In this randomized intervention study with rural Ivorian women, the addition of a dialogue component explicitly addressing gendered social inequalities and norms alongside economic empowerment programming significantly reduced past year physical IPV among women who participated in more than 75% of the program with their male partner. The combined intervention also significantly reduced economic abuse and altered attitudes regarding the justification and acceptance of IPV among all women in the study. Encouragingly, reductions in IPV were also observed in ITT analyses, though decreases were not significant and the effect size for the overall physical and/or sexual IPV was small. Importantly, this research demonstrates that IPV reduction programs can be rigorously evaluated in a conflict-affected setting—even in the midst of a period of heightened post-election violence, and, moreover, that it is possible to observe reductions in this under-addressed yet very prevalent form of violence in such challenging and unstable contexts.
The current RCT findings are broadly consistent with the one prior South Africa-based study (the IMAGE intervention) that examined the impact of combining gender equity components with economic empowerment programming on IPV, [11, 16] offering further strength to the evidence that the inclusion of gender training to socio-economic programming can offer social and health benefits, and also extending such findings to a conflict-impacted region. However, it should be noted that the current intervention possessed unique features that included activities that specifically involved male partners and the use of a financial lens with couples to talk about gendered power dynamics in the home. Moreover, the current Cote d’Ivoire intervention was a shorter intervention than the IMAGE trial intervention, which also included a larger community mobilization component.
There are also differences regarding study design. In this study, we assessed the incremental impact of adding gender equity components onto a VSLA program, while the IMAGE intervention assessed the impact of a combined micro-credit and gender training program compared with no intervention at all. In the current study, we saw significant changes among those women who were highly adherent to the intervention (or who achieved high intervention exposure), while in IMAGE, the impact was significant among all participants. Secondary analyses of IMAGE data suggest that the gender training component of the intervention was critical to its success . Future research is needed to identify exact pathways of change for the current intervention. Taken together, findings illustrate the potential benefits of adding gender sensitization components onto livelihood programs for women in both conflict affected and non-conflict affected settings.
Study findings must be interpreted within the context of limitations. Firstly, as with most stigmatized health issues, IPV self-reporting may be subject to social desirability bias. Also, prior research has suggested that participants in IPV reduction interventions may over-report IPV due to increased awareness . However, the directionality of such bias is difficult to determine as it is unclear whether one arm would be more or less likely to under or over report; biases away from the null would be present if over-reporting was only present in the treatment arm. Future research in post-conflict settings that includes similar quantitative assessments of IPV perpetration and gender attitudes among men would strengthen understanding of the impacts of the type of programming that was evaluated herein. Second, participants, facilitators, and researchers were not blinded to treatment status. Given that both treatment and control groups took place in the same villages, there is a chance of contamination, and thus a bias towards null findings. However, pre-study consultations with community leaders suggested that the use of a waitlist control design in large villages would minimize such threats. No crossover was observed between treatment arms per administrative records. Also, in eight villages with a smaller population, groups were only randomized to one arm due to contamination concerns. Regarding external validity, while a community-based sample was recruited, participants may differ from women who chose not to participate in the investigation.
The study was likely to be underpowered as certain villages had more widowed participants than anticipated and because of our inability to mobilize as many villages as expected, thus reducing the analytic sample. Moreover, given the relatively scarce number of interventions conducted to date that focus on the program components and the populations in the current study, findings reported herein are preliminary in nature. Additionally, since some of the villages were inaccessible during a period of post-election violence in 2010, regular fidelity monitoring of VSLA activities was not possible. However, in anticipation of possible post-election violence, the IRC team conducted training prior to the start of activities. Attendance records indicate regular meetings, including meeting at undisclosed locations during threats of intense violence. While widespread violence may have influenced the physical mobility of our study population, which may in turn have affected whether couples stayed together physically during times of insecurity, post-hoc analyses indicated that cohabitation of partners in the year preceding the endline survey did not significantly differ by treatment group, adherent group, or reporting of IPV. Although financial issues/lack of confidence in VSLA activities were frequently cited as reasons for leaving the program among those who dropped out, it is unclear if this fully explains why drop-outs differed by treatment status given that all women received VSLA simultaneously. A related potential limitation is that we were unable to assess how the success of VSLA activities influenced overall group dynamics or effectiveness of the GDGs, as the incremental effects of GDGs may be correlated with the success of the economic component of the intervention. Notably, attrition was not significantly related to IPV at baseline or endline, nor were groups with high levels of drop-outs collapsed, which minimized contamination concerns. Moreover, due to post-election violence, the start of the GDGs were delayed and could not be completed as of August 2011 as planned initially. Thus, the past year assessment of IPV is inclusive of 8 months in which the GDG was being delivered and it is unclear if the GDGs, in the context of economic empowerment programming, would have been able to reduce IPV in this shortened period. In addition, while per protocol analyses assessed adherence to the GDG sessions, we do not know which sessions were skipped. It can be argued that the sessions regarding financial stress and household economy may have more directly addressed IPV and inequitable gender norms than other sessions. Therefore, non-attendance to these sessions may have reduced the impact of the GDG component. Notably, while not all sessions were developed to explicitly discuss IPV, reports from facilitators indicate that the topic of IPV was spontaneously discussed by groups at each session. As we were able to assess adherence to the GDG intervention and impacts on summary measures of different forms of IPV, we could not examine the severity or frequency of IPV events in our data in accordance with a dose–response relationship.
These limitations notwithstanding, the current RCT has important strengths. It was done in partnership with a non-governmental organization with a long working history in Côte d’Ivoire, and incorporated specific components to maintain positive relationships with the community (e.g. inclusion of non-partnered women, non-use of a non-interventional control group due to voiced ethical concerns). Moreover, despite the use of a comparison arm that received only the economic intervention (versus a pure control), reductions in IPV and changes in attitudes towards justification of IPV were observed. All responses were prospectively assessed, and loss to follow-up was minimal despite ongoing post-election violence. Finally, although not all results reached statistical significance, the findings from both ITT and PP analyses are consistent with one another. As argued by other violence researchers and community interventionists, for complex interventions such as the one evaluated in Côte d’Ivoire, consistency and directionality of findings, in additional to statistical significance, are important components to consider [15, 30, 34]. The significant finding regarding improvement in attitudes towards justification of IPV is particularly important, as attitudinal changes may be proxies for norms, which may precede changes in IPV ; longer-term research is needed to determine this theory of change. Further, trends in IPV reduction are encouraging, given the overlap of follow-up period and delayed intervention delivery. Related to the short follow-up time, the effect size for the overall outcome measure (physical and/or sexual IPV) may have been small due to overlap between women who reported both sexual IPV and physical IPV. This overlap, combined with the possibility that it may be more feasible to influence physical IPV levels in shorter time frames compared to sexual IPV, may in part explain why effect sizes for the overall outcome (physical and/or sexual IPV) were small and did not reach significance. Longer follow-up time is needed for future work. Findings that attitudes related to the justification of physical IPV were significantly improved, but not sexual IPV further suggest the need for future research to investigate the relative difficulty of reducing sexual IPV in comparison to physical IPV. In addition, while significant reductions in sexual IPV were not found, it is possible that the GDGs may have prevented increases in sexual IPV.