Interventions Focusing on Men's Violence Against Women in Sub-Saharan Africa to Address HIV: A Systematic Review of the Evidence

Background and Purpose: Seventy percent of newly HIV infections in the world occurred in sub-Saharan Africa [SA], most of them passed by heterosexual contacts. Existent research reveals that inequitable gender-based power and Gender-Based Violence  (GBV) are correlated with HIV rates among women in SA. Considering significant advances in antiretroviral therapy (AVT), violence against women remains a major cultural problem in SA. This research investigated the range of HIV interventions in SA focusing on men’s behavior, HIV risk including condom use, multiple partnerships, and violence against women outcomes.

Method: We reviewed 20,340 abstracts, 81 of which met our inclusion criteria for full review, resulting in 14 articles that are the current analysis. The search strategy used in the original review was rerun on 19 databases and a Cohen Kappa was calculated [k= 0.80]. Each reviewer was required to look at 81 abstracts and a>80% agreement had to be reached. The “Checklist for Measuring Quality of  Interventions” (Thomas, 2002) was used. Outcomes were compared by duration and target of the intervention and were analyzed based on statistical significance and effect size.

Results: Studies were conducted in both urban and rural settings in seven sub-Saharan African countries. The shortest intervention tested was implemented for a month and a proof-of-concept intervention consisted of two 5-hours workshops presented in same-gender groups. In contrast, long interventions were community-based; with 8-12 weeks of intervention delivery and 0.5-2 years follow up periods. Physical and/or sexual violence were the most widely measured exposure; six studies reported positive shifts in men’s report on violence although effect sizes were mixed RR≤ 1, d < 0.5 and ORs ≤  1 and  OR ≥1. Gender appeared to have an effect on outcomes in studies with male and female participants because men reported a greater magnitude of change compared to women. Studies reported generally positive or mixed results for key biological outcomes with ORs < 1.0 and d ≥0.5; however positive HIV seroconvertion was reported in one study. Four studies reported a decrease in sexual risk behavior through reduction in multiple partnerships (OR ≤1), transactional sex (OR ≤1), and sex with concurrent (OR ≤1) or casual partner (OR ≥1) while other four studies reported moderate increase in frequency of condom use with OR ≤1 and d ≤0.5.

Conclusions and Implications:  A major finding of our study is that studies that address GBV in relation to HIV are still underdeveloped. A thorough evaluation of key outcomes including the incidence and transmission process should be examined in the context of social equality, cultural gender inclusivity, and a sustained education where women and men feel that they are equal and can significantly contribute to the development of their families. Thus, considering significant advances in antiretroviral therapy (AVT), violence against women is still a big public health issue impacting HIV prevention.

Thomas, H. (2000). Quality assessment tool for quantitative studies. Effective Public Health Practice Project. Toronto: McMaster University.