Four randomised controlled trials were included (n=718): three rated high quality and one rated low methodological quality (due to lack of blinding).
The results indicated that the interventions were effective in improving the mental health of informal stroke caregivers in comparison with control. The overall mean weighted effect size was 0.277 (95% confidence interval: 0.118, 0.435, p<0.001, four randomised controlled trials, n=718), for educational programmes only an mean weighted effect size 0.354 (95% CI: 0.087, 0.621, p<0.01, two randomised controlled trials, n=239), and for support programmes only an mean weighted effect size of 0.234 (95% confidence interval: 0.037, 0.432, p=0.02, two randomised controlled trials, n=479). There was evidence of statistical heterogeneity between studies (p=0.05).
Sensitivity analysis based on study validity rating indicated that interventions in high quality studies were significant in improving mental health of stroke caregivers (mean weighted effect size 0.354, 95% confidence interval: 0.175, 0.534, p<0.001, three randomised controlled trials. There were no statistically significant differences between intervention and control groups in the one low-quality study. There was evidence of publication bias (Fail-safe N=13).