Physician Practices In Response To Intimate Partner Violence In Southern India: Insights From A Qualitative Study

Abstract

Health care providers in India are often the only institutional contact for women experiencing intimate partner violence (IPV), a pervasive public health problem with adverse health outcomes. This qualitative study was among the first to examine Indian primary care physicians’ IPV practices. Between July 2007 and January 2008, 30 in-depth interviews were conducted with physicians serving low-to-middle income women aged 18–30 in southern India. A modified grounded theory approach was used for data collection and analysis. Study findings revealed a distinct subset of ‘physician champions’ who responded to IPV more consistently, informed women of their rights, and facilitated their utilization of support services. Findings also offered insights into physicians’ ability to identify indications of IPV and use of potentially culturally appropriate practices to respond to IPV, even without training. However, physician practices were mediated by individual attitudes. Although not generalizable, findings offer some useful lessons which may be transferable for adaptation to other settings. A potential starting point is to study physicians’ current practices, focusing on their safety and efficacy, as well as enhancing these practices through appropriate training. Further research is also needed on women’s perspectives on the appropriateness of physicians’ practices, and women’s recommendations for IPV intervention strategies.

Introduction

The widespread prevalence of intimate partner violence (IPV) has been well-established the world over, as has IPV’s adverse impact on women’s physical, mental, and reproductive health (Garcia-Moreno et al. 2006). According to the most recent national survey conducted in 2005–06 among reproductive-age women in India, 27.8% reported experiencing physical violence from an intimate partner, and 7.7% reported experiencing physical and sexual violence (Silverman et al. 2008). Prevalence estimates were higher (27% to 50%) in Karnataka state (Krishnan et al. 2010; Rocca et al. 2009; Chandrasekaran et al. 2007). Moreover, exposure to IPV increases women’s risk of physical injuries, chronic pain, unintended pregnancies, miscarriages, and sexually transmitted diseases, including HIV/AIDS (Jejeebhoy 1998; Newmann et al. 2000; Stephenson, Koenig, and Ahmed 2006; 2006), and mental health disorders, including attempted suicide (Patel et al. 2006; Varma et al. 2007).

The foregoing evidence has prompted researchers to highlight the need for health care providers (HCPs) to prevent the escalation of IPV and to manage the resulting adverse health outcomes (Jejeebhoy 1998; Muthal-Rathore, Tripathi, and Arora 2002; Prasad 1999). Across South Asia women who experience IPV rarely seek help from the police or other support agencies (Andersson et al. 2009; Naved et al. 2006), (Panchanadeswaran and Koverola 2005). However, women’s utilization of healthcare in general is increasing (National Family Health Survey 2005–06), and increased interaction with HCPs raises women’s awareness of services available to them and may favorably alter their future health care-seeking practices (Sugathan, Misra, and Retherford 2001; Ram and Singh 2006). These data suggest that HCPs may have a unique opportunity to assist women experiencing IPV.

Despite this evidence, limited published research has examined Indian HCPs practices in response to IPV. During their education, most HCPs do not receive training on gender sensitization or responding to IPV(Majumdar 2004). No guidelines exist on how or when providers should intervene against IPV, and little information exists on what specific actions are warranted in the Indian context (Deosthali and Malik 2003). Other responses to IPV in India range from legislation to protect and enforce women’s rights to various support services offered by non-governmental organizations (NGOs) and state governments (Poonacha and Pandey 1999). Such services are limited in number and reach, and vary in quality, but are increasingly available in cities. Nevertheless, most women are unaware of these services and resort to them only in extreme—typically near-fatal—situations (Panchanadeswaran and Koverola 2005). HCPs, as a consistent point of institutional contact, might play an important role in educating women about these services and encouraging utilization.

This qualitative study was among the first in India to examine primary care physicians’ practices in response to IPV, with some attention also focused on the underlying attitudes in which they were partially grounded.

Methods

Study setting and sampling

This study was ancillary to the Samata Health Study (SHS), a longitudinal observational study among married women between the ages of 16 and 25 years, residing in two urban poor communities in Bangalore, the capital of Karnataka state. The SHS examined the relationships between economic and social factors and women’s reported experience of IPV and susceptibility to sexually transmitted infections. It was conducted in collaboration with the city municipal corporation (Bruhath Bengaluru City Corporation, BBMP), and for logistical reasons restricted to Bangalore’s east zone. In response to the high prevalence of IPV observed in SHS (Krishnan et al. 2010; Rocca et al. 2009), the current formative research was undertaken to examine primary care physicians’ practices in response to IPV.

Given the dearth of data on this topic in India, we used purposive sampling, a sampling strategy frequently employed in qualitative research to identify participants with detailed knowledge to provide rich illustrations on the topic of interest and to capture the full range of practices in place. We purposefully oversampled physicians specializing in women’s health issues and with large female clientele as such physicians were more likely than their peers to encounter women’s health issues. Study eligibility criteria included practicing medicine in Bangalore’s east zone, where the SHS was conducted, and specializing in family medicine, internal medicine, obstetrics and gynecology, or pediatrics. We also included roughly equal numbers of physicians from the public and private sectors. The latter were included because even among lower-income families in urban India, use of private healthcare services for primary care tends to be high despite the higher costs incurred (Peters, Rao, and Fryatt 2003).

Participant selection began by obtaining publicly available lists of registered public and private sector primary care physicians working in Bangalore’s east zone. Trained project staff then conducted community walks to locate these physicians and verify study eligibility using checklists. This involved visiting each physician’s clinic at multiple times of the day to verify: (1) the size of the female client load by observing patients in the waiting area, and (2) the physician’s specialty by observing clinic signs posted. In addition, the staff had informal conversations with key stakeholders—BBMP senior staff, public health center staff, community residents, and pharmacists—to gather their perspectives on the physicians most frequently used by low-to-middle income women in the 18–30-year age group. Through this process a total of 52 potential participants were identified and screened for eligibility. Eight did not meet study eligibility criteria because their self-reported female clientele was less than 50%. Among the remaining 44 (85%) who were eligible, three declined participation due to lack of time, and three were unwilling to sign the informed consent form. Another six verbally consented to participate but were eliminated after failing to respond to repeated scheduling attempts, and two were not re-contacted after an initial screening visit because theoretical saturation (Strauss and Corbin 1998) had been attained after 30 interviews, and no new data collection was required. Thus, in total, 44 participants were found to be eligible, and 68% of those eligible participated.

Data Collection

We followed a modified grounded theory approach to data collection, dividing it into two phases—phase one between July and August 2007 and phase two between October 2007 and January 2008—to enable interaction between data collection and preliminary data analyses (Strauss and Corbin 1998). Sixteen interviews were conducted in phase 1. Preliminary analysis and open coding followed, along with minor modifications to the study guide to ensure that we were thoroughly exploring the properties and dimensions of the emerging concepts. Fourteen interviews were completed in phase 2, and, consistent with grounded theory principles, data collection was terminated after 30 in-depth interviews were completed because theoretical data saturation (the point at which no new themes were emerging) (Strauss and Corbin 1998) had been attained. Screening and interviews were conducted by the lead author and a research assistant in a private location at each physician’s clinic. Before each interview, participants received a detailed description of the study, its purpose and objectives, and reviewed the consent form, with the opportunity to ask for clarifications. Interviews were conducted after written and signed informed consent was obtained and lasted approximately 90 minutes. A semi-structured in-depth interview guide was developed which drew on previous studies (Velzeboer 2003; Center for Enquiry into Health and Allied Themes 2007) and included questions on physician knowledge and practices relating to IPV. Physicians were also asked closed-ended questions about their practice, training, and clientele. Interviews were conducted in English, audio-taped, and transcribed. Participants did not receive compensation for participation. Human subjects approval was obtained from the Indian Institute of Management, Bangalore, the University of California, Berkeley, and RTI International.

Data Analysis

Pursuant to the grounded theory approach, multiple rounds of coding and interpretation were employed, primarily by the lead author. Open coding generated initial concepts, which were tested and further explored in subsequent interviews. These concepts were shared with senior SHS staff through memos and in-depth discussions to ensure that individual assumptions and biases were not unduly influencing the analytical process and that necessary attention was being paid to alternative explanations and interpretations. Further, to ensure reliability, the lead and second authors independently coded four transcripts, and each developed a codebook. Through collaborative reconciliation of the two authors, the two codebooks were merged to create a master codebook from which redundant codes were removed, and codes and their meanings were discussed and refined to ensure consistent interpretation of codes. Any differences of opinion between authors regarding codes and their definitions were reconciled by going back to the initial four transcripts, with all final decisions on codes being based on participants’ words and phrases. At this stage, some codes were also merged into broader and more meaningful categories to arrive at a master code list. For example, the codes ‘history taking’, ‘probing’, and ‘indirect questioning’ were collapsed into the category titled ‘screening practices’. The master code list was then applied to the remaining interviews by the lead author. A phase of selective coding (the grounded theory process of integrating and refining theory) followed to explore how the different concepts and categories were interconnected (Strauss and Corbin 1998). This entailed developing memos and diagrams to describe in detail the emerging categories and their properties and dimensions, and to group the categories together and study how the different categories fit together. The data were organized with the support of the Atlas.ti (version 5.5) software package.

Results

Background characteristics

Eighty percent of the sample was female (Table 1). More than 50% of the sample specialized in general medicine, over a third in obstetrics and gynecology, and the rest in internal medicine or pediatrics. The majority had over 10 years of experience. Less than one quarter reported receiving training on gender sensitization or on responding to IPV reported by patients.

Table 1.

Socio-demographic characteristics of sample

Variable Description	N (%)
Practice Setting
Public clinic (urban health center or maternity hospital and 2 respondents from regional public hospitals)	14 (47)
Private clinic (15 from private clinics and 1 respondent from a private hospital)	16 (53)
Gender
Female	24(80)
Male	6 (20)
Specialty
General Physician	16 (54)
Gynecologists	11 (36)
Other	3 (10)
Additional Administrative Duties
Yes	14 (47)
No	16 (53)
Years in practice
5–10	11 (37)
> 10 years	19 (63)
Received Training on gender sensitization and IPV
Yes	7 (23)
No	23 (77)

Physician practices in response to IPV

Physicians’ IPV response can be broadly categorized into four core practices (Figure 1): (1) noticing indications of IPV; (2) screening and assessment; (3) counseling; and (4) referral. All physicians, however, did not follow these practices in a step-wise progression, nor did they employ all four practices consistently. Instead, physician response involved an iterative and evolving process of reflecting on what was seen or heard, attempting to make a correct diagnosis, and constructing an appropriate response.

Figure 1.

Physicians’ four core practices in response to IPV

The analysis also revealed a subset of physicians (n=10) whose IPV practices involved a higher level of engagement. These physicians reported screening more routinely, informing women of their rights and choices, and facilitating their utilization of support services. Some of these physicians reported having trained support staff in responding to IPV, and routinely networking with support organizations. These physicians attributed their deep commitment to promoting women’s rights and well-being to training on responding to IPV and/or influential role models, including faculty, senior students, and former colleagues. The majority among this subset were from the public health sector. Henceforth, this subset of respondents will be referred to as “Physician Champions,” and the remaining physicians (n=20) as “Non-Champion Physicians.” Below we describe the four practices, indicating variations in practice, wherever substantial, between these two subsets of physicians.

1. Noticing indications of IPV

All 30 respondents reported following this practice, explaining that it occurred subconsciously without any additional effort from them and was interspersed with routine clinical procedures. Most frequently observed were physical injuries ranging from minor cuts, bruises, and burn marks to more severe conditions of broken limbs, fractures, and swollen or black eyes. The majority also reported observing “vague symptoms” that included body ache, abdominal pain, headache, back ache, giddiness, insomnia, loss of appetite, and “general weakness.” They explained that such symptoms prompted them to take additional steps: check patient’s vital signs and order clinical tests. If patients returned with the same complaints even after test results were normal, or after sustained use of medication, respondents explained that they would begin to suspect IPV.

Fifty percent of respondents reported noticing signs of sexual violence, which they described as women’s inability to make independent reproductive choices and experiences of coercive sex. Respondents reported seeing patients who were not allowed by their partners or extended families to “follow family planning” and, at the other extreme, patients who were forced to terminate their pregnancies to continue working and contributing to the household income. Physician Champions’ clearly described the observations that led them to appreciate patients’ experiences of coercive sex. One Champion reported noticing strange requests from a patient for medication to render her husband impotent. In-depth conversations with this patient revealed that she was experiencing coercive sex “three to four times a day,” and “if she refused she was beaten up.” Another Champion reported noticing that some new mothers feigned “giddiness or vomiting” to be re-admitted to the clinic after delivery. Inquiry revealed that these patients sought temporary respite from the coercive sex awaiting them at home.

Physician Champions additionally reported noticing signs of psychological distress among their patients such as being uneasy, withdrawn, fearful or unhappy. They further explained that even slight shifts in a patient’s body language could indicate that the patient wants to disclose something and desires more privacy. Many of these Champions reported having taken creative steps to accommodate this need, such as instituting a no attendant policy or requesting attendants to “wait outside [during the physical examination] because [of the] small room”.

2. Screening and assessment for IPV

Perceived as interrelated processes, screening and assessment (referred to hereafter as screening), were selective and entirely at the physician’s discretion. Most respondents employed indirect questioning which entailed “leading” yet non-threatening questions about the patient’s personal life and relationships, and whether she was “happy at home” or “having any problem[s]”. Based on the patient’s disclosure, respondents assessed the situation, and in some cases transitioned to direct screening. They explained, however, that they used direct questioning in cases of clearly visible physical injuries.

Respondents’ “orientations” toward screening can broadly be classified under three categories: (1) universal screening, (2) trigger-driven screening, and (3) choosing not to screen. Two Physician Champions reported practicing universal screening, where they routinely asked all patients if they were being physically or psychologically harassed at home. They believed that universal screening ensured greater identification of IPV and would help engender rights-consciousness among women and a greater awareness of support services. As one Champion said:

“…if we identify [IPV] in the beginning even before it becomes serious then [women] know it is okay [to talk about this] … we make them understand that first you are a human being, you also have rights… so that tomorrow if they face [IPV] they understand … [we should] not accept this.” (Public sector physician)

The remaining eight Physicians Champions and most Non-Champion Physicians (n=19) primarily screened in response to certain triggers. All screened when they observed physical injuries or patients’ disclosures, but champions reported additionally screening in response to “vague symptoms” and signs of psychological distress. Respondents further explained that they occasionally delayed screening until later visits to establish better physician-patient familiarity and patterns in patient behavior or symptoms.

The final orientation was choosing not to screen. One respondent reported never screening, believing that women could choose to not “tolerate IPV” and that outside intervention was futile for those who accepted IPV. Other respondents described specific instances in which they chose not to screen. About 50% believed that domestic conflict occurs in all families, and medical professionals should not screen for “small” problems (e.g., financial difficulties and differences of opinion within the couple), given physicians’ limited knowledge of the family and inability to judge right from wrong in a given situation. Another attitude that was reported by many respondents was that IPV was higher among lower socioeconomic status (SES), and all respondents were asked indirectly how this attitude influenced their practice. Only one respondent, who otherwise described a nuanced and detailed screening practice, candidly admitted that she avoided screening women belonging to lower SES:

“If she is a hard core slum woman [lower SES] even if she [reports] severe symptoms, I still don’t empathize and don’t converse much. Sometimes I feel she deserves it for her way of behavior, her lifestyle, she also is quite aggressive… and into habits like alcoholism, and they are all the time arguing with the husbands unreasonably…whereas with a housewife and a lower middle-class [woman], I identify more…and I feel that their problems are more genuine.” (Public sector physician)

3. Counseling

Almost 80% reported counseling patients who were identified as experiencing IPV. Counseling encompassed a breadth of activities including active listening, consoling, validating the injustice of IPV, encouraging women to be hopeful, advising against suicide, and disseminating other harm-reduction messages. The extent to which these activities were practiced differed widely among respondents, with Champions performing more activities on average.

An examination of the messages that respondents reported giving to their patients further highlights the variation in practice between Physician Champions and Non-Champion Physicians. Most Champions’ messages reflected their belief in gender equality and human rights. They felt that women and men deserve equal rights, and it was “physicians’ duty as human beings” to validate the injustice that women were experiencing, and to inform them of their rights, choices, and options for support services. Explaining how she encouraged women to defend themselves against IPV while deliberating on the implications of their actions, one Champion said:

“…make sure that you prevent it [IPV]. Whether you are going to retaliate…or you are going to run out of the house… you are going to scare your husband by saying that the doctor has told me to complain, it is your strategy but make sure that you are not hurt anymore.” (Public sector physician)

In contrast, most Non-Champion Physicians’ (n=15) messages reflected their beliefs in traditional norms prevalent in Indian society. These include tacitly ignoring IPV in certain situations, accepting a subordinate position for women, and propagating a view of gender roles in which women bear primary responsibility for the household and childrearing and should “adjust to the family problems.”

The remaining five Non-Champion Physicians seemed ambivalent about responding to IPV. Their messages reflect internal conflict between concepts of human rights and equality on the one hand, and societal beliefs that suppress women on the other. For example, they reported counseling patients to “be bold” and “leave [their] husbands”, but then immediately followed up by saying that “women have to put up with everything for the sake of their children”.

Respondents reported employing two counseling strategies. One was to counsel husbands. While all respondents reported encouraging women to bring their husbands to the clinic, opportunities to interact with husbands were rare, and only 14 respondents (4 Champions and 10 Non-Champion Physicians) could describe such interactions. Five respondents (two among them Champions) further emphasized centering the conversations with husbands on the connections between IPV and women’s adverse health, toward minimizing women’s risk of future IPV. For example, one respondent said:

“… I stress more on the fact that the wife is not doing well in her pregnancy, that she definitely needs to get all the tests done, and she needs a better diet because she is not putting on weight… I don’t tell them…I heard that you’re beating your wife, nothing like that. I don’t want them to feel threatened in any way and think that the wife is sneaking or telling me things.” (Private sector physician)

Another strategy was to meet with members of the extended family. Respondents explained that that the entire family should agree with changes that the couple might be asked to make, and understand why women should not be mistreated. Only five respondents, two of them Champions, could describe such interactions.

4. Referral

None of the twenty Non-Champion Physicians referred patients to support services. They attributed low referral levels to their lack of awareness about, and the limited availability of, support services. With little or no interaction between themselves and support organizations, they reported being uneasy referring patients to unfamiliar organizations. Limited time and multiple clinical responsibilities further were cited as preventing them from investigating social services and supporting patient efforts to pursue these services.

In contrast, all Physician Champions reported referring patients to support services. They had taken the initiative to learn about organizations providing services in their communities, and took personal steps to facilitate women’s utilization of these services. One Champion reported forming a network among the non-governmental organizations (NGOs) near her clinic, enabling her to connect her patients with the organizations best suited to meet their needs.

Discussion

Findings of this study offer insights into the ability of a select sample of largely female physicians with largely female clientele to identify indications of IPV and potentially use culturally appropriate practices to address this pervasive problem. Four practices—noticing indications of IPV; screening and assessment; counseling; and referral—were frequently followed, albeit to very different degrees and through an iterative and evolving process. Findings further revealed a distinct subset of Physician Champions who responded to IPV more consistently, taking additional steps to inform women of their rights and facilitating their utilization of support services.

The limitations of this study, however, should not be minimized. Purposive sampling results in selection bias and limits generalizability of this study’s results. Such bias may have been a particular issue in this study, given the sensitive nature of IPV, and may have contributed to our more positive findings. However, in a context in which little is known about Indian physicians’ IPV practices, our formative study was aimed at exploring some practices and perceptions by a small, largely female, group of physicians in response to IPV. Our sampling approach was thus useful in uncovering in-depth information regarding such physicians’ current IPV practices. The study’s results should be somewhat transferable—a criterion in qualitative research by which rigor and utility of qualitative research may be assessed (Curry, Nembhard, and Bradley 2009). Although some argue that transferability, or the extent to which results can be applied in other settings and population groups, may be best considered on a case-by-case basis, a “rich and vigorous” presentation of findings, as well as a clear and detailed description of “the culture, context, and selection and characteristics” of participants and data collection procedures (Graneheim and Lundman 2004) can enhance opportunities for broader, analytic generalization or transferability (Firestone 1993).

Although we described variation in physician practices between Physician Champions and Non-Champion Physicians because this division in practice emerged naturally from the data, the small sample size limited other relevant comparisons. Future research is needed with larger samples to corroborate and expand on our findings e.g., with respect to differences in physician IPV practices by gender, age, specialty, and practice setting. Another limitation was the lack of double independent coding for all transcripts to ensure reliability of coding. Due to the constraints of time, we used double coding for four transcripts and developed a master code list which was then applied to the remaining interviews. Finally, all data were self-reported by respondents and therefore subject to potential recall and social desirability biases, particularly relevant given the sensitive topic under study. To help reduce these biases, interviewers received extensive training emphasizing how to avoid communicating personal values and expectations during the interview, opening the discussion of IPV with the broader term “conflict in the home,” and using the participant’s own terms (e.g., “bashing up,” “stress,” and “forced sex”) rather than IPV during the remainder of the interview.

Despite this study’s limitations, it offered some useful lessons which may be transferable. Key among these was the example of Champions’ higher level of engagement, which suggested the possibility of improving physicians’ IPV practices through training, a well-established determinant of physician IPV practices in the West (Gutmanis et al. 2007; Lapidus et al. 2002; Stayton and Duncan 2005; Short, Hadley, and Bates 2002). Seventy percent of Physician Champions worked in the public health sector, and as part of a major health system they may have had greater exposure to training and influential role models. This exposure, in turn, may have enhanced their knowledge about, and trust in the utility of, support services; perceived self-efficacy in responding to IPV; and beliefs regarding their professional responsibility to help women. Further, Champions’ practices and perspectives demonstrated personal attributes of self-reflection, self-motivation, and commitment to pursue an issue that garners little public attention. Interestingly, many of these attributes mirror those being promoted in the Western countries as characteristics of physician leaders (Holmboe et al. 2003; Mouradian and Huebner 2007). Our encounters with these physicians highlighted the potential within the Indian health system to build advocates for women’s rights, who in turn, may have the power to influence the next generation of healthcare professionals’ practices.

A second finding, while not conclusive due to the use of small and non-representative sample, suggested that even without training and guidelines physicians may be able to identify indications of IPV and, through trial and error, some have developed responses that may be culturally appropriate to the Indian social context and worth building upon. For example, most of the physicians in our small sample predominantly used indirect questioning to screen for IPV. They carefully framed their questions within patients’ broader family context, employed a personalized approached to screening, and sought to foster trusting physician-patient relationships to enable future IPV disclosure. In a society that may implicitly ignore IPV and in which women may accept IPV under certain conditions (Go et al. 2003; Jejeebhoy 1998), this approach might be preferable to universal screening with direct questioning that comes unexpectedly and may offend some women (Rodriguez et al. 1999; Roelens et al. 2006; Velzeboer 2003; Gerbert et al. 1999). The indirect screening approach is not specific to India. In Western countries, even after over twenty years of guidelines on responding to IPV being in effect, universal screening remains controversial (Rodriguez et al. 1999; Ramsay et al. 2002); and screening is typically selective, in response to specific patient attributes(Rodriguez et al. 1999; Roelens et al. 2006; Gerbert et al. 1999; Chamberlain and Perham-Hester 2002), and employs a mix of direct and indirect questioning(Gerbert et al. 1999). Further, physicians with expertise in IPV advocate reorienting universal screening goals away from trying to “fix” the problem and toward building trusting physician-patient relationships and providing patients with opportunities for disclosure (Gerbert et al. 1999).

Another potentially culturally appropriate response, although not conclusive, was some respondents’ counseling strategy of engaging with and winning the trust of husbands and families. This response, particularly in a patriarchal society such as India (Raj et al. 2006; Jejeebhoy and Cook 1997), may in fact be necessary for more effective IPV prevention efforts. Unlike in many Western countries, in India, such practices do not violate laws or ethics rules concerning patient privacy. Moreover, members of a woman’s marital family often are responsible for instigating conflict within a couple or even directly perpetrating IPV (Hyder, Noor, and Tsui 2007; Raj et al. 2006; Clark et al. 2010). In contrast, members of the natal family can be a source of support, which, in turn, may lower women’s risk of IPV (Clark et al. 2010). However, to avoid further jeopardizing women, physicians must receive extensive training on the appropriate use of this mode of intervention, including in particular skills on how to ensure patient safety, solicit information on family dynamics, and tailor counseling messages to each family’s circumstances.

Finally, although a detailed analysis of attitudes was beyond the scope of this study, the findings in our small study sample revealed that physician practices were mediated by individual attitudes. The screening of women from the lowest SES may be suppressed, for example, by a belief that such women’s lifestyles and behavior were responsible for their problems. While only one respondent stated this belief and the influence on her practice explicitly, it may be a more generally held practice that other respondents (typically belonging to a higher SES than their patients) were unwilling to disclose during the interviews due to social desirability bias. Other examples include views concerning what problems are too “small” to warrant attention; counseling messages reinforcing traditional gender roles; and acceptance of IPV under certain circumstances. Our finding of the unfavorable influence of attitudes on physician practices corroborate findings from previous studies (Nicolaidis, Curry, and Gerrity 2005; Garimella et al. 2002). Future research, including participant observation, is necessary to delineate physician norms and attitudes, and examine how these influence physician practice to ensure that select groups of women are not being intentionally or unconsciously neglected.

Conclusions

Although not generalizable, the present study findings offer a unique window into IPV practices among a small sample of largely female Indian physicians, including identification of this pervasive problem and use of potentially culturally appropriate practices to address it. The existence of ‘Physician Champions’ further points up the potential for building advocates for women’s rights within the Indian healthcare system.

A starting place for health sector IPV interventions might involve building on existing IPV practices. Physicians’ efforts could be supported with training that enhances screening and counseling skills and their ability to connect women with sources of support. Further, because physicians’ IPV practices were found to be unfavorably influenced by their attitudes, training programs should also address negative stereotypes concerning involvement with IPV cases. To better identify these and other promising practices, a comprehensive evaluation of existing domestic and global IPV training programs for physicians is needed, with lessons learned incorporated into the development of future and widely available training programs. These programs must also be periodically updated to ensure that physicians’ IPV responses are current with ongoing efforts to increase societal knowledge and awareness about women’s rights and ill-health associated with IPV.

In parallel, future research is needed on the safety and effectiveness of physicians’ current IPV practices, as well as women’s perspectives on the appropriateness of these practices and suggestions for health facility-based IPV interventions that would be most beneficial them.

Another avenue, shown to be promising elsewhere (Naved et al. 2009), is training for all support staff—such as paramedics, nurses, social workers, counselors, and outreach workers—so they can assist physicians in implementing core IPV prevention and treatment practices, including detailed case management support. Given the large client volumes, multiple responsibilities, and time shortages that primary care physicians in India face, a concerted effort by teams of HCPs may over time facilitate earlier detection, assessment, and referral for IPV.

Finally, and in tandem with a concerted health-sector response to IPV, the quality and reach of support services should be enhanced. Mechanisms are needed to enable the creation and strengthening of linkages between HCPs and these support services. Such linkages may help ensure that all providers recognize and respect each others’ unique skills, enabling a collaborative and timely response to IPV in the future.