Sixteen trials and 2 meta-analysis were used to evaluate treatment for severe depression.
Four trials and one meta-analysis were used to evaluate treatment for melancholic depression.
Statistical results were not generally given thus making interpretation difficult. Some results were only given in the text and were not included in the tabular presentation of study characteristics.
Severe depression in inpatients: comparison of selective serotonin re-uptake inhibitors (SSRI) vs TCA (6 trials and one meta-analysis): in 4 trials and the meta-analysis, the reduction in HAM-D score was comparable for SSRI and TCA and in 2 trials the TCA was more effective.
Mixed inpatients and outpatients (2 trials with 132 patients treated for 6 weeks and one meta-analysis): no statistically significant difference in changes in HAM-D scores in one trial and the meta-analysis. The other trial favoured SSRI to TCA.
Outpatients: 2 trials with 794 patients treated for between 6 and 8 weeks: compared SSRI, TCA and placebo: inconsistent results with one trial reporting greater improvement with SSRI compared to TCA and placebo and the other reporting SSRI to be equal to TCA with both being better than placebo (P < 0.05).
Melancholic patients (3 trials and one meta-analysis evaluated SSRI): some patients appear to have had HAM-D scores of less than the 25 required to diagnose severe depression.
Tolerability of SSRI s vs TCAs in severe depression: a number of adverse anticholinergic and cardiovascular effects were significantly more common with TCA than SSRIs in other trials no significant difference in adverse effects was reported. No supporting data was presented.
Venlafaxine: one study with 67 patients compared venlafaxine with fluoxetine in hospitalised patients with severe depression over 6 weeks and reported results as changes in Montgomery-Asberg Depression Rating scale. One study with 93 patients compared venlafaxine with placebo in melancholic depression over 4 weeks and reported results as changes in Montgomery-Asberg Depression Rating scale. Both studies favoured venlafaxine. No data was given for changes in HAM-D.
Mirtazepine: one study with 133 patients compared mirtazepine with fluoxetine and reported no significant difference at 6 weeks. One study with 150 patients compared mirtazepine, TCA and placebo and reported both active treatments to be significantly better than placebo at 6 weeks.
Nefazodone: one study with 120 patients compared nefazodone with placebo over 6 weeks and reported nefazodone to be superior to placebo.
Bupropion 2 studies with 141 patients compared bupropion with placebo over 4 weeks and reported bupropion to be significantly superior to placebo.
Combination therapy for severe depression: one retrospective uncontrolled trial included 3 patients considered to be treatment resistant.
Electroconvulsive therapy: one study was mentioned. No details were given.