Brief Report: Sexual Violence Against HIV-Positive Women in the Nyanza Region of Kenya: Is Condom Negotiation an Instigator?

For people living with HIV, exposure to sexual violence (SV) is associated with decreased adherence to antiretroviral medication, a primary predictor of their survival. Identification of risk factors for SV is a pressing issue in sub-Saharan Africa, where the global majority of HIV-positive women live and the prevalence of SV against women is high. We used qualitative data to examine SV against HIV-positive women enrolled in HIV care in Kenya. Respondents identified husbands as perpetrators of SV in the context of women's efforts to use condoms as directed by HIV care providers.

INTRODUCTION

The association between intimate partner violence (IPV) and HIV infection among women is well established[1] and thought to be bidirectional.[2] Although prospective data show that IPV is a risk factor for subsequent HIV infection among women,[3,4] the evidence for how HIV infection triggers subsequent IPV is emerging only now. Respondents in qualitative studies of HIV+ women living in South Africa and Swaziland state that diagnosis and disclosure of HIV to an intimate partner is itself a trigger for IPV.[5,6]

The objective of this study is to address this research gap, using a qualitative design to explore how HIV+ status might lead to IPV with a focus on sexual violence (SV), a form of violence against women that can occur in partnerships. IPV is known to have a particularly damaging effect on mental and HIV health.[7-11] We used data from recently published qualitative mental and HIV health care needs assessment[12] to identify the context and potential pathways of SV against HIV+ women receiving HIV care from the Family AIDS Care Education and Services (FACES) program in the Nyanza region of Kenya. Nyanza has a high prevalence of HIV (18.7% in Kisumu County) and SV against women (51.9% lifetime prevalence among ever-married women aged 15–49 years).[13,14]

METHODS

Study Location

FACES is a Kenya Medical Research Institute and University of California San Francisco collaborative in Kisumu County, Kenya. Established in 2004, FACES is a President's Emergency Plan for AIDS Relief–funded prevention, care, and treatment collaboration serving more than 140,000 HIV+ individuals in western Kenya. This study took place at the FACES Lumumba Clinic.

Study Patients

The parent study, a qualitative mental and HIV health care needs assessment (April–June, 2013), included 30 in-depth interviews and 4 focus group discussions with a total of 61 study participants.[12] Participants were female HIV+ FACES patients, community leaders, and/or clinic staff. Inclusion criteria included age of 18 years or older, ability to give verbal informed consent, attend the duration of the interview, and absence of severe cognitive dysfunction interfering with participation. Interviewers were 2 men and 2 women. Study procedures were approved by the research ethics boards of the University of California San Francisco and Kenya Medical Research Institute.

Interviews and Study Domains

Interviewers followed a structured interview guide consisting of open-ended questions and standardized probes. Respondents were asked to report on the experience of HIV+ women in their community, as opposed to their own personal experiences. Although the parent study was designed to investigate the breadth of violence against HIV+ women and related mental health care needs, SV was a probe for the question: “What kinds of violence do HIV+ women suffer?” with specific follow-up questions focusing on who the typical perpetrator was, when, and in what setting the violence typically occurred. Questions on SV were not added for the substudy.

Data Collection and Analysis

Interviews were audiorecorded with no participant identifiers and were transcribed into Dholuo/Kiswahili and then translated into English. Data analysis was guided by grounded theory,[15] which was used to identify concepts, categories, and propositions and by seeking similarities and differences among interviews. Concepts were grouped to develop broader categories, which were then compared and grouped to develop themes.

Primary and secondary codes were established through a process of open coding on specified domains of all (100%) interviews (S.M.M.) and by comparing results with a separate open coding of a random selection of 15% of interviews (S.L.D.). Discrepancies across coders were discussed and resolved with slight modification of primary and secondary codes. The resulting codebook was applied by another researcher to a randomly selected subsample of 50% of the interviews. Application of codes was discussed with minor modifications to secondary domains. The remaining 50% of interviews were coded independently. The overall concordance rate was calculated to be 98.5% across the parent study interviews based on the random reliability checks of 50% of interviews. For this study, we analyzed all data excerpts coded as “sexual violence,” without restriction.

RESULTS

Overall, 74% of all interviews (n = 25) in the parent study mentioned SV against HIV+ women (21 individual interviews and all 4 focus groups). All of the substudy's in-depth interview participants were female (100%); 5 were community health workers (20%), 3 were health care providers (12%), and the remaining were HIV+ women receiving care at the FACES clinic. The 4 focus groups ranged in size from 4 to 10 participants and consisted of (1) community health care workers (9 participants, 6 female); (2) FACES patients (all female, Luo-speaking); (3) FACES patients (all female, Swahili-speaking); and (4) community advisory board members (4 participants, 1 female). Participants' actual names are replaced with pseudonyms. Categories were grouped to identify key themes regarding the context of SV against HIV+ women. See Table 1 for a summary of findings. Here, we focus on the most common themes. The percentages reported in Table 1 and below are the number of interviews referencing the theme, divided by the total number of interviews from the parent study that referenced SV.

TABLE 1.

Overview of Key Themes Regarding the Context of SV Against HIV+ Women in Kisumu, Kenya

Agents and Context for SV Against HIV+ Women

All respondents who identified perpetrators of SV (56%, n = 14) stated that husbands of HIV+ women enacted SV against them.

Transmission and Superinfection

As part of HIV management, health care workers advised women to use condoms to reduce the risk of transmission or superinfection. The majority of participants (72%; n = 18) stated that SV occurred when HIV+ women requested that their partner use condoms.

A FACES provider stated:

Here at the clinic we encourage and give condoms, that when they go back home they should use them so that it prevent any infection in case it occurs. Then and when they reach home and maybe all of them are positive and there he says that we are all positive and why should we use the condoms or something like this. So the man will end up having sex without using the protection and as the woman will not be willing to give in; it will be like a rape case in such a situation (Akili, participant in focus group of health care providers).

Disagreement About Fertility Preferences

Many respondents stated that local HIV+ women wished to use condoms to prevent pregnancy because of concerns regarding the impact of HIV on the child's health. Some participants (20%; n = 5) mentioned their wish to avoid pregnancy was a facilitator of male SV. Multiple interviews (24%; n = 6) cited husband's wishes for additional children as an instigator of SV against HIV+ women. FACES patients stated:

Women are being forced to have sex since they (men) say they want children (Esther, female FACES patient, in-depth interview).

DISCUSSION AND RESEARCH IMPLICATIONS

Our study suggests that there may be particular characteristics of SV against HIV+ women that could be leveraged for prevention. Our findings are consistent with prior research demonstrating that relationship inequity and IPV are associated with incident HIV infection among women.[16-20] To the extent that SV against HIV+ women is informed by the gender norms facilitating violence against all women, gender transformative programs should be investigated for their potential to reduce SV against HIV+ women. Programs such as Now![19] and Safe Homes and Respect for Everyone[21] have promising results for decreasing social acceptance of IPV among men and women and reducing IPV, including SV against women.

Condom Negotiation Between Seroconcordant Spouses Superinfection

One prominent characteristic of SV against HIV+ women in our study is that respondents reported the condom negotiation to be an instigator of SV against HIV+ women. Although research has established that IPV against women is associated with reduction in their confidence to negotiate condom use with their partner,[22,23] respondents in this qualitative study in Kenya perceived that condom negotiation, itself, led to SV against HIV+ women in their communities.

Condom use is clearly an important public health intervention to prevent HIV transmission in serodiscordant couples. However, condom use continues to be recommended for seroconcordant couples (both partners HIV+) based on the potential risk of “superinfection” (infection with more than 1 strain of HIV), although data from Kenya did not find deleterious health effects among women with superinfection and the use of antiretroviral therapy further limits the concern for superinfection.[24] The chain of events leading to SV against HIV+ women commonly described in these interviews is that women were tested for HIV, found to be positive, and were then asked by HIV care providers to disclose their status to their intimate partner/husband and initiate condom use, even if their partner was also known to be HIV+. Efforts to enforce condom use by HIV+ women were met with SV from their husbands.

Although it is not the focus of this report, it is important to note that Kenyan men lag behind women in HIV testing[13]—seroconcordant couples may be more common than current numbers suggest. As we describe elsewhere,[12] HIV+ women who report their status to untested husbands are often accused of bringing HIV into the family and suffer abuse from husbands and in-laws.

Disagreements Regarding Childbearing/Vertical Transmission

A substantial proportion of interviewees stated that HIV+ women requested condoms to prevent pregnancy, reflective of health care provider counseling on the risk of vertical transmission and maintaining health while living with HIV. However, respondents in sub-Saharan Africa qualitative studies stated that men preferred to have more children and/or have children at shorter intervals.

Short-Term Research Recommendations

This study suggests that there may be particular characteristics of SV against HIV+ women in Kenya that could be addressed through improved application of HIV couples testing. Couples-based HIV testing is associated with decreased IPV and better adherence to condom use guidelines. Couples testing may be a promising intervention to mitigate SV against HIV+ women, alone, or as a component of a larger gender transformative program.

To the extent that couples testing improves HIV testing among men, it may help identify more seroconcordant couples, presenting an opportunity for HIV care providers to tailor and adapt risk reduction strategies that promote condom use, including consideration of partnership dynamics, especially the risk of SV. Regardless of whether couples are seroconcordant, couples testing should be used as an opportunity for health care professionals to discuss SV (and broader IPV) against HIV+ women and contribute to prevention efforts.

In addition, couples testing may provide an opportunity for HIV care providers to review the principles of family planning with both partners, including the protocol necessary to prevent vertical transmission. To the extent that couples testing shifts the burden of negotiating condom use from HIV+ women to the health care system, it may foster understandings of the couple as a partnership, which may lead to decreased SV against HIV+ women. Implementation research on for HIV couples testing is needed.

LIMITATIONS

This is a substudy from an existing qualitative assessment that included 1 question on SV. As such, the frequency and types of SV among HIV+ women in this population might be underrepresented by the available data. However, given the limited data now available on the topic, this study will assist in identifying areas for future research. Second, this is a relatively small study and focused on 1 location within Kisumu County. Although saturation on key themes was reached, as with all qualitative studies, generalizability is not a goal. Third, to prevent retraumatization, study participants were asked to provide information regarding HIV+ women served by FACES, as opposed to their personal experiences. Some participants spontaneously elected to provide personal information. Thus, our data analysis reflects a mix of data focused on perceptions of the self and others, and the “second-hand” information may be less reliable.

CONCLUSIONS

Participants in this study reported that the efforts of HIV+ women to follow health care provider recommendations to use condoms to reduce the risk of superinfection and transmission (sexual and vertical) increased their risk of sexual victimization. We identified 3 key areas for future research, including continued gender transformative research, an analysis on the overall health impact of recommending condoms solely to prevent superinfection for women in sub-Saharan Africa, and implementation research to speed the scale-up of couples testing and counseling.