Thirty studies were included for review (number of patients not reported; sample size ranged from 16 to 406). Twenty studies reported random sampling. Ten studies were rated as high quality; 19 as moderate quality and one as poor quality. Follow-up ranged from immediate post-test to more than six months.
The effect of support groups on psychological well-being, burden and social consequences:
Support groups moderately improved caregivers' mental health (g=-0.44, 95% CI -0.73 to -0.15; 19 studies, 1,779 participants, Ι²=86.6%) and depression (g=-0.40, 95% CI -0.72 to -0.08; 17 studies, participant number unclear, Ι²=86.03%). There was evidence of large statistical heterogeneity for both outcomes.
Support groups had a moderate effect on social consequences (g=0.40, 95% CI 0.09 to 0.71; 13 studies, participant number unclear) and a small effect on burden (g=-0.23 95% CI -0.33 to -0.13; 24 studies, participant number unclear). There was evidence of large statistical heterogeneity for social consequences (Ι²=82.21%) but not for burden (Ι²=12.59%).
Sensitivity analyses of psychological well-being showed a moderate effect at different time points. Sensitivity analyses of burden showed a move from small to small to moderate effect at four to six months follow up. There was no evidence of publication bias.
The effect of different variables on outcomes:
Groups that used theoretical models (QB=4.55) and that lasted more than eight weeks with more than 16 hours (QB=7.17) improved psychological well-being significantly more than shorter groups or groups that did not use a model. Significant differences were also observed for depression. Other variables associated with an increased effect size for psychological well-being or burden were reported in the article. Meta-regression analysis demonstrated that a higher proportion of female participants was associated with better outcomes for psychological well-being and depression. Older participant average age was associated with less favourable outcomes for social consequences.