Addressing Intimate Partner Violence and Power in Intimate Relationships in HIV Testing Services in Nairobi, Kenya

Intimate partner violence (IPV) undermines women’s uptake of HIV services and violates their human rights. In a two-arm randomized controlled trial we evaluated a short intervention that went a step beyond IPV screening to discuss violence and power with women receiving HIV testing services during antenatal care (ANC). The intervention included training and support for HIV counselors, a take-home card for clients, and an on-site IPV counselor. One third (35%) of women (N = 688) reported experiencing IPV in the past year; 6% were living with HIV. Among women experiencing IPV, program participants were more likely to disclose violence to their counselor than women receiving standard care (32% vs. 7%, p < 0.001). At second ANC visit, intervention group women were significantly more likely to report that talking with their counselor made a positive difference (aOR 2.9; 95% CI 1.8, 4.4; p < 0.001) and felt more confident in how they deserved to be treated (aOR 2.7; 95% CI 1.7, 4.4; p < 0.001). Exploratory analyses of intent to use ARVs to prevent mother-to-child transmission and actions to address violence were also encouraging. The study took place in Kenyatta National Hospital’s (KNH’s) ANC clinic in Nairobi, Kenya, from February 2015 to August 2015. KNH is a teaching hospital and the oldest and largest public referral hospital in the East Africa region, with KNH’s ANC clinic receiving upwards of 1000 clients per month. All ANC clients at KNH receive HTS as part of their first ANC visit of their current pregnancy. Women can choose to opt out of HIV testing. While IPV screening is supposed to be part of standard HTS counseling in the ANC clinic, it is inconsistently implemented. Referrals for IPV are to KNH’s Gender-Based Violence Recovery Centre (GBV Centre), located within the KNH complex, about 5 min walk from the ANC clinic [18]. The GBV Centre provides comprehensive post-rape care that includes emergency care such as post-exposure prophylaxis (PEP); treatment for sexually transmitted infections (STIs); trauma counseling; and psychological support via individual counseling, support groups, and connecting women to social workers. The Centre also arranges referrals to safe houses if the client’s home environment is unsafe, to police, and to organizations that offer legal services. For HIV-negative mothers, standard HTS counseling in the ANC clinic consists of discussing a birth plan, couple counseling if accompanied by a partner, information on danger signs during pregnancy, importance of attending scheduled ANC visits, maternal nutrition, STI risk reduction, postpartum family planning, hygiene and self-care, infant feeding, and male partner engagement in ANC/childcare. Women are also supposed to be screened for IPV and referred to the GBV Centre if they are experiencing violence. Mothers with HIV, in addition to the preceding, receive counseling on CD4 counts, ART, importance of adherence, and specialized care such as mother-to-mother groups and nutrition clinics. The pilot intervention was implemented from March to July 2015 and included four main parts. The first component was a provider training. All providers in the ANC clinic participated in a 1-day off-site training on GBV. While providers had previous training at KNH in IPV screening and care, this training provided a refresher and an orientation to the intervention and study. The training included participatory activities—for example, myths and facts about IPV, and how relationship inequalities affect women’s well-being—and evidence on the impact of IPV and power inequalities in relationships on HIV prevention and care. The HTS counselors also received hands-on practice conducting the intervention. The 28 HTS counselors—25 females and 3 males—included lay counselors with between 2 and 10 years of experience, and nurses with between 4 and 20 years of experience. The second intervention component included counseling aids for providers to use during the post-test counseling session. One was a small tri-fold card with key messages and resources regarding IPV, power, and women’s health. Cards developed by Futures Without Violence [19] for clinical settings in the U.S. were adapted by the study team to meet the objectives of this study and setting. The cards included a panel on facts, such as prevalence of IPV in Nairobi, and that being in a relationship characterized by IPV or low power for the woman increases her risk for HIV and STIs and can harm her baby’s health. Another panel posed reflective questions about the nature of the relationship—for example, is my partner’s communication honest and open? Who is mostly responsible for making decisions? Does my partner value my opinions and respect my choices? Is my partner respectful and kind to me? Relationship control and IPV were the focus of a third panel and asked about controlling behavior, physical and sexual abuse, and whether the woman was afraid to ask about condom use or tell her partner she has an STI or HIV. A highlighted panel included messages to remember such as, “You matter. You have a right to be free from violence,” “You are not to blame if you are experiencing any form of abuse,” and “There are people here who can help you.” Two additional panels covered possible things that a woman could do if she is being hurt or feels powerless, and provided contact information for four places to turn to for help in Nairobi: The GBV Centre at KNH, the Women’s Rights Awareness Programme (which provides safe shelter), Nairobi Women’s Hospital Gender Violence and Recovery Centre, and Centre for Rights Education and Awareness. HTS counselors gave the card to the woman as a resource to keep or share. Counselors used a counseling script that elaborated the main messages on the card. The script guided counselors to introduce the topic of IPV by saying something like, “We’ve begun giving this card to all our ANC clients so they know how to get help for themselves if they need it, or so they can help a friend or family member who might need it.” For the card’s relationships panel, for example, the script read, “I am talking with all my clients about their relationships. Even if you are not in a relationship now, this is still important information for the future. I want you to think about these questions… Is your partner kind to you? Does he respect you and support you? If so, that is great. That makes it easier for you to protect your health and your children’s health.” Other key messages in the script included, “Remember, you have a right to be free from violence and you deserve to be treated with respect” and “Even if you don’t need any of this support now, relationships change—remember you can always come back to KNH for help.” The enhanced counseling took on average 6.5 min longer to implement than standard HTS (29 min vs. 23 min), the most common length (mode) of both enhanced and standard HTS sessions was 10 min. Third, an IPV counselor from the GBV Centre was posted on-site in the ANC clinic to handle all IPV referrals immediately, with the intent of cutting down logistical and time barriers women face when following-up on referrals to clinics elsewhere in the KNH complex. To further facilitate access to the IPV counselor, a volunteer peer counselor accompanied referred women to the IPV counselor and/or GBV Centre. Finally, HTS counselors involved in the study attended support group sessions facilitated by experienced counselor supervisors from KNH’s mental health unit. The purpose of the sessions was to offer a space for the counselors to debrief and process vicarious trauma, and to discuss questions and strategies about how to handle different situations. A randomized controlled trial assessed indicators on the hypothesized pathway of change—i.e., whether clients were screened for IPV, whether clients disclosed IPV, whether they received a referral, and whether they followed up on a referral—and intermediate outcomes—i.e., our primary outcome: IPV knowledge; and secondary outcome: whether clients felt supported. While the study was not designed to assess whether the intervention helped women address IPV or improve antiretroviral (ARV) adherence (the intervention’s longer term aims), we examined intent to adhere to ARVs to prevent mother-to-child transmission and actions taken to address IPV as exploratory outcomes. All first-visit ANC clients during the study period (882) were approached and screened for eligibility. Out of 852 eligible women, 698 (82%) agreed to participate. Reasons given for declining to participate in the study were time concerns and competing obligations. Consenting participants were randomly assigned 1:1 to either intervention (IPV-HTS) or control (standard HTS) counseling. To minimize bias that could be the result of counselors’ skills, experience, or personality, all counselors provided both the intervention and the standard counseling. While this increased the risk of spillover, the importance of adhering to the protocol was emphasized in training and in regular check-ins. In addition, the clients’ folders included standard HTS materials or IPV-HTS materials, depending on the ANC clients’ assignment to the intervention or control group. Because IPV screening is part of standard HTS counseling, if women in standard care disclosed IPV to their HTS counselor, the counselors explained the services available, i.e., IPV counseling in the ANC clinic and the GBV Centre; and referred her based on the women’s situation and needs. Review of clients’ ANC card and HTS forms supplemented survey questionnaires administered to participants. Participants were interviewed immediately after receiving their ANC and HTS services, and then interviewed again approximately 1 month later at a subsequent ANC visit. The first-round interview was conducted after the intervention for ethical and research design reasons. Had we interviewed women before they received their HTS, and if a participant disclosed experiencing violence to the interviewer, the interviewer would, for ethical reasons, offer empathy and refer the woman for IPV counseling before she even saw an IPV-HTS or standard HTS counselor. As identifying women who are experiencing violence and referring them for IPV counseling are part of the intervention, such support and referrals by the interviewer would have pre-empted the intervention. Thus, both women in the intervention and control groups were interviewed post-HTS. In the post-HTS interviews, survey questions asked every participant about IPV, and if women responded that they had experienced IPV, the interviewer referred the woman for counseling. The interviewers did this regardless of which study arm the woman was in. Inclusion criteria were: first-time ANC client (i.e., a woman presenting for her first ANC visit of her current pregnancy) and being aged between 15 and 49 years. Exclusion criteria were having received an HIV test in the previous 6 months. Following their post-test HIV counseling, clients were interviewed by research staff using a structured questionnaire. Surveys were translated into Swahili and scales’ internal consistency (reliability) was assessed after adaptation to the local language. Questions covered socio-demographic information, clients’ relationship status, partner characteristics, condom use, HIV testing, couples HTS, disclosure of HIV status, knowledge of partner’s status, and perceived difficulty disclosing or asking about a partner’s HIV status. Power inequalities in the relationship were assessed using the Sexual Relationship Power Scale (SRPS) (Cronbach’s α = 0.67) [12]. As recommended, the Relationship Control and Decision-making Dominance subscales of the overall SRPS were combined and rescaled to have a final score ranging from 1 to 4, with a higher score indicating greater relationship power for the woman. Attitudes toward gender norms were measured using a South African adaptation of the GEM Scale (Cronbach’s α = 0.80) [20, 21]. The total scale items were averaged so that the overall scale ranged from 1 to 3, with a higher score indicating greater endorsement of gender equitable norms. We asked about women’s experience of emotional, physical, and sexual violence in her lifetime, in the past 12 months, and by current and other partner/s. For example, for emotional violence, women were asked whether in the past 12 months her current partner had insulted her or made her feel bad about herself; belittled or humiliated her in front of other people; did things to scare or intimidate her on purpose; or threatened to hurt her or someone she cared about. If the answer to each of these questions was ‘no,’ they were followed up with whether her current partner had ever done any of these things; whether another partner had done these things in the past 12 months, or if anyone had ever done these things. We also asked whether the woman had ever sought help for IPV. For most analyses, we operationalized IPV as any IPV—emotional, physical, or sexual—experienced in the past 12 months. Where noted, we also looked separately at those women who reported physical and/or sexual violence only in the past 12 months. Finally, the questionnaire also asked women about their HTS counseling experience, including whether she disclosed violence to the counselor and her follow-up or intent to follow-up on the referral. Information on HIV status was extracted from participants’ ANC medical charts. A second interview was conducted with participants at a subsequent ANC visit (mean duration between interviews was 4.4 weeks) to ascertain whether women had followed up on referrals, assess whether any beneficial effects were sustained, and to monitor whether she had experienced any adverse outcomes due to the intervention. We asked relevant behavioral and self-efficacy questions, such as confidence in ability to regularly take ARVs to prevent maternal to child transmission, to explore whether the intervention had the potential to improve these outcomes. Sample size was calculated to detect a 13% point difference between study arms in IPV-related knowledge and awareness. As no prior data on these indicators existed, 57% baseline prevalence was assumed using endorsement of IPV from the DHS as a proxy variable [22]. Calculations were based on the following additional assumptions: β = 0.80, α = 0.05, 10% non-response rate, and 25% attrition while applying the continuity correction. The study was not powered to assess changes in subgroup analyses (e.g., among women living with or without HIV). Statistical analysis was performed in R Studio Version 3.1 (R Studio Inc., Boston, MA), using an intent-to-treat approach. Chi-square and two-sample t-tests were used to assess whether women systematically differed on sociodemographic characteristics, endorsement of gender norms, or relationship characteristics with regard to study arm or attrition status. Bivariate analyses were also used to assess cross-sectional differences between intervention and control group participants at first and second follow-up. Pairwise missing data were excluded from analysis. We used multivariate logistic and linear regression analysis to assess differences in outcomes at each follow-up time by intervention assignment. Differences were considered statistically significant at p < 0.05. The study protocol was reviewed and approved by the Population Council Institutional Review Board (New York), and the Kenyatta National Hospital/University of Nairobi Ethics and Research Committee. Written informed consent was obtained from each respondent prior to study participation. ClinicalTrials.gov #{"type":"clinical-trial","attrs":{"text":"NCT02577380","term_id":"NCT02577380"}}NCT02577380.