Forty-eight studies (n=2,218) were included in the review, of which 18 had an untreated control group.
Forty-five of the 48 studies concluded that group psychotherapy was effective. Forty-three of the 46 studies presenting adequate data found that group therapy significantly reduced depression.
Studies with an untreated control group (18 studies): all of these studies found that depression decreased in both the treatment and the control groups. Fourteen studies found that group therapy significantly reduced depression compared with delayed control treatment. Four studies found no significant difference between the treatments.
Group therapy versus no treatment (15 studies reporting adequate data): compared with no treatment, group therapy significantly reduced depression at post-treatment and at follow-up; the ES post-treatment was 1.026 (range: -0.07 to 2.30; 95% CI: 0.565, 1.487); the ES at follow-up (10 studies) was 1.178 (95% CI: 0.586, 1.764).
Clinical significance of change: at baseline, patients allocated to group therapy were moderately depressed according to Robinson norms for BDI scores. Patients treated with group therapy still had higher BDI scores post-treatment than the general population (P<0.001). Group therapy increased the proportion of patients with clinically meaningful improvement post-treatment in comparison with no treatment (9 studies, 356 patients), 48.2% versus 18.5%.
Group versus individual therapy (9 studies): the results differed across the studies. Five studies favoured individual therapy and 4 studies favoured group therapy. There was no statistically-significant difference in ES between group and individual therapy across the 5 studies presenting adequate data; the ES was 0.15 (CIs included zero, but no values were given).
Cognitive-behavioural therapy versus psychodynamic therapy (8 studies): the benefit of cognitive-behavioural therapy over psychodynamic therapy just reached statistical significance; the ES was 0.295 (95% CI: 0.001, 0.589).
Higher quality studies (8 studies meeting two quality criteria): compared with no treatment, group therapy significantly reduced depression in higher quality studies; the ES was 0.94 (95% CI: 0.17, 1.70)
Research versus clinical setting: there was no statistically- significant difference in ES between treatments conducted in research and clinical settings; the mean decrease from baseline to post-treatment was 51.0% (median 53.7%) in research settings (16 studies) versus 44.8% (median 41.7%) in clinical settings (18 studies). There was no significant difference between locations in baseline BDI scores: 25.3 for research versus 23.3 for clinical (P>0.10).
The drop-out rates ranged from 0 to 63.3% across all studies.