Violence-Related Coping, Help-Seeking and Health Care–Based Intervention Preferences Among Perinatal Women in Mumbai, India

Abstract

Domestic violence is a significant public health issue. India is uniquely affected with an estimated 1 in 3 women facing abuse at the hands of a partner. The current mixed-methods study describes violence-related coping and help-seeking, and preferences for health care—based intervention, among perinatal women residing in low-income communities in Mumbai, India. In-depth interviews were conducted with women who had recently given birth and self-reported recent violence from husbands (n = 32), followed by survey data collection (n = 1,038) from mothers seeking immunization for their infants ages 6 months or younger at 3 large urban health centers in Mumbai, India. Participants described fears and other barriers to abuse disclosure, and there was a low level of awareness of formal support services related to violence. Qualitative and quantitative findings indicated that formal help-seeking is uncommon and that informal help sources are most frequently sought. Quantitative results revealed that, while few (<5%) women had been screened for violence in the health care setting, most (67%) would be willing to disclose abuse if asked. When presented with a list of possible clinic-based violence support interventions, participants endorsed crisis counseling and safety planning as most helpful (90.9%). Findings provide direction for violence-related intervention services for perinatal women. A multipronged approach that includes strengthening the informal support system, for example, neighbors and family members, as well as facilitating access to formal services building on the health care system, warrants exploration in this context.

Introduction

Violence against women is a significant public health issue with significant health and human rights implications (García-Moreno, Jansen, Watts, Ellsberg, & Heise, 2005). A large and growing body of research demonstrates the negative impact of violence on women’s mental, physical, and reproductive health (Campbell, 2002; Decker et al., 2008; García-Moreno et al., 2005; Hurwitz, Gupta, Liu, Silverman, & Raj, 2006; Lee & Hadeed, 2009; Panchanadeswaran & Koverola, 2005; Raj, Liu, McCleary-Sills, & Silverman, 2005; Silverman, Decker, Reed, & Raj, 2006; Silverman, Decker, Saggurti, Balaiah, & Raj, 2008; Silverman, Raj, & Clements, 2004). These data highlight the need for both primary prevention efforts to reduce violence against women, as well as support systems to help victims of violence cope with these traumatic experiences. Some evidence suggests that victims of violence can benefit from disclosure of abuse and help-seeking from the formal sector. Community-based support services have been found to provide critical support to survivors (Bennett, Riger, Schewe, Howard, & Wasco, 2004; Ullman, 1996; Wasco et al., 2004), and can impart reductions in posttraumatic stress symptoms (Wasco et al., 2004), self-blame (Starzynski, Ullman, Filipas, & Townsend, 2005), and subsequent violence (Bybee & Sullivan, 2002; Sullivan, 2003; Sullivan & Bybee, 1999). Globally, a significant barrier to obtaining this support is women’s reluctance to disclose abuse and seek help for their experiences of trauma, often for fear of being blamed or not believed (García-Moreno et al., 2005). Other reasons for not disclosing abuse include perceptions of violence as normal or not serious, fear of consequences such as revictimization by the perpetrator, or the threat of losing children (Fanslow & Robinson, 2011; Miller et al., 2010). Given the documented benefits of abuse disclosure in a supported environment (Bennett et al., 2004; Bybee & Sullivan, 2002; Sullivan, 2003; Sullivan & Bybee, 1999; Ullman, 1996; Wasco, et al., 2004), research to clarify patterns of help-seeking is necessary to guide services, particularly in areas uniquely affected by violence.

India is one such setting. National data indicate that over one in three married women have experienced physical or sexual partner violence (International Institute for Population Sciences [IIPS] and Macro International, 2007; Silverman et al., 2008), and recent cross-national comparisons suggest a higher prevalence of both lifetime and recent physical violence as compared with other settings (Larkin & Morris, 2009). The South Asian context confers additional barriers to abuse disclosure including cultural prohibitions against discussion of family problems that would bring shame and stigmatization to the family reputation, and related fears of jeopardizing family honor (Kanukollu & Mahalingam, 2011; Lee & Hadeed, 2009; Naved, Azim, Bhuiya, & Persson, 2006). Perhaps in response to this constellation of barriers to disclosure, quantitative evidence from India (IIPS and Macro International, 2007; Kamat, Ferreira, Motghare, Kamat, & Pinto, 2010; Sarkar, 2010) and elsewhere in the region (Naved et al., 2006), as well as South Asian immigrant populations in other settings (Lee & Hadeed, 2009), illustrates that few women share experiences of abuse, and far fewer seek formal services. Those who do disclose appear more likely to turn to informal sources of support (e.g., family and friends) rather than formal sources (IIPS and Macro International, 2007; Naved et al., 2006; Raj & Silverman, 2007).

To date, little is known about the nuances of help-seeking patterns, including the nature and usefulness of help received, or experiences with formal sources of support within the Indian context. Evidence from Indian women recruited from shelter samples (Panchanadeswaran & Koverola, 2005) and immigrant South Asian women (Ahmad, Driver, McNally, & Stewart, 2009; Lee & Hadeed, 2009; Raj & Silverman, 2007) suggests a trajectory of help-seeking that starts with informal sources and advances to more formal sources of support as abuse continues or informal support networks fail to provide help (Panchanadeswaran & Koverola, 2005). To date, help-seeking patterns and preferences among community-based samples of women in India remain unclear. Finally, despite the development of clinic-based screening and intervention protocols in the United States and elsewhere (Hamberger & Phelan, 2004; Rabin, Jennings, Campbell, & Bair-Merritt, 2009), little is known about clinic-based violence screening experiences and intervention preferences within India.

Such investigation is particularly relevant during the perinatal period. Pregnancy is simultaneously a high-risk period for violence victimization (Brownridge et al., 2011; Devries et al., 2010), as well as one in which women often have more frequent contact with medical services and may be more open and accessible for violence-related interventions (McFarlane, Soeken, & Wiist, 2000). Understanding patterns of coping with violence, including help-seeking behaviors, among abused women who have recently given birth is an important step in developing the infrastructure to effectively intervene during this critical period.

Against this backdrop, the current study sought to describe violence-related coping and help-seeking behaviors, screening experiences, and intervention preferences among a clinic-based sample of low-income women with infants in India. Findings are intended to inform perinatal health care practitioners and provide insight into ways to help abused women get the support they need.

Method

The current study consists of secondary data analysis of the mixed-methods “Mechanisms for Relations of Domestic Violence to Poor Maternal and Infant Health’” study, which primarily sought to investigate the pathways by which domestic violence relates to maternal and infant health concerns. In-depth interviews (qualitative) and survey (quantitative) data were collected from mothers of infants ≤6 months presenting for infant care at urban health centers (UHCs) in Mumbai, India (Raj et al., 2011). The qualitative research preceded the quantitative study with the goal of informing survey development for the latter component. All procedures were approved by the Institutional Review Boards of Harvard School of Public Health and National Institute of Research on Reproductive Health (NIRRH) via the Indian Council of Medical Research. Current analyses are specific to examination of coping, help-seeking, and intervention preferences.

Qualitative Method

The UHC of the Shivaji Nagar slum community of Mumbai, India, served as the recruitment setting for in-depth interview participants. State and city government reports estimate that Shivaji Nagar is home to an estimated 100,000 persons; the majority (85%) are Muslim. The community is made up of clusters of 20 to 25 or more households with no roofs or nonconcrete roofs, which typically lack drinking water, toilets, and drainage. The vast majority of women (97%) do not work outside the home; husbands commonly work as tailors, domestic servants, and day laborers. Over half of the women (65%) are literate, a proportion larger than that seen for Indian women as a whole (IIPS and Macro International, 2007; Raj et al., 2011); however, this literacy is typically in Urdu rather than Hindi, Marathi, or English (the main languages spoken in Mumbai).

Eligible participants were women aged 15 to 35 years who had given birth in the past 6 months, were attending the health center for child immunization or other infant care, and reported emotional or physical abuse from husbands either in the year prior to their most recent pregnancy (i.e., the pregnancy resulting in their current infant), during this index pregnancy, or since this index pregnancy, assessed via a modification of the Abuse Assessment Screen (McFarlane, Parker, Soeken, & Bullock, 1992). Over a 5-month period, participants were recruited from patient lines following a recruitment schedule that rotated times of day and days of the week. Women were initially screened on the basis of their infant’s age; those meeting this component of eligibility criteria and expressing interest in participation were invited to a private room within the health center for abuse-related eligibility screening, written informed consent, and data collection with the assurance that their place in the line would not be lost due to their study participation. These procedures generated a sample of 32 women. Following determination of eligibility and informed consent, participants completed a 90 to 120 min in-depth interview that explored women’s experiences of abuse by husbands and other family members. Their experiences of abuse, help-seeking, and perceived health consequences were assessed via open-ended questions (e.g., “How were you treated?,” “Did you tell anyone about your experience, and why or why not?”). Extensive probing as to the nature of abuse, what happened during help-seeking, and health impact elicited a narrative response in which participants described their stories.

All interviews were conducted in Hindi by master’s-level trained female research staff from the NIRRH. To ensure participant comfort and reduce risk for distress, participants confirmed that they were comfortable discussing the conflicts in their marriage. During the interview, participants were carefully monitored for potential distress; the interviewers periodically asked women how they were feeling with the topics and provided ongoing validation for their experiences throughout the interview. Following the interview, participants were offered an escort to a hospital-based domestic violence-related service in an attempt to promote service utilization.

Interviewers took detailed notes during the interviews; within 24 hr of each interview, interviewers detailed their notes fully, typed, and translated them into English using Microsoft Word (digital audio recording was not found to be acceptable by this population). Translated and typed Word files were sent to project investigators for analysis.

Using a grounded theory approach (Glaser, 1967; Raj et al., 2011; Strauss, 1987; Strauss & Corbin, 1990), the research team used an iterative process to identify mutually exclusive but possibly linked codes or themes across interview transcripts. Two members of the research team independently coded each in-depth interviews and identified key themes. Coding schemes were compared, with discrepancies and differences in interpretation discussed among the larger team. Codes were generated, built upon, and revised via an iterative process involving the coders and other study investigators. Subsequently, two graduate-level research assistants independently coded interview data. Additional codes and subcodes were identified iteratively in this coding process, and reapplied to previous interviews as needed. Intercoder reliability across coding was reached via a standard approach (Carey, Morgan, & Oxtoby, 1996; Raj et al., 2011). Specifically, coders came to agreement on all codes; in the case of disagreement, a final decision was made by a doctoral-level study investigator who provided oversight during the coding process. Coded data were sorted manually; that is, coded and sub-coded data were cut and paste into Word documents labeled by code, with coded data labeled by interview. Tree diagrams depicting study domains, codes specific to these domains, and subcodes within each code were also constructed following the same iterative process as that used for the study coding as a whole (Raj et al., 2011; Schensul, Schensul, & LeCompte, 1999). The current article focuses on the themes of coping with, and help-seeking for, abuse. Data are presented based on the type of coping; subcodes from each code and one to two quotes best illustrating each subcode are presented.

Quantitative Method

Following collection and analysis of in-depth interview data described above, survey data collection was conducted at three large UHCs located in major slum areas of Mumbai, India: Shivaji Nagar, Bail Bazaar, and Mohili village. Eligible participants were mothers, aged 15 to 35 years seeking immunization for their infants aged 6 months or younger at one of the three participating clinics. Unlike the qualitative study phase, eligibility was not limited to those who had experienced abuse so as to enable an estimate of abuse prevalence. Recruitment sites were selected on the basis of size (>100,000 residents), as well as the presence of a UHC in the community. The UHC immunization clinics provided the setting for participant recruitment and data collection. Recruitment occurred during all immunization hours from August to December 2008. During recruitment, mothers were screened for eligibility (having an infant ≤6 months) and willingness to participate in the survey by clinic-based community health volunteers and outreach workers prior to their infant immunization. Following infant immunization, nursing staff asked women selected for recruitment if they would be interested in hearing about the research study. Those expressing an interest in participation were accompanied by the recruiter to a private room within the clinic to speak with a trained research staff member. Once in the private room, research staff verified interest and eligibility, obtained written informed consent, and collected survey data. Surveys were interviewer-administered based on literacy concerns; to enable participant comfort in discussing sensitive topics, all research staff members were female master’s-level employees of the NIRRH trained in survey research, ethics, maternal and child health, and domestic violence. Surveys were administered in Hindi and took approximately 30 to 40 min. Participants were screened for emotional distress and provided with referrals for legal and mental health assistance following their participation in the study. A total of 1,830 women were approached for recruitment of whom all were known to be eligible based on infant age. Of these, 60% (1,108/1,830) agreed to learn more about the study, 94.6% of whom (1,049/1,108) agreed to participate, provided written consent, and completed the survey. Of the 1,049 participants, 11 were dropped from further analysis due to incomplete responses, resulting in a final sample size of 1,038.

All data were self-reported; the interview instrument was developed based on standardized measures for perinatal health and domestic violence from the National Family Health Survey (IIPS and Macro International, 2007) that were then adapted as necessary for relevance to the current setting based on the qualitative findings, for example, context-specific forms of abuse (Raj et al., 2011). Outcomes examined in this article include a series of investigator-initiated measures that assessed women’s experiences with violence-related help-seeking, health care—based screening, willingness to disclose abuse, and preferences for clinic-based interventions. Participants reporting physical or sexual abuse were asked “the last time you experienced severe violence or harassment, did you tell anyone or seek any help?” Those endorsing this item were asked a follow-up question concerning the individual or source sought help from (e.g., parents, in-laws, husband, friend, neighbors). Clinic-based screening and intervention items were asked of all participants. Assessments addressed experiences with screening (i.e., “Has a doctor or nurse ever asked you about violence/conflict?”), willingness to disclose abuse (i.e., “Would you be willing to tell a doctor or nurse about violence/conflict if they ask?”), and desire for clinic-based intervention (i.e., “Would you want to receive information or assistance from a doctor or nurse to help with violence/conflict?”). Finally, participants were given a list of potential interventions ideas that were generated from the qualitative phase, and asked to indicate whether or not they would be helpful. Potential interventions were evaluated independently (i.e., sequentially) by participants, and included counseling services for only the woman to receive support for conflict they face in the home and provide them with ways to increase their safety, counseling services for women with their husbands or in-laws to reduce conflicts and improve marital communication, a class for pregnant women and their husbands to teach men about the health needs of pregnant women and new mothers, a class for pregnant women and their in-laws to teach in-laws about the needs to relieve pregnant women and new mothers of some household burdens, nutrition needs of pregnant women, a class for husbands to teach them how to be nonabusive in their interactions with their wives, and help the women find a place for her and her children to go if she were in danger. Basic descriptive analyses were conducted on all variables.

Results

Qualitative Results.

In-depth interview participants (n = 32) ranged in age from 16 to 35 years (mean age = 23.9 years); all were married at the time of interview, and most (n = 29) were Muslim with the remaining participants (n = 3) Hindu. Participants predominantly had one child (n = 11) or two children (n = 11); the remaining women had three children (n = 9) or seven children (n = 1). Of women reporting an educational level (n = 29), 16 indicated primary education or less.

Self-Contained Responses to Violence.

Participants described a range of responses and coping mechanisms that did not involve other individuals or formal institutions.

Silence.

A dominant coping response to mistreatment and abuse described by our participants was one of silence. This silence reflected simultaneously a sense of resignation, often one that built over generations, as well as salient barriers to disclosure.

Resignation to Violence

Most participants described violence as a normal part of married life, and a situation for which no viable alternatives exist.

As I told you my husband is very arrogant person. It’s very difficult to stay with him but I don’t have any other source to survive. He used to beat me till death. He showed very controlling behavior. Some time I feel he is lunatic (mathephiru). He always discriminate me. I don’t have any respect for him. But nobody can change destiny”

–32 years, Muslim, married 20 years, illiterate, joint family, two children

This type of arguments happens in my house also. Moreover my husband slaps me also. It’s not a new thing. Husbands are made to beat their wife, which is very common.

–24 years, Muslim, married 8 years, education sixth in Urdu medium, nuclear family, three children

One participant noted that her partner’s abuse included forcing her to have sex with others; her sense of resignation and lack of alternatives prompted her compliance.

He (her husband) started forcing me to have sex with other man. After few days I got ready to do what he suggested, as there was no option for me.

–27 years, Muslim, married 14 years, illiterate, joint family, three children

Barriers to Disclosure to Family and Neighbors

Fear of disclosure to family and neighbors was also an overwhelming theme, and one that appeared to be underpinned by fear of social repercussion and increased violence from perpetrators.

In many cases, participants described these fears as preventing disclosure either in general, or to specific individuals such as parents and neighbors as illustrated:

All these happened to me but I never told my parents I was always afraid, if I told everything to my parents then my husband would ask me for divorce. For that reason only I was bearing everything.

–19 years, Muslim, education level seventh, married 2 years, joint family, one child

[Interviewer: What about your neighbor? I mean to say did you get any support from them?] “No, they don’t know anything about my personal life. They just know, we are living happily with our children and as it is, what they will do if I said all this things except laughing on me and my husband. That is the reason I don’t say all these things to anyone including my brother.”

–24 years, Muslim, married 8 years, education sixth in Urdu medium, nuclear family, three children

I can’t go to my parents place, they will kill me if they come to know about all these things about me.

–27 years, Muslim, married 14 years, illiterate, joint family, three children

Independent action.

Active strategies to reduce violence that did not require outside assistance were rarely described by participants; those active strategies that did emerge included desire for economic independence and sexual activity with partners as a way to stave off conflict.

Economic Independence

Participants’ dependence on their husbands for their daily needs and requirements was a dominant theme. Some expressed a desire for economic independence as a means of freedom from conflict and unwanted interference. Despite these goals, no participant described actually achieving economic independence, likely reflecting a confluence of the impoverished nature of the community as well as the low status of women therein.

Why he (her husband) wanted to keep the child in the orphanage? As we don’t have money for our child, after that I decided to work outside … Now I am fed up all this things and thinking to divorce my husband. Really I am too much tired of all these things and wants to go somewhere with my child. I want to go where there is no husband or no one else to interfere in my life. I will earn and give food to my child.

–18 years, Hindu, illiterate, married 2 years, nuclear family, one child

Sexual Activity

A few participants described sexual activity with husbands as a strategy to resolve existing marital disputes.

Actually in our community we have the tradition that husband and wives are not allowed to sleep together till 40 days. But in my case whenever my husband had fights with me, with the help of sex I resolve the fights.

–23 years, Muslim, education ninth standard, married 4 years, three children

Sexual activity was also described as a strategy for heading off potential conflicts including husband extramarital relationships. This strategy was underscored by a sense of isolation and lack of alternatives perceived by participants who described it.

When I was pregnant according to the Doctor we should not have any physical relationship between us but my husband is not going to listen … but I never refuse to make physical relation. If he will not make with me then he will bring another woman then what will I do? I do not have anybody here to whom I can contact or tell anything.

–23 years, Muslim, education seventh standard, married 5 years, nuclear family, two children

Help-Seeking for Abuse and Mistreatment.

The vast majority of help-seeking involved use of informal support sources, with few participants expressing awareness of formal support services. Among those who were aware of such services, some participants also described barriers to, and experiences with, formal services for violence support.

Informal help.

Informal help was the dominant form of help received by participants, most commonly from family members and neighbors in close proximity. Informal help was often prompted by instances of violence that were uniquely severe or public, such that participants were unable to hide what they were experiencing.

Help From Neighbors and Family During Partner Abuse

The help provided by neighbors and family predominantly included assistance such as food, shelter, and occasionally intervention with the perpetrator(s). At times, these resources did not provide sufficient protection from further violence.

When I was pregnant for the second time then a fight took place between me and my husband. He pushed me with his hand and I fell on the wall. That time my neighbors gave me shelter for two days. They also advised my husband but he did not listen.

–23 years, Muslim, education seventh standard, married 5 years, nuclear family, two children

The sense of shame that kept women silent about their experiences also emerged in their experiences of receiving help from well-meaning neighbors.

Daily we were having fight. Sometimes I was feeling so much hungry and my husband and his 1^st wife would not allow me to eat. I was only given one chapatti and rice with dal. No vegetable, nothing so I also got very angry and started abusing my husband. And we had fights. Afternoon time also she used to give me stale food. (Respondent starts to cry). You must be knowing women feel so much hungry when she is pregnant, she needs love and support from the family members. And I was getting mistreatment only. Sometimes my neighbors gave me proper meal. I felt so much ashamed that time.

–17 years, Hindu, married 1 year, education ninth standard, nuclear family, one child

This sense of shame was particularly pronounced when the help provided did not curtail the abuser, and instead exposed concerned neighbors and family members to violence.

As I told you, now days he always gets irritated on me. After the fight with my mother-in-law and father-in-law, I came to my sister’s place. At that time he had come to meet me. He just came out of the house and started abusing me and my sister. It was disgraceful for me and my family.

–24 years, Muslim, married 6 years, education higher secondary, joint family, two children

Formal help.

Overall, participant awareness of formal support services was quite low, and actual use of services even lower, resulting in limited qualitative data concerning formal support services. Many that were aware of such services indicated minimal intention to use them. Unfortunately, those who had sought formal services often reported negative experiences. Use of health services, or reliance on health-related support programs did not emerge in discussion.

Limited Awareness and Intention to Use NGOs

While knowledge of existing support programs was limited, some participants also described a desire for such programs as illustrated by the following participant:

No, I don’t know about any woman organization, but I have problem from my husband. If you give me suggestion, sister then it will be very helpful.

–24 years, Muslim, education fourth standard, married 7 years, joint family, two children

In contrast, those who were aware of formal services often described a futility in seeking their support, reflecting a sense that the services have a limited capacity for intervening and creating change, particularly relative to the personal obligation to seeing their own circumstances through.

I cannot do anything. I know some organizations working for women’s problems. But nothing happens; everybody has to see his own life. [Interviewer: Have you taken help from this kind of social organization?] No, I have only heard about it that there are these kinds of organizations who help women. They help to get their answer, that’s all.

–23 years, Muslim, education seventh standard, married 5 years, nuclear family, two children

Negative Experiences

For those who had sought formal support services, negative experiences created a sense of resignation and hopelessness, often stemming from a sense that the services were ineffective in reducing violence.

It was really hard time for me. Actually I went to the one N.G.O for help but there is no difference in the situation. Now I am become very world weary.

–28 years, Hindu, married 3 years, education secondary, joint family, one child

Similar concerns were raised for police interventions; with the few participants who had gone to police for support turned away based on police perception that it is a domestic issue.

Interviewer: As you told me that he usually ran away, so did you ever file a complaint against him in the police station? Respondent: Yeah, once we went there also but they were not ready to file it. They said it’s the matter between husband and wife. We couldn’t do anything in this issue. See today I brought his photo also.

–24 years, Muslim, married 8 years, education sixth in Urdu medium, nuclear family, three children

Quantitative Results.

Demographic characteristics and details of experiences of abuse have been reported elsewhere (Raj et al., 2011). Just over one in three survey participants (36.0%) reported physical or sexual violence by husbands in the year prior to pregnancy, or during the pregnancy or postpartum period (Table 1). Help-seeking was uncommon among participants reporting abuse, with only 31.7% of abused women reporting that they told someone about their experiences or sought help during their most recent experience of violence. Corroborating the qualitative findings, the most common sources of help were family members including parents (48.3%), in-laws (17.0%), and neighbors (14.4%).

Table 1.

Abuse Prevalence and Help-Seeking Experiences Among Female Survey Participants (n = 1,038).

	% (n)
Experienced physical or sexual abuse from husband	36.0 (374)
The last time you experienced severe violence or harassment; did you tell anyone or seek any help? (n = 372)	31.7 (118)
If yes, whom did you tell or seek help from? (n = 118)
Parents	48.3 (57)
In-laws	17.0 (20)
Neighbor	14.4 (17)
Sister or brother	9.3 (11)
Friend	7.6 (9)
Police	1.7 (2)
Lawyer	0.9 (1)
Community Health Worker (CHW)	0.9 (1)
Others from natal home	0
Community leader	0
Mahila Mandal (violence support NGO)	0
Doctor/nurse	0

Across the entire sample, few women (4.7%) had been screened for violence in the health care setting (Table 2). Over two thirds (67.1%) of participants indicated a willingness to share such experiences with a health care professional if asked, and a similar proportion (67.9%) indicated wishing to receive information or assistance from a doctor or nurse if faced with abuse. When participants were asked to consider the helpfulness of a series of potential clinic-based violence-related interventions, crisis counseling with safety planning emerged as the most helpful, with over 90% endorsing this intervention. Participants were generally receptive to other types of interventions that were proposed based on the qualitative findings and the study team’s expertise, including couples or family counseling to reduce conflict and improve communication (87.9%), classes about perinatal health for husbands and wives (87.1%), classes for wives and in-laws about perinatal health (86.9%), and batterer intervention in the form of classes for husbands to reduce abuse perpetration (86.5%). Almost two thirds (64.8%) of participants indicated that a shelter for women and children would be helpful.

Table 2.

Health Care-Based Screening Experiences and Preferences Among Female Survey Participants (n = 1,038).

	% (n) Indicating yes
Has a doctor or nurse ever asked you about violence/conflict?	4.7 (49)
Would you be willing to tell a doctor or nurse about violence/conflict?	67.1 (696)
Would you want to receive information or assistance from a doctor or nurse to help with violence/conflict?	67.9 (705)
Would it be helpful if the clinic offered women who were being abused at home …
… counseling services for only the woman to receive support for conflict they face in the home and provide them with ways to increase their safety?	90.9 (943)
… counseling services for women with their husbands or in-laws to reduce conflicts and improve marital communication?	87.9 (912)
… a class for pregnant women and their husbands to teach men about the health needs of pregnant women and new mothers?	87.1 (904)
… a class for pregnant women and their in-laws to teach in-laws about the needs to relieve pregnant women and new mothers of some household burdens, nutrition needs of pregnant women, etc.?	86.9 (902)
… a class for husbands to teach them how to be nonabusive in their interactions with their wives	86.5 (898)
… help to find a place for her and her children to go if she were in danger from her husband or in-laws?	64.8 (673)

Discussion

Findings from this mixed-methods study demonstrate significant barriers to abuse-related help-seeking among urban perinatal women receiving health services in Mumbai, India. Only one in three participants sought external help during their most recent experience of violence. Informal help sources, including family members and neighbors, emerged as the primary support system in situations of abuse, providing food and shelter, and occasionally intervention with abusers. Quantitative findings suggest the acceptability of clinic-based screening for violence, and suggest the utility of a multipronged approach to violence intervention that blends both formal and informal sectors to protect the health and safety of perinatal women suffering abuse in this South Asian community.

Consistent with prior research not specific to the perinatal period (Bauer, Rodriguez, Quiroga, & Flores-Ortiz, 2000; Naved et al., 2006), qualitative findings illustrate that the silence with which women respond to abuse is characterized by resignation to abuse and perceptions of limited power within the relationship, as well as significant social barriers to disclosure. Findings echo and extend prior research from South Asia as well as Asian immigrant communities illustrating significant barriers to help-seeking in response to violence (Bauer et al., 2000; Lee & Hadeed, 2009; Naved et al., 2006; Raj & Silverman, 2007). Reasons for nondisclosure were not assessed quantitatively; however, quantitative evidence that over two thirds of abuse survivors do not disclose such abuse suggests the pervasiveness of the barriers to disclosure.

Among those who do disclose abuse, the vast majority relied on informal support sources, consistent with prior evidence (Ahmad et al., 2009; Naved et al., 2006; Raj & Silverman, 2007). Current qualitative evidence suggests a low level of awareness of existing formal support services; consistent with that which has been demonstrated elsewhere (Ahmad et al., 2009). Coupled with the perception that formal services are of limited value in reducing violence, lack of awareness of existing services may in part explain women’s reliance on informal support sources.

New to this body of work, quantitative findings suggest the acceptability and potential utility of healthcare-based interventions in helping abused women cope with abuse and mistreatment. While few participants (<5%) reported being screened for violence in the health care setting, most (67%) would be willing to disclose abuse if asked. Health facilities may serve as a vital link in both providing necessary services as well as serving as a referral point for other necessary interventions. Participants endorsed crisis counseling and safety planning as the most helpful type of intervention (90.9%); screening patients and providing referrals could enable access to this type of support program. Given the demonstrated need to ensure access to information about violence support services even in the absence of disclosure (Chang et al., 2005), making information available in health settings could similarly raise awareness about services and promote access. Emergency crisis shelters were only endorsed by 64%, perhaps reflecting social and cultural pressures to preserve the familial unit even in cases of violence (Bauer et al., 2000; Lee & Hadeed, 2009). In light of qualitative evidence speaking to the role of the informal sector, and the need for more responsive formal sector support services, survey items concerning possible intervention approaches were limited to those that could be developed for the formal sector.

Taken together, findings provide direction for violence support interventions for perinatal women in urban Mumbai. A multipronged approach that blends strengthening the existing informal support system to allow women more comfort and support in disclosing violence, coupled with interventions that build on the health care system through screening and support, may be promising in this setting. An important structural goal is creating a climate that allows for nonstigmatizing disclosure of abuse and help-seeking from both formal services and informal support systems. Public awareness efforts to guide family members and neighbors to respond appropriately to violence may be a mechanism to support victims and facilitate access to services. Similarly, further work is needed to develop and test health care—based interventions within this setting. Efforts to integrate screening, brief intervention, and referral to services are underway in multiple international settings (Joyner & Mash, 2011; Kataoka, Yaju, Eto, & Horiuchi, 2010; Laisser, Nystrom, Lindmark, Lugina, & Emmelin, 2011; Miller et al., 2011). Current evidence that screening is acceptable to this population suggests the potential value in its inclusion in multipronged violence intervention efforts in urban India.

Additional limitations that should be considered include the use of a single geographic area; the extent to which findings generalize to rural areas or other urban settings elsewhere in India is unclear. Similarly, reliance on a postpartum sample of women who successfully sought and obtained immunizations for their infants, and agreed to participate precludes conclusions about nonperinatal women and those not seeking immunization of their infants. Findings concerning help-seeking should be considered in light of the response rate; it is possible that those reluctant to participate could be similarly less likely to disclose abuse and seek help, rendering our estimates of abuse conservative. Findings may be particularly useful in informing clinic-based interventions for this important population. The qualitative phase was primarily designed to understand abuse patterns, perpetrators, and responses in this setting; discussion of clinic-based help-seeking did not emerge naturally, nor were participants asked to focus specifically on this area. So too, the quantitative formative inquiry concerning appropriate and acceptable intervention approaches focused solely on possible formal sector services based on the gap in formal services identified in the quantitative phase. Quantitative data concerning informal sector support were limited to the abuse disclosure overall, and source of help sought. These priorities reflect the mixed-methods process, the sequential nature of the study, and space considerations given the large study in which this inquiry was embedded. Further research to clarify the role and appropriate blend of formal and informal support services will benefit from more complete blending of inquiry regarding each support source. Further formative research is needed to guide the development of multipronged approaches to violence intervention. Given the relatively small sample size, qualitative findings are descriptive and do not enable comparisons across demographic or other groupings; further research may be able to disentangle these nuances. While research assistants were trained in soliciting sensitive information including issues of domestic violence, the use of an interviewer-administered survey may have introduced reliability and social desirability issues. Finally, given the low levels of awareness of formal help sources, and limited utilization, the qualitative findings in this domain are limited.

Findings clearly illustrate significant barriers to abuse-related help-seeking among perinatal women in this urban Indian community. Moreover, they underscore the need for a multipronged approach to violence intervention and demonstrate the acceptability of health care—based efforts within such an approach. Further community-participatory research is needed to develop and test interventions so as to protect women’s health and well-being in this uniquely affected region.

Biography

Michele R. Decker, ScD, MPH, is an Assistant Professor in the Department of Population, Family and Reproductive Health at the Johns Hopkins Bloomberg School of Public Health. Trained as a social epidemiologist, her research addresses gender-based violence, and its sexual and reproductive health implications, including those related to HIV. Much of this work is focused on marginalized populations including urban women, adolescents and those involved in transactional sex or sex work.

Saritha Nair, PhD, a demographer by training, is a Socio-Behavioral Scientist at the National Institute for Research in Reproductive Health in Mumbai, India. Dr. Nair works in the areas of sexual and reproductive health, gender issues and male involvement in reproductive health.

Niranjan Saggurti, PhD, is an Associate with the Population Council’s HIV and AIDS program, based in New Delhi, India. He is a demographer and biostatistician by training, and has approximately 15 years’ experience conducting research on sexual and reproductive health, HIV/STI, gender-based violence, and the intersection of these issues in India and other highly affected settings.

Bushra Sabri, PhD, MSW, is a post-doctoral fellow at Johns Hopkins University. She is a social worker by training, and her professional experiences include work in mental health, substance abuse and victim service settings. Dr. Sabri’s research focuses on interpersonal violence, health outcomes, and racial/ethnic and gender disparities in health.

Meghna Jethva holds an MA in Clinical Psychology, and served as a research assistant with the National Institute for Research in Reproductive Health in Mumbai, India with a focus on gender-based violence and sexual and reproductive health.

Anita Raj, PhD, is a Professor in the Division of Global Public Health, Department of Medicine at the University of California San Diego School of Medicine. She is a developmental psychologist by training, with over 20 years of experience conducting research on sexual and reproductive health/HIV/STI, gender-based violence, substance misuse and abuse, and the intersection of these issues. Her current research is based in North America, South Asia and Russia.

Balaiah Donta, PhD, MSc, is a statistician by training and is Scientist F (Deputy Director Sr. Grade) at the National Institute for Research in Reproductive Health in Mumbai India. His research interests include sampling techniques and large scale sample surveys, enhancing male partnership in sexual and reproductive health, and expanding contraceptive choices.

Jay G. Silverman, Ph.D., is a Professor in the Division of Global Public Health, Department of Medicine, at the University of California, San Diego, School of Medicine. Dr. Silverman’s research focuses on the health-related consequences and prevention of gender-based violence against adolescent and adult women (e.g., intimate partner violence, sexual assault, sex trafficking) globally. His recent research has centered on implications of intimate partner violence and commercial sexual exploitation for heterosexual HIV/STI infection and for reproductive and child health.