy or psychotherapy may primarily depend on costs, time and labour required, the availability of qualified personnel in the particular clinical setting, and the individual patient's preferences. The patients should also be informed that the longer term relative efficacy of these treatments is not known, that it is uncertain whether medication has a healing effect, and that little is known about side effects of medication. Cognitive therapy should therefore, when a choice is possible, cautiously be recommended.