Díaz-olavarrieta, Claudia PhD*; García, Sandra G. ScM, ScD*; Feldman, Becca Seitchik MSc†; Polis, Alberto Martinez MD*; Revollo, Rita MD, MPH‡; Tinajeros, Freddy MPH‡; Grossman, Daniel MD§
SYPHILIS IS A MAJOR CAUSE of adverse pregnancy outcomes and remains a public health problem in many developing countries despite advances in its diagnosis and treatment.1,2 Maternal syphilis is easily detected and treated when syphilis screening is offered as a part of routine antenatal care.3 In areas where women tend to make at least 1 visit to antenatal clinics before giving birth, syphilis detection and screening programs represent a cost-effective means of control.4,5 Yet congenital syphilis continues to pose a global public health challenge, in part due to biologic, cultural, and socioeconomic factors that impede effective prevention, diagnosis, and treatment in women.
Social perceptions and norms about women’s roles make them more vulnerable to risky sexual behavior, sexually transmitted infection (STI) transmission and progression to disease. In many cultures a double standard about gender roles may make it acceptable for men to have multiple sexual partners while women with multiple partners are seen as adulterous.6–8 This inferior status and lack of control over their decisions and bodies makes women especially vulnerable to violence by a partner. Intimate partner violence (IPV), which can be sexual, physical, and/or psychological, heightens women’s vulnerability for contracting STIs.9,10 Furthermore, the fear of violence or aggression that disclosure may evoke from a male partner may prevent abused women from disclosing a STI to a partner.
While partner notification is a well established component of STI control strategies, and integral to maternal syphilis control and treatment, partner notification may increase women’s risk for violence or other negative outcomes.11 Lessons from the HIV epidemic suggest that a woman disclosing a positive diagnosis may be put at risk of intensified partner-related violence.4 This has important implications for maternal syphilis control strategies in contexts such as Latin America and the Caribbean, which have some of the highest rates of IPV in the world.12,13 It is estimated that in Bolivia, the site of this current study, there are 100,000 cases of violence against women each year and only 20% of them are reported.14 Furthermore, in provincial Peru, the neighboring Andean region that shares cultural and socioeconomic similarities with Bolivia, 61% of women have ever experienced physical or sexual violence, compared to 13% in Japan and 23% in Thailand.13
There is a paucity of empirical evidence about the connection between syphilis status and partner-related violence, particularly in areas such as Bolivia, where IPV is high and often goes undocumented. A more in-depth understanding of this connection is imperative in order to guide the implementation of a safe and culturally appropriate partner notification strategy. While data on partner-related disclosure violence were not available for the current study, evidence from the HIV literature shows that women who have experienced violence in the past are more likely to fear violence upon disclosure.11 The current study assesses the association between partner-related violence and partner notification using data from a cross-sectional survey and in-depth interviews with women who tested positive for syphilis, and their male partners.
In 2004, the Population Council’s Regional Office for Latin America and the Caribbean conducted a large feasibility and acceptability study introducing rapid syphilis tests in antenatal clinics in four provinces in Bolivia that were selected according to high prevalence rates. The provinces included were La Paz, Santa Cruz, Cochabamba, and Chuquisaca. For a complete description of study sites, please refer to García et al., 2007, in this issue (pp. S37–S41).
A few months after this study was launched, coinvestigators implemented 2 nested studies related to IPV and partner notification. For this current study, we used data from the partner notification study, which also included questions on IPV.
As part of the partner notification study, we collected qualitative and quantitative data in order to cross-validate information obtained through each technique. The surveys were developed by the research team from the Population Council’s offices in Mexico City and Bolivia. Participants signed informed consent forms and were explained the voluntary nature of the study. The protocol was approved by the Population Council’s IRB.
Survey research for the nested partner notification study, in the form of self-administered cross-sectional surveys, was conducted with women who tested positive for syphilis and their male partners who were recruited via partner notification cards. After receiving a positive diagnosis, all women completed a survey that included questions on barriers for partner notification, whether they had experienced partner-related violence in the past year, and reasons for seeking and completing syphilis treatment. Women then received counseling on partner notification. A trained social worker gave consenting women individual counseling sessions on how best to notify their partners of their syphilis diagnosis. Women were informed of the need for their partners to receive treatment in order to avoid reinfection. Doubts and misconceptions regarding transmission of syphilis were clarified and participants were told they did not have to notify their partners if they feared or suspected violence. Women who feared violent responses by a partner or otherwise indicated a history of violence were given a list of community referrals for IPV and were told that a health worker could notify the partner if the woman did not feel safe or comfortable doing it herself.
Male partners who were notified of their partner’s positive syphilis diagnosis were invited to participate in the study. After obtaining informed consent, a trained social worker then administrated a survey in the participant’s language of choice about barriers and alternative strategies for partner notification. They were also asked a series of questions on IPV based on questions used in the 2003 Demographic and Health Surveys that ask men whether they use physical force with women in current relationships in order to correct or educate them.15 Demographic data were taken from the larger syphilis study database. Data were analyzed using SPSS 10.3.
Qualitative research was conducted using in-depth semistructured interviews. After completing the survey, men and women were asked if they wanted to participate in the interviews during the following months. Participants who consented were contacted shortly afterwards. The interviews took place in the antenatal clinic in a confidential and private location, where they were explained the nature of the study and also told that a decision not to participate would not affect the care they or their newborns would receive at the clinic. Translators were present for those women who were not fluent in Spanish as some may have spoken indigenous languages such as Quechua or Aymara. If the woman agreed to participate in the study, she signed the informed consent and the IDI was conducted. All the interviews were tape-recorded.
The dependent variable, whether a woman notified her partner or not, was dichotomous and defined as yes/no. There were 3 independent variables analyzed as main predictors, which represented correlates of IPV. The first predictor was whether the woman reported IPV in the past year and was defined as 1 if the woman answered yes to either physical or sexual abuse in the last 12 months, and 0 if no. The second was whether the woman perceived partner-related violence as a barrier to notification; this was coded as 1 if she answered that she was scared of talking with her partner about her infection, would not talk to him about syphilis, or was scared of him inflicting violence, and 0 if she answered affirmatively to other barriers (other people finding out, her fault that she had syphilis, that her partner would leave her, that her children would be taken away from her, or that her partner would not go for treatment at the clinic). The last predictor was whether the woman would protect herself during sexual relations knowing her partner had syphilis, which was defined as 1 if she answered that she would use a condom consistently, would stop having sex with her partner, or would attend a health clinic to treat the syphilis, and 0 if she would not protect herself.
We used descriptive statistics to assess for differences in characteristics between women who had notified their partners (n = 137) and women who had not notified (n = 72). We carried out χ2 tests and in cases where cell size fell <5, we used Fisher exact tests. Univariate analyses were conducted to assess the association between the outcome and the 3 main predictors. Lastly, we conducted 3 multivariable logistic regression models to further clarify the association between whether the woman notified or not and the main predictors controlling for demographic variables (education, occupation, marital status, household economic conditions, language(s) spoken at home, number of previous pregnancies, and whether current pregnancy was desired). We tested for interactions between the main predictors and age and household economic conditions based on suggestions about effect modification from the relevant IPV and STI literature. The models were estimated using a backward selection process (with 0.10 inclusion criteria).
In-depth interviews were carried out by trained social workers. The interviews were recorded and took place in antenatal clinics. We conducted 50 interviews in total with 25 women and 25 men. Interviews were transcribed and then analyzed using Atlas.Ti version 5.0,16 which offers a systematic approach to analyzing unstructured data. We identified concepts and themes in a technique called open coding. Researchers had previously identified main themes about partner notification and IPV and open coding revealed a series of quotes.17
All women and men approached and asked to fill out the survey agreed to participate. We collected 209 surveys from female participants and 137 from male participants who had been notified of their wives’ positive diagnosis. Approximately two-thirds of the women (65.0%; n = 137) chose to notify their partners about their positive syphilis status compared to one-third who did not (35.0%; n = 72). Two-fifths of the women reported abuse in the last year (39.2% n = 82) and more than a quarter (28.2%; n = 60) of the women mentioned fear of violence as one of their first two barriers to notifying. Almost all of the women answered that if they learned that their partner had syphilis, they would take actions to protect themselves (92.0% n = 267).
Women who chose to notify their partners had a similar mean age (26 years) and years of schooling (7.5 years) as women who did not notify their partners (Table 1). However, more women who notified were married (24%; n = 50) compared to women who had not (12.5%; n = 26) (P <0.001) while more women who did not notify were single (29.2%; n = 61) compared to those who did notify (4.4% n = 9); (P <0.001). Males who were notified by their partners were also similar to males of partners who did not notify. Their mean age was 29 years and slightly less than three-quarters had completed high school. Approximately one-third (35.0%; n = 48) of the male partners that were notified by their partners reported that they used physical violence at home and, of the men that reported violence, 27% (n = 37) said that they used physical force with their wife when she was pregnant.
Univariate results (Table 2) provide evidence that women who reported that they would not take actions to protect themselves if their partners had syphilis were less likely to notify their partners than women that would protect themselves (OR = 0.06; CI [0.049–0.656]; P <0.0001). However, we found no statistically significant association between not perceiving violence as a barrier to notification and notifying a partner. There was no evidence of an association between women who reported violence in the last year and whether they notified their partner.
Multiple logistic models (Table 2) show that women who reported that they would not take actions to protect themselves if their partners had syphilis were significantly less likely to notify their partners than women that would protect themselves (OR = 0.09; CI [0.02–0.48]; P <0.01). Also, women who did not perceive violence as a barrier had a marginally greater odds of notifying their partner than women who reported violence as a barrier (OR = 1.82; CI [0.93–3.60]; P = 0.08). However, there was no association between whether women experienced violence in the past year and whether they notified their partners. It was also assessed if experiencing sexual violence predicted the probability of notification differentially than if the woman experienced physical violence. Both associations were not statistically significant. There was no evidence of effect modification between the main effects and age or economic status.
Most of the 25 women interviewed had notified their partners (84%), though some had not notified at all (16%). All of the 25 men interviewed had been notified of their partner’s positive syphilis diagnosis. There was no attempt to recruit pairs of partners and all interviews were conducted with the individual participant. Among the female respondents, 64% were married or cohabitating while 84% of the males were married or cohabitating. Table 3 provides demographic information on the men and women interviewed.
Fears of Notifying.
When women were asked about disclosing their syphilis status to their partners, almost all reported fear of notifying. Both those who notified their partners themselves and those whose health care providers notified indicated that they were scared of their husbands’ responses. Common fears were that he was going to leave, blame her for being unfaithful, and get angry that she infected him. As one woman explained, she was scared, “because I thought I was going to be left alone, all by myself with my two children, feeling alone and with no support …. That is why I could not notify him.” Ultimately, respondents were scared of their partners becoming aggressive either psychologically or physically as evidenced by the following quotes:
“… I was afraid that he would get drunk one day and then would hit me, because he is going to blame it on me …. ” (woman who had a health care provider notify partner)
“… I think that what hurts the most … is when you are beaten psychologically … when he abuses us and tell us things like stupid, idiot, etc …. ” (woman who notified partner herself)
Interestingly, women who reported fear of telling their husbands still tended to personally notify them.
Women’s Actual Experience Notifying.
Of the women who notified their partners, most described their experiences of notifying much smoother than they had anticipated. The majority indicated that their husbands responded in a caring and understanding manner and did not assert blame or become aggressive. Women used words like “relaxed,” “calm,” and “sympathetic” to describe their husband’s responses and reported little violence despite their initial fears. However, several women reported feeling uneasy about the experience of notifying and their partner’s reaction. The following quotes are examples of the uncertain and potentially dangerous situation that notifying put them in:
“… At first I was afraid, afraid that he would probably yell at me or that I would be physically abused ….”
“… I had fear, I feared that he would tell me I was a liar or something … fear that he would get upset or that he would somehow want to hurt me ….”
Women’s views were inconsistent regarding whether pregnancy was protective or not against IPV. One woman explained that her partner’s anger would have come out more if she had not been pregnant. One respondent suggested, “If I were not pregnant, I think he would have reacted violently against me ….” However, another woman described that their husband blamed her: “At the beginning he blamed it on me, that I was the one who had the infection, he was aggressive ‘cause he told me it was my fault that we had the baby, we had an argument ….”
Partners Describing How Other Men Would React.
All of the men who were interviewed had been notified of their partners’ positive syphilis status. Most of the men implied that in general men would act aggressively and angrily upon hearing that their partners were infected. They said that they could imagine that men would blame their partners, get upset, react violently, beat them, or leave them. Said one person, “Other men could be much more violent than I am, they could beat the shi.. out of them, and probably everyday, there are other men more aggressive than I am.” Partners emphasized that guilty feelings on the side of both men and women tend to explain the anger and violence that surrounds disclosure. Suggested one respondent, “… the way I see it, there could be an argument for sure …. yes, there could be arguments, and they will try to get rid of all the guilt … he or she …”
Three men, however, said they, too, would get angry under certain circumstances. They explained that if they found out their partner was cheating on them, they would punish them as exemplified by the following quotes:
“If the woman was the one to blame, at least myself …I would beat her up, or at least yell at her.”
“… If the woman gets involved with someone else while the husband is at work, that would be cheating on him, I would think that this woman is not worth it and I would leave her ….”
Overall, male partners appeared to be understanding of their partners’ syphilis diagnoses; however, there were strong emotions and cultural beliefs dictating their responses.
The majority of the women in this study were able to notify their male partners of their syphilis diagnosis successfully and reported the actual experience as being less fearful than originally anticipated. However, most women reported having experienced IPV or fearing IPV as a result of disclosure. Furthermore, some male partners openly acknowledged they would “discipline” their female partners “as needed.” Taken together, these findings suggest that syphilis control strategies also could screen for IPV during pregnancy, provide partner notification counseling, and offer prompt and easy referrals to the appropriate social services. It is possible that doing so could help women in abusive relationships to communicate their test results safely to partners.9
Women who reported that they would be unable to protect themselves if their partner had syphilis were less likely to notify their partner of their positive status than those women who reported they could protect themselves. Fear of partner-related violence had a slightly positive association with not notifying a partner. Studies conducted with HIV-positive women show that women often refuse to report their infection status to their partners for fear of their reaction.9 However, in the current study, women who reported partner-related violence were just as likely to notify their partner as those who did not report violence. Likewise, of the 10 women interviewed that reported fear of violence, 8 of them went on to disclose their status to their partners. This may be partly explained by the fact that all participants received counseling sessions on partner notification strategies after they were notified of their positive syphilis test. The counseling may have provided them with the necessary support and confidence to notify despite their initial fears about IPV. The null association between IPV and partner notification found in this study is encouraging and could be due to the intervention. However, without a control group, it is not possible to determine what the association would have been in the absence of counseling. Thus, this finding can not be generalized to other low resource pregnant Bolivian women in the absence of the same intervention.
The men’s survey and interview responses overall confirmed women’s reports. Over one-third of male respondents surveyed reported using physical force at home, and more than half thought it was unacceptable for their partner to refuse sex for no apparent reason. The males who were interviewed more in-depth reported that they thought men in general would react negatively to the disclosure with aggression and anger towards their partners. However, the majority reported that they themselves did not react this way. This is reassuring, as disclosure violence could have occurred despite women receiving effective partner notification counseling. The men did suggest that their own reactions would have been more severe if they thought that their partner’s syphilis was a consequence of her infidelity.
This study has several limitations. First, findings from this study cannot be generalized beyond women who attend antenatal clinics in Bolivia. This was a self-selected convenience sample of women at antenatal clinics. Therefore, participants may have different testing and partner notification behaviors than nonparticipants attending the clinics as well as other women who do not receive antenatal care. Women who attend antenatal clinics also may be more informed and educated than women who do not get antenatal care in Bolivia. Second, we did not have survey data on the men that were not notified by their partners; thus, we were unable to assess the association between men reporting violence and whether women disclosed their syphilis status to them. Third, couples were not recruited for the interviews thus we could not analyze experiences, behaviors, or opinions by couples. Lastly, women reported favorable disclosure experiences, despite stating their prior fears of violence as a barrier to notification. This may be due to the counseling, which provided them communication strategies to be able to open up to their partners.
This study provided an initial exploration of the association between IPV and partner notification in antenatal clinics in Bolivia. The use of dual methodology provided cross-validation of each technique and also contributed both partners’ experiences and fears regarding infection and notification. The study clarified areas where subsequent research is needed in this setting to more fully understand how to guide the implementation of syphilis control programs. A crucial area for future investigation is assessing whether the episodes of partner-related violence change in frequency or intensity due to disclosure of their syphilis status. Few STI control strategies include an IPV screening and a counseling component, which may be important to help women disclose test results to their partners safely.11 Further investigation is needed to determine the necessary information to provide in the initial partner notification counseling at the antenatal center as well as the appropriate frequency of continued screening after the initial visit.
These findings suggest that syphilis control programs which include an IPV screening procedure administered to all women followed by brief counseling may benefit women who are in need of extra guidance and disclosure strategies in Bolivia. It is acknowledged, however, that in order to effectively add an IPV screening and referral strategy to syphilis programs in Bolivia, considerable barriers will need to be overcome. These barriers exist at the level of the community, provider, and government and include assuming extra costs, training personnel, increasing logistical support, and strengthening political investment to implement these policies.4 Bolivia elected its first indigenous president in 2006 and women’s health has been put on the national agenda with promises for increased funding and focused policies. Hopefully the incoming government will recognize that a combined strategy is feasible and needed given the burden of syphilis and IPV as well as the availability of cost-effective diagnosis and treatment tools.4