Deborah K. Brokaw, PharmD

ABSTRACT

BACKGROUND: Selective serotonin reuptake inhibitors (SSRIs) are commonly used as first-line treatment for panic disorder. However, comparative efficacy trials are lacking between older antidepressants, specifically the tricyclic antidepressants (TCAs), and SSRIs in the treatment of panic disorder. The authors use data gathered from efficacy trials to compare the efficacy, safety, and tolerability of SSRIs and TCAs used in the treatment of panic disorder.

POPULATION STUDIED: This meta-analysis included double-blind, placebo-controlled efficacy trials of SSRIs for panic disorder in patients with or without agoraphobia. The trials that met these established criteria were published from 1990 to 1998 and included 1741 patients (mean sample size: 145 patients). A comparison between the study populations could not be made since the trials did not contain patient demographic information for the SSRI and non-SSRI groups. This analysis excluded uncontrolled trials, case reports, and long-term or follow-up trials.

STUDY DESIGN AND VALIDITY: The authors used MEDLINE, PsychLIT, discussions with colleagues, and reference sections from related articles to identify double-blind, placebo-controlled efficacy trials of SSRIs for panic disorder. The authors conducted an effect-size analysis on the 12 trials identified. They compared these findings with the results of a recently published meta-analysis using non-SSRI treatments for panic disorder. In the fixed-dose trials, only the effective doses of SSRIs were used in the calculation of effect sizes.

OUTCOMES MEASURED: The main outcome is the effect sizes of the SSRI and TCA groups. The authors calculated the effect size by subtracting the mean score of the post treatment comparison group from that of the post treatment active treatment group and then dividing by the standard deviation of the post treatment comparison group. Tolerability was assessed by using the dropout rates for each study group.

RESULTS: The mean effect size for acute treatment outcome in the SSRI group compared with placebo was 0.55, a number not significantly different from that of the non-SSRI group (0.55) or, more specifically, the imipramine group (0.48). The older but smaller SSRI trials were associated with larger treatment effect sizes, whereas the larger, more recently published SSRI trials showed a smaller benefit. In addition, a funnel plot analysis showed that smaller studies with a lower effect size were missing (publication bias against ”negative” studies). The difference in the dropout rates between groups treated with SSRIs (24.6%), which were weighted to give larger trials a greater contribution, was not significantly different from that of the other antidepressants (25.4%), specifically impramine (22.4%). Using dropout rates as the only measure of tolerability may not be optimal. Not every patient who experienced adverse effects to the drug dropped out of the study. Patients may have also dropped out of the study for reasons other than poor tolerability to the drug.

ABSTRACT

BACKGROUND: Selective serotonin reuptake inhibitors (SSRIs) are commonly used as first-line treatment for panic disorder. However, comparative efficacy trials are lacking between older antidepressants, specifically the tricyclic antidepressants (TCAs), and SSRIs in the treatment of panic disorder. The authors use data gathered from efficacy trials to compare the efficacy, safety, and tolerability of SSRIs and TCAs used in the treatment of panic disorder.

POPULATION STUDIED: This meta-analysis included double-blind, placebo-controlled efficacy trials of SSRIs for panic disorder in patients with or without agoraphobia. The trials that met these established criteria were published from 1990 to 1998 and included 1741 patients (mean sample size: 145 patients). A comparison between the study populations could not be made since the trials did not contain patient demographic information for the SSRI and non-SSRI groups. This analysis excluded uncontrolled trials, case reports, and long-term or follow-up trials.

STUDY DESIGN AND VALIDITY: The authors used MEDLINE, PsychLIT, discussions with colleagues, and reference sections from related articles to identify double-blind, placebo-controlled efficacy trials of SSRIs for panic disorder. The authors conducted an effect-size analysis on the 12 trials identified. They compared these findings with the results of a recently published meta-analysis using non-SSRI treatments for panic disorder. In the fixed-dose trials, only the effective doses of SSRIs were used in the calculation of effect sizes.

OUTCOMES MEASURED: The main outcome is the effect sizes of the SSRI and TCA groups. The authors calculated the effect size by subtracting the mean score of the post treatment comparison group from that of the post treatment active treatment group and then dividing by the standard deviation of the post treatment comparison group. Tolerability was assessed by using the dropout rates for each study group.

RESULTS: The mean effect size for acute treatment outcome in the SSRI group compared with placebo was 0.55, a number not significantly different from that of the non-SSRI group (0.55) or, more specifically, the imipramine group (0.48). The older but smaller SSRI trials were associated with larger treatment effect sizes, whereas the larger, more recently published SSRI trials showed a smaller benefit. In addition, a funnel plot analysis showed that smaller studies with a lower effect size were missing (publication bias against ”negative” studies). The difference in the dropout rates between groups treated with SSRIs (24.6%), which were weighted to give larger trials a greater contribution, was not significantly different from that of the other antidepressants (25.4%), specifically impramine (22.4%). Using dropout rates as the only measure of tolerability may not be optimal. Not every patient who experienced adverse effects to the drug dropped out of the study. Patients may have also dropped out of the study for reasons other than poor tolerability to the drug.

ABSTRACT

BACKGROUND: Selective serotonin reuptake inhibitors (SSRIs) are commonly used as first-line treatment for panic disorder. However, comparative efficacy trials are lacking between older antidepressants, specifically the tricyclic antidepressants (TCAs), and SSRIs in the treatment of panic disorder. The authors use data gathered from efficacy trials to compare the efficacy, safety, and tolerability of SSRIs and TCAs used in the treatment of panic disorder.

POPULATION STUDIED: This meta-analysis included double-blind, placebo-controlled efficacy trials of SSRIs for panic disorder in patients with or without agoraphobia. The trials that met these established criteria were published from 1990 to 1998 and included 1741 patients (mean sample size: 145 patients). A comparison between the study populations could not be made since the trials did not contain patient demographic information for the SSRI and non-SSRI groups. This analysis excluded uncontrolled trials, case reports, and long-term or follow-up trials.

STUDY DESIGN AND VALIDITY: The authors used MEDLINE, PsychLIT, discussions with colleagues, and reference sections from related articles to identify double-blind, placebo-controlled efficacy trials of SSRIs for panic disorder. The authors conducted an effect-size analysis on the 12 trials identified. They compared these findings with the results of a recently published meta-analysis using non-SSRI treatments for panic disorder. In the fixed-dose trials, only the effective doses of SSRIs were used in the calculation of effect sizes.

OUTCOMES MEASURED: The main outcome is the effect sizes of the SSRI and TCA groups. The authors calculated the effect size by subtracting the mean score of the post treatment comparison group from that of the post treatment active treatment group and then dividing by the standard deviation of the post treatment comparison group. Tolerability was assessed by using the dropout rates for each study group.

RESULTS: The mean effect size for acute treatment outcome in the SSRI group compared with placebo was 0.55, a number not significantly different from that of the non-SSRI group (0.55) or, more specifically, the imipramine group (0.48). The older but smaller SSRI trials were associated with larger treatment effect sizes, whereas the larger, more recently published SSRI trials showed a smaller benefit. In addition, a funnel plot analysis showed that smaller studies with a lower effect size were missing (publication bias against ”negative” studies). The difference in the dropout rates between groups treated with SSRIs (24.6%), which were weighted to give larger trials a greater contribution, was not significantly different from that of the other antidepressants (25.4%), specifically impramine (22.4%). Using dropout rates as the only measure of tolerability may not be optimal. Not every patient who experienced adverse effects to the drug dropped out of the study. Patients may have also dropped out of the study for reasons other than poor tolerability to the drug.