BACKGROUND:
Due to intimate partner violence (IPV), a proportion of some women seek safety from their abuser in shelters. Little is known yet about whether the received shelter interventions, offered during and after women's stay, are effective. Therefore, a meta-analysis of studies was performed, testing the effectiveness of interventions administered to female IPV victims during and/or after their residence in shelters, in terms of mental health, re-abuse and social outcomes.

METHODS:
From January 1985 through July 2011, five databases were searched for English-language articles. Original research articles evaluating the effects of interventions provided to IPV victims during and after shelter residence were identified. Hedges' g effect sizes and 95% confidence intervals (CIs) were calculated and pooled if three or more studies including one outcome type were available.

RESULTS:
Ten original studies, including a total number of 726 female IPV victims, were analysed. We found that shelter interventions were effective in improving mental health outcomes (g = -0.39; 95% CI: 0.24-0.54), in decreasing abuse (g = 0.32; 95% CI: 0.08-0.55) and in improving social outcomes (g = 0.71; 95% CI: 0.54-0.88) in shelter-based abused women.

CONCLUSION:
This analysis suggests that interventions provided during and after stay in a shelter are effective in improving mental health, abuse and social outcomes, but further research has to confirm this. Moreover, future studies should compare different type of interventions and should evaluate cost-effectiveness.