An effect-size analysis of the relative efficacy and tolerability of serotonin selective reuptake inhibitors for panic disorder

CRD COMMENTARY: The review question was stated. Some study selection criteria were also stated, although these were only clear for the intervention and study design. The dates over which the literature search was conducted were not provided, and details of the included studies (participants and outcome measures) were not given. No attempt was made to search for unpublished studies; the results of the funnel plot analysis indicate possible publication bias. The pooled effect size approach may have been appropriate but it is difficult to tell without seeing what the original outcome measures were. No measures of variance of the pooled effect sizes were given, so it is unclear whether the effect size for SSRIs versus placebo was statistically significant. The review's authors then undertook an indirect comparison with effect sizes from a previous meta-analysis of other drugs versus placebo, and compared the effect sizes using t-tests. This was not an appropriate approach as it uses indirect comparisons rather than direct comparisons from RCTs. Thus, the authors' conclusions, which are drawn from this final comparison, should be treated with great caution.