Technology-based and digital interventions for intimate partner violence: A systematic review and meta-analysis.

Abstract

A growing body of research shows the promise and efficacy of technology-based or digital interventions in improving the health and well-being of survivors of intimate partner violence (IPV). In addition, mental health comorbidities such as anxiety, post-traumatic stress disorder (PTSD), and depression occur three to five times more frequently in survivors of IPV than non-survivors, making these comorbidities prominent targets of technology-based interventions. Still, research on the long-term effectiveness of these interventions in reducing IPV victimization and adverse mental health effects is emergent. The significant increase in the number of trials studying technology-based therapies on IPV-related outcomes has allowed us to quantify the effectiveness of such interventions for mental health and victimization outcomes in survivors. This meta-analysis and systematic review provide critical insight from several randomized controlled trials (RCTs) on the overall short and long-term impact of technology-based interventions on the health and well-being of female IPV survivors. To synthesize current evidence on the effects of technology-based or digital interventions on mental health outcomes (depression, anxiety, and PTSD) and victimization outcomes (physical, psychological, and sexual abuse) among IPV survivors. We examined multiple traditional and grey databases for studies published from 2007 to 2021. Traditional databases (such as PubMed Central, Web of Science, CINAHL Plus, and PsychINFO) and grey databases were searched between April 2019 and February 2021. In addition, we searched clinical trial registries, government repositories, and reference lists. Authors were contacted where additional data was needed. We identified 3210 studies in traditional databases and 1257 from grey literature. Over 2198 studies were determined to be duplicates and eliminated, leaving 64 studies after screening titles and abstracts. Finally, 17 RCTs were retained for meta-analysis. A pre-registered protocol was developed and published before conducting this meta-analysis. We included RCTs targeting depression, anxiety, PTSD outcomes, and victimization outcomes (physical, sexual, and psychological violence) among IPV survivors using a technology-based intervention. Eligible RCTs featured a well-defined control group. There were no study restrictions based on participant gender, study setting, or follow-up duration. Included studies additionally supplied outcome data for calculating effect sizes for our desired outcome. Studies were available in full text and published between 2007 and 2021 in English. We extracted relevant data and coded eligible studies. Using Cochrane's RevMan software, summary effect sizes (Outcome by Time) were assessed using an independent fixed-effects model. Standardized mean difference (SMD) effect sizes (or Cohen's d) were evaluated using a Type I error rate and an alpha of 0.05. The overall intervention effects were analyzed using the Z-statistic with a p-value of 0.05. Cochran's Q test and Higgins' I 2 statistics were utilized to evaluate and confirm the heterogeneity of each cumulative effect size. The Cochrane risk of bias assessment for randomized trials (RoB 2) was used to assess the quality of the studies. Campbell Systematic Reviews registered and published this study's protocol in January 2021. No exploratory moderator analysis was conducted; however, we report our findings with and without outlier studies in each meta-analysis. Pooled results from 17 RCTs yielded 18 individual effect size comparisons among 4590 survivors (all females). Survivors included college students, married couples, substance-using women in community prisons, pregnant women, and non-English speakers, and sample sizes ranged from 15 to 672. Survivors' ages ranged from 19 to 41.5 years. Twelve RCTs were conducted in the United States and one in Canada, New Zealand, China (People's Republic of), Kenya, and Australia. The results of this meta-analysis found that technology-based interventions significantly reduced depression among female IPV survivors at 0-3 months only (SMD = -0.08, 95% confidence interval [CI] = -0.17 to -0.00), anxiety among IPV survivors at 0-3 months (SMD = -0.27, 95% CI = -0.42 to -0.13, p = 0.00, I 2 = 25%), and physical violence victimization among IPV survivors at 0-6 months (SMD = -0.22, 95% CI = -0.38 to -0.05). We found significant reductions in psychological violence victimization at 0-6 months (SMD = -0.34, 95% CI = -0.47 to -0.20) and at >6 months (SMD = -0.29, 95% CI = -0.39 to -0.18); however, at both time points, there were outlier studies. At no time point did digital interventions significantly reduce PTSD (SMD = -0.04, 95% CI = -0.14 to 0.06, p = .46, I 2 = 0%), or sexual violence victimization (SMD = -0.02, 95% CI = -0.14 to 0.11, I 2 = 21%) among female IPV survivors for all. With outlier studies removed from our analysis, all summary effect sizes were small, and this small number of comparisons prevented moderator analyses. The results of this meta-analysis are promising. Our findings highlight the effectiveness of IPV-mitigating digital intervention as an add-on (not a replacement) to traditional modalities using a coordinated response strategy. Our findings contribute to the current understanding of "what works" to promote survivors' mental health, safety, and well-being. Future research could advance the science by identifying active intervention ingredients, mapping out intervention principles/mechanisms of action, best modes of delivery, adequate dosage levels using the treatment intensity matching process, and guidelines to increase feasibility and acceptability.