UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Untying our hands : the social context of nursing in relation to violence against women Varcoe, Colleen Marie 1997

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1997-251799.pdf [ 21.8MB ]
JSON: 831-1.0088320.json
JSON-LD: 831-1.0088320-ld.json
RDF/XML (Pretty): 831-1.0088320-rdf.xml
RDF/JSON: 831-1.0088320-rdf.json
Turtle: 831-1.0088320-turtle.txt
N-Triples: 831-1.0088320-rdf-ntriples.txt
Original Record: 831-1.0088320-source.json
Full Text

Full Text

UNTYING OUR HANDS: T H E SOCIAL CONTEXT OF NURSING IN RELATION TO VIOLENCE AGAINST WOMEN  by Colleen Marie Varcoe R.N., The Royal Columbian Hospital School of Nursing, 1973 B.S.N., The University of British Columbia, 1979 M.Ed., The University of British Columbia, 1987 M.S.N, The University of British Columbia, 1994 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN NURSING in THE FACULTY OF GRADUATE STUDIES School of Nursing We accept this thesis as^p^forming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA September, 1997 © Colleen Marie Varcoe, 1997  In presenting this thesis in partial fulfilment  of the requirements for an advanced  degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department  or  by his or  her  representatives.  It  is understood that  copying or  publication of this thesis for financial gain shall not be allowed without my written permission.  Department of  /lufyin  The University of British Columbia Vancouver, Canada  DE-6 (2/88)  ABSTRACT Violence against women and children is acknowledged to be a health problem of epidemic proportions, yet the health care response has been inadequate at best. This ethnographic study examined the relationship between the social context of practice and the way in which nurses recognize and respond to women who have been abused. Data collected over two years in two hospital Emergency units and communities included about 200 hours of field work, as well as interviews with 30 health care providers and five women who had been abused. These data support other research showing that violence against women is neglected within health care. Abuse was largely unrecognized, with "blatantly obvious" consequences of physical violence being most recognized. Importantly, violence was often anticipated predominantly among poor and "non-white" people. When abuse was recognized, intervention focused on the physical results of violence, and the social and emotional consequences were often ignored. When abuse was addressed, such efforts were often attempts to influence the woman toward choices that she may not want or actually have, or worse, choices that may endanger her further. However, health care providers also intervened by offering and respecting choices in a manner that was congruent with the needs of the women interviewed. Data analysis suggested that the predominant pattern of routine practice in Emergency, in which patients are efficiently processed in accordance with an ideology of scarcity, fosters the neglect of violence. It is argued that violence and abuse are neglected because the power of dominant interests is exercised through ideologies which are  congruent with neglect. As individuals work within a social context in which dominant interests shape their everyday worlds and provide the lenses through which they interpret their world and personal experiences, practices are mostly congruent within dominant interests. In order to attend to violence against women in a meaningful manner, it is argued that individuals at all levels of decision-making in society, from the corporate elite, to health care policy-makers, to nurses at the bedside, must develop a critical consciousness regarding domination, and the ways certain interests in society are served.  TABLE OF CONTENTS Page ABSTRACT  ii  T A B L E OF CONTENTS  iv  List of Figures  ix  List of Tables  x  Acknowledgments  xi  Dedication  xii  CHAPTER ONE - T H E PROBLEM  1  The Research Questions Conceptual Issues and Definition of Terms Summary  6 6 9  CHAPTER TWO - THE LITERATURE  The Scope of the Problem of Violence Against Women The Health Consequences of Violence Against Women The Social Response to Violence Against Women How Violence is Conceptualized Power, Oppression, Resistance and Perspectives on Violence The Role of the State Public Versus Private Violence The Health Care Response to Violence Against Women Rates of Identification of Women Who Have Been Battered Reasons for Non-identification Detrimental Responses to Women Who Have Been Battered Why is the Health Care Response Inadequate? The Nursing Response to Violence Against Women Nursing Research and Violence Against Women Summary of the State of Knowledge  11 13 16 17 19 23 24 26 27 30 33 36 39 41 47  CHAPTER THREE - T H E METHOD  Research Design Theoretical Perspective on Power Theoretical Perspective on Violence The Sample Emergency Nurses Other Health Care Providers  48 58 61 63 64 67  V  Women Who Have Experienced Abuse Data Collection Interviews Phase 1 Phase 2 Phase 3 Field Work Data Analysis Analysis of Interview Data Developing a Conceptual Framework Developing Theoretical Arguments Developing a Central Thesis and Reconceptualizing the Data Analysis of Field Work Data Document Analysis Pursuing Rigour Auditability Reflexivity Applicability Ethical Considerations Autonomy, Privacy and Informed Consent: Nurses Autonomy, Privacy and Informed Consent: Patients Intervention Versus Non-intervention Potential Use of Findings Care for the Participants Politics and Action Summary  68 68 69 71 72 72 73 77 79 79 81 83 83 84 84 85 85 86 87 87 91 94 95 97 97 98  CHAPTER FOUR - INTERPRETATIONS: AN OVERVIEW  Reproducing and Resisting Dominant Interests Dominant Interests Individual Enactment of Practice: Reproducing and Resisting Dominant Interests Allocation and Rationing of Resources Domination and Resistance in Emergency Practice Domination and Resistance in Relation to Violence Against Women  100 102 105 107 108 111  CHAPTER FIVE - INTERPRETATION: NURSING PRACTICE IN EMERGENCY  The Purposes of an Emergency An Ideology of Scarcity The Pace and Patterns of Practice The Predominant Pattern of Practice: Efficient Processing Stripping the Patient Down The Manageable Problem The Recorded Body  119 123 126 132 132 147 149  vi  Creating the Recorded Body Enlisting Support in Creating the Manageable Problem Processing the Patient Emptying the Stretchers The Central Conflict Practices of Congruence: Going with the Flow The Efficient Nurse The Strong Nurse Influencing Each Other The Deserving Patient Practices of Resistance Preserving the Person Supporting Each Other Influencing the Context The Patient as Agent Summary  151 154 156 159 165 173 173 176 178 179 183 183 187 188 193 194  CHAPTER SIX - INTERPRETATION: INDIVIDUAL PRACTICE  Nursing Practice in Relation to Violence: Practice is Personal Education As a Limited Source of Knowledge Popular Media as a Limited Source of Knowledge Personal Experience as a Basis for Practice Experiences of Violence at Home No Experiences of Violence at Home Experiences of Violence at Work Experiences of Violence at Work as a Limited Source of Knowledge Creating Understanding from Personal Experience ' Heightened Awareness and Understanding Through Experiences of Violence Experiences of Violence at Home as a Limited Source of Knowledge Challenging Assumptions Taking Other Perspectives. Challenging the Importance of the Context Exploring Alternative Ways of Acting Inaccessible Personal Experiences Summary  196 197 202 205 208 210 211 213 217 220 221 222 227 232 235 239 241  CHAPTER SEVEN - INTERPRETATION: RECOGNIZING VIOLENCE AND ABUSE  Recognition of Violence and Abuse The Presenting Complaint Inferences from the Patient's Appearance and Behaviour: Anticipating and Not Anticipating Abuse The Pauperization of Violence The Racialization of Violence  244 246 248 255 259  vi  The Impact of the Racialization and Pauperization of Violence Information Provided by the Patient: Disclosure and Non-disclosure of Abuse Willingness to Disclose Abuse: Health Care Provider's Perspectives Willingness to Disclose Abuse: Women's Perspectives Asking Questions: Perspectives of Health Care Providers Asking Questions: Perspectives of Patients The Influence of the Context on Recognizing Abuse: Toward Congruence The Influence of Individuals on Recognizing Abuse: Toward Resistance Congruence and Resistance: Index of Suspicion For Abuse Oblivious to Violence and Abuse Limited Knowledge of Indicators of Violence and Abuse Not Wanting to Know Not Knowing What To Do Routine Questioning Summary  262 264 268 271 275 277 279 282 284 285 286 287 289 290 297  CHAPTER EIGHT - INTERPRETATION: DEALING WITH VIOLENCE  Doing Nothing: The Undeserving Victim Whose Interests Are Served? Shifting Responsibility Influencing Choices: The Deserving Victim Disclose the Abuse Leave Your Partner Call the Police Fixing Problems Influencing Choices As Problematic The Victim Only Taking Control Ideas About Choice Limits To Choice Whose Interests Are Served? Frustration Making Matters Worse Offering Choices: The Woman with Agency Giving Up Control Listening Respecting Decisions Accounts of Success Whose Interests Are Served? Toward Untying Our Hands Summary  301 305 309 314 315 318 320 322 324 324 328 329 331 334 335 337 339 341 343 345 347 347 348 350  viii  CHAPTER NINE - UNTYING OUR HANDS: CONCLUSIONS AND IMPLICATIONS  Conclusions Ideology Scarcity in Health Care Critical Awareness at the Level of Policy Critical Awareness at the Level of Practice Deservedness Discourses of Violence Rereading Theories of Violence Further Excavating Discourses of Violence Racialization of Violence Personal Experiences Collective Practices  352 354 355 357 360 364 366 368 370 3 71 374 376  REFERENCES  379  Appendix 1: Information Letter for Nurses and Other Health Care Personnel  393  Appendix 2: Information Letter for Patients  396  Appendix 3: Consent to be Observed  398  Appendix 4: Consent to be Interviewed  400  Appendix 5: Letter for Health Care Personnel Not At Study Sites  402  Appendix 6: Initial Interview Questions  404  Appendix 7: Additional Interview Questions  405  Appendix 8: Description of Observations  406  Appendix 9: Data Collection Time Sequence  407  Appendix 10: Number of Participants and Interviews  408  Appendix 11: Letters Inviting Feedback on Draft Analysis  409  Appendix 12: Nurses Experiences of Violence at Home and at Work  411  ix  List of Figures Figure 4-1: Conceptualization of relationships between dominant interests, and individual enactment of local practices applied to the distribution of resources in health care  101  Figure 4-2: Conceptualization of the individual enactment of practice in relation to the predominant pattern of practice in Emergency  109  Figure 4-3: Conceptualization of the neglect of violence and abuse as a specific case of the predominant pattern of practice (efficient processing)  113  Figure 4-4: Conceptualization of patterns of practice in relation to violence against women  116  X  List of Tables Page Table 2-1: Studies Reporting the Percent of Patients Presenting to an Emergency Department Who Have Been Battered  29  ACKNOWLEDGMENTS To have created knowledge together is an honour; to have created friendship is everything. A project of this intensity cannot but leave one transformed. Most of all I value the friendships that developed through this process, and how those who have helped me have also shown me how to be a better friend. Thank you To all the staff nurses (who cannot be named) who worked with me on this project, for opening your practice and lives to my scrutiny, sharing your ideas and insights, your commitment to nursing. For your support, encouragement and trust. For being my teachers, becoming my friends. To all the women who shared their experiences of violence, for reliving your pain and opening my eyes. Special thanks to Angie and Janice who remind me why we do this work. To all the health care and service providers who welcomed me, for sharing your ideas and your practice. Special thanks to Jean Fike, Shirley Stacey and Joyce Rigaux for helping me to become part of the community of those dedicated to ending violence. To the many excellent teachers I have had through these past four years. To Leslie Roman, for opening spaces and possibilities for me to do this work, and giving me many beginning tools. To Joan Bottorff, for the opportunities and support you provided in helping me begin to learn how to function in the world of academic scholarship. And to Sally Thorne, for the opportunities and encouragement, practical support and intellectual challenge. To my superb committee. Katharyn May, for keeping the spaces open, for being able to give exactly the right amount of encouragement and direction. Joan Anderson, for your brilliance, elegance and grace, in thought and deed. Angela Henderson, for keeping me grounded, for helping me find balance. To my friends and colleagues. Debbie Hollands, for always believing in me. Paddy Rodney and Lynne Young, for leading the way and sharing the journey. Janice McCormick for your depth and brilliance, shared laughter and tears. Rose Steele, for always being there. Leighanne McKenzie, Sheila Turris and Lori Irwin, for your enthusiasm, support and commitment when I most needed it. Seonag McCrae, Bev Pister, Helene Jospe, and Karen Mahoney, for hanging in there, despite my inattention, tardiness, and single-mindedness. To my children, Aaron, Alex and Megan, for cups of tea, shoulder massages and back cracks. For your tolerance and cheerfulness, and most of all, for keeping me in the world. Last, but not least, to my husband, Jim Reich. Your love and support makes me strong.  For my mom, Marion Varcoe loving memory of her son, my brother Vance Emmett Varcoe (1956-1974)  CHAPTER ONE - THE PROBLEM Violence against women is a problem of epidemic proportions, a global epidemic that spans all "races", all classes and most countries (Heise, 1994; Heise, Pitanguy & Germain, 1994). The 1  most recent survey of the incidence of violence against women in Canada estimates that one in every two women over the age of 16 has experienced physical or sexual abuse (Rodgers, 1994). Most assaults against women occur within the context of intimate relationships (Dobash & Dobash 1988; Heise, 1994; Koss, 1994; Rodgers, 1994). Although all estimates of violence in intimate relationships "underestimate reality" (British Columbia Institute on Family Violence, 1994, p. 1), in 1993, 29% of Canadian women who had ever been married or lived in a commonlaw relationship reported being physically or sexually assaulted by a marital partner at least once during the relationship (Johnson, 1996; Rodgers, 1994). Dobash and Dobash (1988) found that in Great Britain, 25% of all violent crimes are wife assaults, and conclude that "it is through taking on the position of wife that women are most likely to become the victims of systematic and severe violence" (p. 57). Violence against women incurs enormous health consequences, although it is rarely seen as a public health issue (Heise, 1994; Heise et al., 1994). At a global level, the health burden from rape and domestic violence is comparable to that posed by cancer, human immunodeficiency virus (HIV), and cardiovascular disease (Heise et al., 1994). In the United States, violence against women is the leading cause of female trauma (Campbell & Sheridan, 1989). In Canada, the women who responded to the Violence Against Women Survey (VAWS) said that they had been physically injured in 45% of all cases of wife assault and in 22% of violent sexual assaults (Johnson, 1996). The conditions attributable to gender-based victimization include sexually  1 have put "race" in scare quotes throughout this thesis in order to stress that race is a social construction. I considered this important because the idea of race as I encountered it throughout this study was more likely than other concepts (such as "class" and "culture")' to be treated as "natural facts". 1  2  transmitted diseases (STD), HIV, abortion, depression, alcohol and drug dependence, posttraumatic stress disorder (PTSD) and death from suicide or homicide (Heise et al., 1994). The long term physical consequences of violence and abuse include chronic pain, irritable bowel syndrome, arthritis, neurological damage, chronic pelvic pain, and pelvic inflammatory disease (Campbell, 1993). The psychological consequences of abuse include intrusive symptoms such as nightmares, hyperarousal, and disconnectedness from others (Herman, 1992). The consequences of victimization have been shown to extend long beyond the assault and are reflected in the significantly greater use of health services in comparison to non-victimized people (Koss, 1993). The consequences of violence against women also extend to children exposed to violence, who experience problems such as difficulty in problem solving, adjustment problems, and learning to use violence (Jaffe, Wolfe & Wilson, 1990). Twenty one percent of Canadian women who said they were abused by a marital partner were abused during pregnancy and 40% of these women said that the abuse began during pregnancy (Johnson, 1996; Rodgers, 1994). Despite the size of the problem, and the associated health consequences, violence against women is neglected in health care. Violence and abuse are largely not recognized in health care settings, and when recognized, are not dealt with effectively. The rates of recognition of violence against women by health care professionals are extremely poor (Heise, 1994; Kurz & Stark, 1988; McLeer & Anwar, 1989; Stark, Flitcraft & Frazier, 1979; Warshaw, 1993). Studies of Emergency units suggest that without specific domestic violence screening protocols the recognition is less than 2% of all cases of abuse (Kurz & Stark, 1988; Stark et al., 1979; Warshaw, 1993). Similarly, without specific domestic violence screening protocols abuse is recognized in 5-7% of female trauma cases (McFarlane, Christoffel, Bateman, Miller & Bullock, 1991; McLeer & Anwar, 1989), despite estimates that 20-30%) of all female trauma victims  3  (Stark, et al.; McLeer & Anwar), and 22% of all trauma victims (Goldberg & Tomlanovich, 1984) have been abused. In addition to failing to identify women who are abused, health care professionals tend to blame women for this lack of recognition, and often respond in ways that make women responsible for the abuse and leave them even more isolated (Heise, 1994; Warshaw, 1993; Kurz & Stark, 1988; McLeer & Anwar, 1989; Stark et al., 1979). Ratner (1995) used LISREL® to analyze the VAWS data and concluded that "when abused women come into contact with physicians, nurses, the clergy, or counselors, as a result of the abuse they have experienced, they are likely to receive no gain or incur further losses to their health status" (p. iv). It is especially critical to address the shortcomings of the health care system and nursing in dealing with violence and abuse. The health care system is particularly well placed to allow providers to intervene with women who have been battered because it is the single system with which every woman is likely to interact at some point in her life (Heise, 1994) and nursing's holistic approach, health orientation and clinical concern presage a unique contribution (Campbell, 1992). Because nurses are in contact with women on a regular basis they have the potential to make a significant contribution to the health care of women who are battered and to the prevention of further violence. Racism, classism and negative attitudes toward women by health professionals contribute to the detrimental responses (Campbell, Pliska, Taylor, & Sheridan, 1994; Dobash & Dobash, 1992; Hampton & Newberger, 1988). However, the reasons for the neglect of violence and abuse have not been fully explicated. Analyzing the reasons for the lack of an appropriate and sensitive health care response, Kurz and Stark (1988) conclude that a lack of information and sexism alone are insufficient explanations. They suggest that the response of those within the health care  4  system has been inadequate because, as in the larger society, violence has been conceptualized as an individual problem, and as extensions of the state, health care institutions are not mandated to intervene in "private matters" and health care providers have not developed a mechanism of response. Stark, et al. (1979) further suggest that the political and economic constraints within which the health care system operates are part of an extended patriarchy that is reconstituted by medicine in dealing with women who have been abused. The role of nursing in responding to women who have been battered has not been described separately from the role of other health care professionals, so it is not clear what the differences and similarities are between nurses and other health care professionals in dealing with violence. It is also not clear how nurses influence and interact with others in the health care system and the larger community with regard to violence. Warshaw (1993) has speculated that nurses' conditions of work and relative powerlessness within the health care system may limit their effectiveness with women who have been battered, but the influence of these factors and the fact that nurses are predominantly women has not been studied. In summary, violence against women is known to be an epidemic globally and in Canada, with most assaults against women occurring within the context of intimate relationships. Abuse of women is also known to be a significant health problem with immediate and long term physical and psychosocial sequelae. Although women frequently seek assistance from the formal health care system, it has been shown that professionals fail to identify women who are battered, tend to blame women for this lack of recognition, often respond in ways that make women responsible for the abuse and feel more isolated, and may make health outcomes worse. The lack of an appropriate and sensitive health care response has been suggested to arise from the fact that the health care system, as an extension of the state, is not mandated to  5  intervene in "private matters", and has not developed a mechanism of response to violence. It has been suggested that, as in the larger society, health care professionals conceptualize violence as an individual problem, a view which colludes with the political and economic constraints of an extended patriarchy within which the health care system operates. As nurses are regularly in contact with women, they may be able to contribute significantly to the health of women who are abused and to preventing violence. Nursing has begun to contribute to knowledge regarding violence against women. However, to date the actual practice of nurses in relation to violence has not been studied. Despite growing awareness that violence is a problem of power relations that is deeply gendered, nursing research has not analyzed power relations or oppression in relation to violence. Further, the ways the social context influences and is influenced by nurses' practice in relation to violence has not been studied. The Research Questions The following three research questions were used to guide this study: 1) How does the social context in which health care is provided, and the power relations within that context, shape and constrain nurses' care for women who have been battered? 2) How do nurses affect the structures and relationships within the social context in relation to their practice with women who have been battered? and 3) What are nurses' experiences in providing care for women who have been battered and how do they understand the context in relation to their practice with women who have been battered? Conceptual Issues and Definition of Terms Terminology within the area of violence against women is problematic and highly contested. Despite careful consideration of the conceptual meanings and terminology used in this study, I found that over the course of the study, some of the terms I had drawn upon from the  6  literature became problematic in the field. Within the context of the everyday world of nurses and women who have been abused, I found that the participants ascribed quite different meanings to certain terms than the theoretical meanings I had ascribed. Thus, I have replace several terms used initially so that this work might more accurately reflect the terminology I actually used with participants, and so that it might be more accurately read by participants. Most importantly, the research questions concerned women who have been battered, as I understood battering to be distinguished from abuse on the basis of health risk, and had defined battering as the physical and psychosocial sequelae of abuse (Stark & Flitcraft, 1991). Although this term is used occasionally throughout the study, it has been replaced where possible with the term women who have been abused, for two reasons. First, the women interviewed who had experienced abuse objected to the term "battered" as to them the term connoted physical violence and they insisted that the significance of mental, emotional and economic abuse be recognized. Second, the term also connoted physical violence to nurses in this study, and using the term "battered" colluded with their tendency to focus on physical violence. The other central concepts in the study are the social context of nursing, violence, violence against women and wife abuse. The Social Context of Nursing refers to the entire context of nursing practice, and encompasses the immediate units in which nurses practice (such as the emergency department), health care settings (such as the hospital), the communities in which health care settings are located, the health care system, and society. The larger society also encompasses specific social influences which are particularly relevant to this study, such as the state, the public media, and profession of nursing. Although this study occurred within the context of Canadian society, it is understood that because of globalization, the influences of society are not limited by political boundaries. For example, American media is highly  7  influential in Canada, including both the popular media (with a particularly influential example being the O.J. Simpson trials), and professional literature. As the phrase "social context" was not immediately comprehensible to many participants, I often had to augment this definition by also referring to the "practice environment", the "social environment", and the "community". Violence is defined as an abuse of power (BC Institute on Family Violence, 1994). "Violence" encompasses a vast array of abuses in a wide range of contexts. Violence can be thought of as including all forms .of abuse: emotional abuse, psychological abuse, financial exploitation, physical assault and sexual assault. Violence can include stranger violence, violence in intimate relationships, and more broadly, social violence. I am using the term "violence" to encompass all forms of violence in all contexts. However, in this study I was especially, but not exclusively concerned with the unique problems of violence within intimate relationships, and more specifically the unique problems inherent in the role of wife. While I brought this focus to the study, participants often saw other ideas as central, and when asked about "violence" tended to talk more about violence against children, "stranger" violence or sexual assault. Again, despite my meanings, participants brought their own meanings to this study. Despite defining violence in the above manner, it seemed that participants often understood violence as meaning physical abuse. I attempted to clarify my meanings when possible, but often used the term "abuse" instead of violence, as "abuse" seemed to more clearly communicate the abuse of power to which I referred, and thus in this study the terms "violence" and "abuse" are used together and interchangeably. Violence Against Women clearly refers to violence that is directed toward females, and is a subset of gender-based violence directed toward females throughout the life cycle. Women are  8  commonly the victims of violence, but violence is by no means confined to women. Violence against women cannot be separated from violence against children, because the children of women who are battered witness violence and may also be targets of violence. Mothers of children who are abused may also be violated by the abuse of their children. Men are clearly victims of stranger violence and social violence, and to some extent are victims of violence in at least gay intimate relationships. I am choosing to focus on women with recognition that violence affects all members of our society, with the least powerful being the most violated. As noted earlier, the term battering was initially used and was defined as the physical and psychosocial sequelae of abuse. However, women who have been abused is the phrase that is used throughout this study to refer to women who have experienced the physical and psychological sequelae of violence, which is an abuse of power. This phrase was used in an attempt to avoid using labels such as "battered woman" and "abused woman" which define women only as victims. As noted, the term "abused" was used in preference to the term "battered" as the latter term implied physical violence to the participants. The shift in terminology occurred only as I became aware that despite how I had defined the term "battered", connotations that were made by participants were quite different and were shaping the study. Thus, the research questions, consent forms and so on, use the term "battered", and the term is often used by participants, at least partly in consequence of my use of the term, but in the analysis and discussion I have used the term "abused" whenever possible. This is critical because the most important audience for this work are the participants of the study; thus their meanings are the most significant. Wife Abuse refers to the use of violence against a woman by a male intimate. Following feminist scholars such as Dobash and Dobash (1988), Kurz (1993), and Bograd (1988), this term  9  is used explicitly to draw attention to the vulnerability that accompanies the role of wife, but is not intended to imply that women only experience abuse within the context of a legal marriage. While violence against women pervades all facets of western society, abuse of a women by a male partner within the context of an intimate relationship is one of the most insidious and challenging forms of violence. Use of this term is not, however, intended to mask the fact that women experience violence in many contexts in addition to intimate relationships with men. The State in this study refers to all institutions which serve to order and organize society and define what is appropriate behavior in our relations with each other (Barnsley, 1985). Given this definition, health care institutions are considered part of the state. Summary The reasons for the less than effective response of the health care system to the problem of violence against women have not been fully explored, although suggested explanations have focused attention on the social context. Further, the role of nursing in relation to violence against women has not been examined. Thus, the basis for improving nursing's contribution to care in relation to violence against women in unknown. In light of this problem, this study was undertaken to examine the relationships between the social context and nurses in relation to violence against women. In the next chapter a more comprehensive review of the literature is presented concerning the scope of the problem of violence against women, with particular attention to wife abuse, the health consequence of violence against women, and the health care response to violence against women, with emphasis on nursing and Emergency settings.  10  CHAPTER TWO - THE LITERATURE In the three decades in which violence against women has been openly discussed and researched as a social and health problem, an understanding of the scope of the problem and the nature of abuse has begun to develop. However, many controversies remain, and the social response to violence against women has not had an appreciable effect on the problem. In North America, Great Britain, and around the world, organized efforts to protect women from violent men and to assist women to protect themselves have come largely from women themselves (Dobash & Dobash, 1988; Heise, 1994). Although women frequently seek assistance from heath care institutions and practitioners, professionals within the formal health care system have been largely unresponsive, and perhaps detrimental, to the welfare of women who have been violated (Goldberg & Tomlanovich, 1984, Lempert, 1997; McLeer & Anwar, 1989; Rodgers, 1994; Ratner, 1995; Stark, Flitcraft & Frazier, 1979; Warshaw, 1993). In this chapter, the literature is reviewed to provide an overview of what is known about violence against women, the relationship between nursing and violence against women, and the social context of this relationship. First, in order to establish the importance of the problem of violence against women, the scope of the problem will be reviewed. Particular attention will be paid to wife abuse to emphasize the prevalence of this particular form of violence. Second, the health consequences of violence against women will be reviewed in order to establish that violence is a social problem of grave concern to nurses and other health providers. Third, the response of society to violence against women will be described in order to provide background for discussion of the response of those within the health care system. The reasons for the inadequate social response will be discussed with specific attention to the ways in which violence has been conceptualized. The response to violence against women by those within the health  11  care system is seen as a specific case of the more general social response and, in the fourth section of this review, will be detailed in order to identify the gaps in knowledge and deficiencies in practice. This section will focus on the lack of recognition of abuse and the detrimental responses to abuse within the health care system. In the fifth and final section of this review, particular attention will be paid to the response of nurses to violence against women in order to identify the gaps in knowledge and deficiencies in practice specific to nursing. In the description of the response of nurses and other health care providers, specific attention will be paid to Emergency settings, primarily because Emergency has been identified as one of the primary points of contact between nurses, other health care providers, and women who have been abused. The Scope of the Problem of Violence Against Women Estimates of the prevalence of violence against women are hampered by under-reporting, lack of data, definitional problems, and methodological issues. Despite these challenges, violence against women, and more broadly, gender-based violence, is acknowledged to be a global epidemic. Violence against women occurs in the work place, the home and in society at large, and encompasses rape, sexual harassment, forced prostitution, physical assault, psychological abuse and wife abuse (Hiese, 1994; Hiese, Pitanguy & Germain, 1994). Of all forms of violence against women, wife abuse, or abuse by a male intimate is the most common. Reviewing 35 well designed studies from around the world, Hiese et al. (1994) concluded that despite differences in the studies, in many countries one quarter to one half of women report having been physically abused by a partner or former partner. An even greater proportion of women reported being subjected to ongoing psychological and emotional abuse. In Canada, Statistics Canada conducted a national population survey in 1993, the Violence Against Women Survey (VAWS). In this survey, researchers interviewed a randomly selected  12  national sample of 12,300 women by telephone. The data from this study have been analyzed by various researchers, including Rodgers (1994), Ratner (1995), Johnson (1996), and Kerr and McLean (1996), providing the most comprehensive picture of the problem in Canada to date. The VAWS reported that half of Canadian women over the age of 16 had experienced at least one incident of sexual or physical assault, and that 10% had been the victims of assault in the year preceding the survey (Johnson, 1996; Rodgers, 1994). In congruence with global statistics on wife abuse, 29% of women who had ever been married or lived in a common-law relationship reported being physically or sexually assaulted by a marital partner at least once during the relationship (Johnson, 1996; Rodgers, 1994). Johnson extrapolated these figures to the population, estimating that over 2.6 million Canadian women have experienced physical or sexual assault, and that of the 6.69 million women currently in a marital relationship, 1.02 million (15%) have been assaulted. In particular, British Columbia has the highest reported incidence of violence against women in Canada (Kerr & McLean, 1996). In addition to this appalling level of prevalence, the VAWS (Johnson, 1996; Rodgers, 1994) also provided estimates of the frequency and severity of violence in the context of intimate relationships. In 63% of all cases of wife assault violence occurred more than once, and 32 % of all cases of wife assault involved more than 10 episodes of violence. In almost half of all relationships with violence, a weapon was used at some point, and almost half of those assaults resulted in injury to the woman. In 43% of the situations in which the woman was injured, the woman sought medical attention. In 34% of all cases of assault the woman feared for her life. The prevalence, frequency and severity of wife abuse in Canada is similar to that in other countries around the globe. Thus it can be anticipated that the health burden attributable to wife abuse is similar to that of other developed countries.  13  The Health Consequences of Violence Against Women The impact of violence on health is not fully understood, and as analyzed by Ratner (1995), this understanding has been limited by the lack of studies designed to yield generalizable results, and the lack of longitudinal studies to examine associations between exposure to abuse and health outcomes. The impact has been suggested to include general effects on health, immediate consequences, and long range consequences. One of the estimates of the general impact of violence on the health of women was made by the World Bank through a modeling exercise in which estimates of the healthy years of life lost due to various conditions were made (Hiese, Pitanguy & Germain, 1994). Counting each year lost due to premature death, and each year spent sick or incapacitated as a fraction of a year, it was estimated that rape and domestic violence accounted for nearly 20% of the health burden for women ages 15-44 in developed countries (the percent of the total health burden was lower in developing countries due to the greater overall burden of disease). Globally, the proportion of health burden estimated to be due to rape and domestic violence was comparable to the burden estimated as incurred by cancer, cardiovascular disease, HIV and tuberculosis, which were only outweighed by maternal conditions and STD's. Women's estimates of their general health have been shown to be proportionally poorer in relation to the violence they have experienced (Gelles & Straus, 1988). Based on a national sample in the United States (US), Gelles and Straus found that the greater the violence a woman experienced, the less likely she was to report excellent health, and the more likely she was to report fair or poor health. They found that women who had been abused stayed in bed due to illness twice as often as women who had not experienced abuse. However, analyzing more recent national Canadian data, Ratner (1995) concluded that exposure to wife abuse accounted  14  for very little of the variance in women's perceptions of their health status, perhaps because of the way in which the general public defines health. The specific consequences of violence against women include both the immediate effects of abuse and the long term consequences of having previously been abused or currently living under the chronic stress of violence. The consequences have been described as including physical injury, psychological consequences, alcoholism and drug use. Physical injury is the most fully explored outcome of violence. Physical injury was initially used to define abuse, to the point of overshadowing non-physical consequences of abuse (Gelles & Straus, 1988). Patterns of injury have been used in various studies to identify the incidence of abuse (e.g. Kurz & Stark, 1988) and physical injuries continue to be used to estimate the incidence and severity of abuse, with national crime surveys in both the US and Canada typically reporting numbers and types of injuries (see Campbell, Harris & Lee, 1995; Johnson, 1996). Recently attention has turned to specific injuries that tend to be unrecognized as abuse, such as head, neck and facial injuries (Ochs, Neuenschwander & Dodson, 1996), ocular injuries (e.g. Beck, Freitag & Singer, 1996) and orthopedic injuries (e.g. Varvarao & Lasko, 1993). In the VAWS, more than 40% of women who had been abused by their partner said "yes" to each of having been beaten up, kicked, hit or bit, slapped, pushed, grabbed or shoved, any of which could cause physical injury. In addition, 24% reported being hit with something, 30% reported being choked, and 35% reported being sexually assaulted by their partner. As noted earlier, women were physically injured in 45% of all cases of wife assault and in 22% of violent sexual assaults, and sought medical assistance in 20% of cases of wife assault. In an average year in Canada, 78 women are killed by their partners, and in 1993, about 200,000 women were  15  threatened, slapped, kicked, punched, choked, beaten, or sexually assaulted (Johnson, 1996). Not surprisingly, the severity of injuries varies with the severity of the abuse (Ratner, 1995). The psychological sequelae of abuse have been variously described, and include a range of emotional effects, intrusive symptoms encompassed by Post Traumatic Stress Disorder (PTSD), suicidal ideation and suicide. The emotional consequences of wife abuse, not surprisingly, include anger, loss of trust, fear, and lowered self esteem (Gelles & Straus, 1988; Johnson, 1996; for a detailed discussion of the dynamics of these consequences, see Ratner, 1995). Among women who seek help in relation to abuse, PTSD, which includes nightmares, intrusive memories and hyperarousal, is common (Herman, 1992; Saunders, 1994). Saunders found that in previous research the mean rate of PTSD among women who had been abused was 56%, and in his study of 192 women the prevalence was over 60%. Gelles and Straus (1988) found that women who have been abused are much more likely than non-abused women to contemplate and attempt suicide, a finding subsequently supported by other studies (e.g. Abbott, Johnson, KozoilMcLain & Lowenstein, 1995). In general, the psychological consequences of abuse explored in the studies cited above tend to be worse with the severity and frequency of abuse. Alcohol and the use of prescription drugs have been proposed as coping mechanisms for dealing with abuse (Ratner, 1993, 1995; Stark et al., 1979) and in the VAWS, one quarter of women turned to alcohol, prescription or other drugs to help deal with their situations (Rodgers, 1994). The use of prescription drugs to sleep and relieve anxiety has been suggested as much higher among women who have been battered than those who have not (Groenveld & Shain, 1989). Interestingly, Ratner's (1995) analysis of the VAWS revealed that the strongest determinant of whether a woman took drugs or medications to deal with the abuse was contact  16  with physicians or nurses, a finding congruent with earlier criticisms of health care responses (e.g. Stark et al., 1979; Warshaw, 1993). The impact of wife abuse extends beyond women to their children. Children who witness violence are traumatized by the violence indirectly, and are likely to also be abused. Gelles and Straus (1988) found that children in homes where there was violence were much more likely to have a variety of problems, including having troubles in school, being aggressive and using drugs and/or alcohol. In Bennett's (1991) phenomenological study of adolescents who witnessed abuse, the girls were traumatized by witnessing violence and were also likely to be direct victims of abuse. In their study of 1000 women who had been battered, Bowker, Arbitell and McFerron (1988) found that in 70% of the families where children were present, the men also abused the children. Berman (1996) found that children from families in which there was violence were similar to children of war in terms of the PTSD that they experienced. Violence against women causes tremendous suffering, and social, economic and health costs, which can only be partially quantified in economic terms. The costs of the health consequences of violence against women can only be estimated, and are assessed to be staggering (Day, 1996; Hankivsky & Graves, 1995; Kerr & McLean, 1996). What then has been the response to this epidemic? The Social Response to Violence Against Women The health care response to violence against women is a subset of the larger social response to the problem. The social response to the epidemic of violence against women, especially within the context of intimate relationships, has been inadequate if not destructive. As current studies of the prevalence of abuse illustrate, violence against women continues unabated. Until recently, abuse in intimate relationships has been treated as a private matter, and there has been  17  limited intervention by the state. In the past decade, however, such abuse has been "criminalized" in many countries, with the police and justice systems directed by policy to treat violence against intimate partners as a crime. Such strategies are not intended as primary prevention strategies, and have had limited success in preventing further abuse. The limited social response has been influenced by the way in which violence is conceptualized, by the role of the state in intervening against violence and the division between private and public realms. How Violence is Conceptualized  1  Society's response to violence has been at least partially determined by the way in which violence against women has been conceptualized. At present there appear to be two central problems in conceptualizing violence. First, three distinct views have been used and given rise to contradictory explanations of violence and directions for practice. Second, within those three views, power and oppression have been under-theorized. Under-theorizing power and oppression has limited each of the perspectives and, I believe, created barriers to integration. Theorists from across disciplines have attempted to understand violence by focusing attention on and seeking causal explanation for violence within individuals, within couples or dyads, and within society. Thus three models have developed to explain abuse: the interpersonal model, the family violence model, and the gender-politics model (Stark and Flitcraft, 1991). Bograd (1988) and Gelles and Loseke (1993) label these three predominant "lenses" as the psychological lens, the sociological lens and the feminist lens, referring not to specific disciplines, but rather to the focus of inquiry and presumed causality of violence. Initial attempts to make sense of violence against women tried to explain violence by focusing on the individual. These views emphasized the psychology of the victim, and more  Some of the discussion appearing here has been published along with the implications for nursing research (Varcoe, 1996). 1  18  recently, the psychology of the perpetrator. The initial focus on the characteristics of victims led to victim blaming theories of violence such as the theory of learned helplessness which "merely labels as a peculiarity...what is in fact a reasonable response to an unreasonable situation", thus diverting attention from the situation to the victim (Wardell, Gillespie & Leffler, 1983, p. 76). More recent attention to the psychology of the perpetrator has shifted the locus of causes of violence to the psycho-pathology of the perpetrator, but leaves power and gender relations unexamined. The focus on the individual has been popularized by the media which, Dobash and Dobash (1992) argue, perpetuates "unsubstantiated yet damaging theories about the problem, its victims, perpetrators and solutions [notions that] implicitly assume that this is strictly an individual problem suffered by deviants needing psychiatric care rather than a social problem in need of wider remedies" (p. 32). Bograd (1988) criticizes the focus on the psychology of individuals, because it suggests that violence is an aberrancy of a few husbands (rather than the normal patterns of most men), excuses men, implicates women and concludes that the differences between abused and non-abused women are the causes rather than the consequences of abuse. Causal explanations of violence related to the psychology of the individual leave power and gender relations unexamined and consider violence in isolation from the social and historical contexts in which it occurs. The second set of perspectives on violence focuses on dyads or families, and seeks explanations of the causes of violence in social relations within couples and families. These perspectives, which are used in most research on violence (Silva, 1994) tend to be genderneutral, to treat power inequities as only one factor among many, and to explain violence as resulting from external stresses and breakdown of the family, rather than as a part of most normally functioning families (Bograd, 1988; Stanko, 1988). Straus and Gelles have done  19  perhaps the most influential work on relations within dyads. Straus and Gelles (1986) conceptualized violence as a conflict between parties and, using the Conflict Tactics Scale (CTS), found equivalent violent behavior among men and women. Y116 (1993), Dobash and Dobash (1988), Silva (1994) and others have critiqued the perspective underlying the CTS because it does not critique power or gender relations. As with the focus on individuals, focus on dyads or families limits analysis of the influence of the social context. The third set of perspectives, labeled "feminist", tend to explain violence as arising from the social context and contribute an analysis of the influence of gender and power to theorizing violence (Yllo, 1993). Gelles (1993) and others (e.g. Dutton, 1994; Letellier, 1994) argue that feminism is limited to using a single variable (patriarchy) to explain the existence of wife abuse, and use evidence of men who are not violent and violence in same sex relationships to argue that patriarchal ideology does not account for male violence. However, this is a narrow view of feminism and, as countered by Yllo, reflects a narrow conception of patriarchy. As Renzetti (1994) and others note, many feminists are not exclusively concerned with gender. Tensions and conflicts between these various perspectives have led to very different explanations of violence, and therefore to very different approaches to decreasing violence. The battle between these perspectives continues to be waged, and violence theorists are calling for integrated models (e.g. Dutton, 1994; Miller, 1994, Tolman & Bennett, 1990; Renzetti, 1994). However, the approaches to integration are contentious, and I believe are seriously constrained by the limited ways in which power and oppression have been theorized in relation to violence. Power. Oppression. Resistance, and Perspectives on Violence Analyses of power are generally absent in work from perspectives that focus on the psychology of individuals, leaving an implicit assumption of equality between individuals.  20  Within perspectives that examine violence within dyads or families, power is viewed as one of many factors influencing violence and the view of power is one in which there are two equally opposing forces. From feminist perspectives, power inequalities are assumed and occupy a central position in the way that violence is theorized. However, power may not be explicitly theorized or may be theorized in a variety of ways within each of these perspectives. Three conceptualizations of power initially proposed by Lukes (1974), and applied and refined by Gaventa (1980) highlight the limitations of the various ways in which violence has been conceptualized. The first dimension is a traditional view of power as a contest between two opposing forces in which the "winner" usually has the greatest resources (intellectual, material, personal, experiential). This view is based on the assumption that all individuals and groups have equal opportunity to express dissent. Viewing power in this manner assumes that nonparticipation is the fault of the non-participant and a consequence of apathy or a lack of experience or skill. Powerlessness is explained as a lack of knowledge, communication skill, political expertise or clout (Dykema, 1985), in other words, a deficit on the part of those who are least powerful. This conceptualization of power is congruent with, and supports views that focus on the individual in which the differences between abused and non-abused women can be viewed as the causes of violence. This view of power is also compatible with dyadic perspectives, as exemplified by the work of Straus and Gelles (Gelles & Straus, 1988; Straus & Gelles, 1986), in which violence is viewed as a conflict between two equal and opposing parties. Using this view of power, issues of oppression and resistance do not arise as the parties are equal, women are assumed to be autonomous agents, and questions such as "why does she stay?" are justifiable. The second view of power suggests that some people are excluded from contesting their positions. Dominant beliefs, attitudes, values, institutional rituals and practices operate to benefit  21  certain people or groups. Those who benefit are supported in defense of their position; those who do not benefit are simply excluded'from decision making or suffocated before being heard. This view of power is implicit in most feminist conceptions of violence, which regard the power inequalities which are fundamental to wife abuse as deeply gendered, arising from multiple sources of oppression, and fostered by the state (see for example, L. Hoff, 1992; Kjervik, 1992; MacKinnon, 1993). Devaluing and oppression of women in society are seen as fundamental to violence against women. Because women are excluded from contesting their positions, resistance is limited and women's agency is constrained. From this view of power, the question becomes "what keeps her here?" and is directed toward the structures of society and the state. In the third dimension, power relationships are maintained because the very wants and needs of the dominated are shaped by more powerful others. Gaventa (1980) thinks that this phenomenon occurs 1) as a psychological adaptation to powerlessness, 2) from a lack of opportunity to develop political consciousness, and 3) from inconsistent belief patterns among the dominated. This view of power also underlies many feminist understandings of violence, leading to much more complex analyses of the experiences of women who are abused (e.g. L. Hoff, 1990; Wuest & Merritt-Gray, 1994). Unlike views of power which give rise to theories such as learned helplessness, this view of power explains women's behaviors as adaptations to powerlessness and domination rather than as psychological deficiencies, and sees domination arising not only from the person inflicting abuse, but from an entire system that tolerates, accepts, and perpetuates abuse. It follows that feminists argue that research on violence must be more concerned with oppression than victimization (e.g. Kjervik, 1992; McBride 1992; Yllo, 1993). However, at the basis of criticisms of the narrow concern with gender offered by  22  feminism (e.g. Dutton, 1994; Gelles, 1993; Letellier, 1994), is a very real problem with the ways in which oppression has been theorized by some feminists. Feminist theorizing shifted the discourse on oppression from class as the central source of oppression (which arose from Marxism), to a concern with gender as the central source (e.g. Acker, Barry & Esseveld, 1983; Eistenstein, 1977; Harding, 1987). This shift, born of white middle class western feminism, was important in drawing attention to gender-based oppression. However, placing gender at the center of feminist theorizing erroneously implies that gender is the central defining feature of a woman and "reflect(s) the dominant tendency in western patriarchal minds to mystify a woman's reality by insisting that gender is the sole determinant of woman's fate." (hooks, 1984, p. 14). This focus on gender rests on essentialized notions of "woman" and "patriarchy" (Walby, 1992); that is to say, woman is conceived of as a biologically or socially defined "essence", and patriarchy as a monolithic entity (Alcoff, 1988; Collins, 1989). In treating gender as a category distinct from "race", and class, these too are essentialized. Feminists such as Brewer (1993), Collins (1986), Mohanty (1992), Ng (1993) and Smith (1990) contest the centrality of gender oppression, essentialist conceptions of gender, and the subordination of experiences of "race" and class. Following these critiques, feminists have theorized oppression as arising from multiple sites, most expressly including "race", class and gender. Critiques of uni-causal models of oppression have stimulated pleas for radical pluralism. However, accounting for endless sources of oppression presents crucial challenges to meaningful analysis (Bordo, 1994; Phillips, 1992). At the same time post-structuralists have questioned the utility of analytic categories such as "race", class and gender, declaring such categories to be too internally diverse to be useful (Walby, 1992). These two different challenges have destabilized feminist theory and threatened the very categories by which oppression can be understood and  23  contested (Bordo; J. Hoff, 1994; Phillips; Smith). Alternatively, Brewer proposes a focus on "the simultaneity of oppression" (p 16) which arises from multiple interacting sites, without abandoning the analytic categories of oppression such as "race", class and gender. The way oppression is conceived is critical to theories about violence used to guide research and practice. If gender is the sole source of oppression, then wife-abuse is seen to arise from relations between men and women, and theories locating the causes of violence within the individual and family are sufficient. However, viewing oppression as simultaneity demands a view of violence as arising from the social context. This position is congruent with the feminist focus on the context of violence (Bograd, 1988; Dobash & Dobash, 1988; L. Hoff, 1992; Yllo, 1993). Further, this view of violence as arising from multiple sites of oppression permits and requires analyses of racism, classism, heterosexism, ageism and other experiences of oppression. Wife abuse no longer can be seen as a woman's (or women's) problem, but rather becomes a problem of social dimensions requiring intervention not only with individuals who experience and perpetrate violence, but with other social relations that permit and sustain violence. While the social response to violence against women has been limited by the way violence has been conceptualized, it may be that these theories of violence are functions of social forces that produce and maintain violence, and serve the state by reproducing and maintaining violence. The Role of the State  2  Dobash and Dobash (1992) contend that there are two arguments against state intervention with wife abuse: the private sphere must be maintained (which means that whoever is most powerful is not interfered with) and, the state will support the traditional. Heise (1994) notes that  Although I value the importance of these arguments, using Foucault's view of power (as being enacted in all relationships "bottom up", (see Methods chapter) the role of "state" did not become a focus in this study. Rather, individuals were conceptualized as creating the organization (see McCormick, 1997) and sustaining "the state" through their enactments of power. 2  24  it has taken years of struggle by feminist activists for violence to be regarded as a socio-political issue rather than as a private matter or as a problem of individual psychopathology. Dobash and Dobash ask if the battered women's movement is "seeking a form of intervention in the private to which the state has traditionally had no commitment, and thus for which the state has developed no mechanism for response?" (p. 103). Conceiving violence as a product of multiple sources of oppression, including gender, hooks (1984) argues that violence in the family is preferred by the state rather than violence against the state, thus explaining why the state has not acted significantly to end violence against women in their homes. MacKinnon (1993) concurs, but goes further, arguing that the state deals with violence in ways which support and consolidate male power. MacKinnon asserts that the state is male in the feminist sense. She argues that male is the implicit reference for human and that objectivity is the norm of the state, reflecting a view of society as it exists and calling that view practical rationality. "If rationality is measured by point-of-viewlessness, what counts as reason will be that which corresponds to the way things are, and practical will mean that which can be done with out changing anything" (p. 208). She argues that violence is dealt with by the state, partially through law, not just from a male perspective, but in order to institutionalize male power. As the lack of state intervention is predicated on the protection of the private realm, the public/private dichotomy is an important construct to understanding the social response to abuse. Public versus Private Violence The notion that the home is a private domain has limited the social response to violence against women, and thus sustains such violence. Historically, the ideas of public and private have been essential for understanding gender (Schneider, 1994). The dichotomy between the  25  private sphere of women and the public sphere of men has been basic to the distinction between men and women, and these ideas have functioned to limit the social response to wife abuse. As Johnson (1996) points out, the mythology of the privacy and sanctity of the home and family has "allowed husbands the necessary privacy to beat their wives without the fear of legal interference or other types of social sanctions" (p. xxi). Marcus (1994) argues that the use of modifiers such as "date" and "marital" used with rape, and "domestic" used with violence serve to locate the act or activity in the private sphere. These designations minimize the importance of such violence. And she argues, the attribution of "domestic" serves to remove any issue, including violence, from the civil rights agenda. Marcus proposes that in order to disrupt this silencing, violence against women ought to be reframed as "terrorism in the home". The idea that wife abuse is a private matter limits reporting of abuse by women, and limits the social response when abuse is reported, which in turn limits the usefulness of reporting. Stanko (1988) postulated that criminal events are not reported to the police because of how private individuals think their problem is, how seriously they think it will be treated, and their feelings that nothing can be done. The idea that nothing can or will be done in relation to wife abuse is well founded and inaction is also supported by the public/private schism. Schneider (1994) argues that within the criminal justice system the public/private dichotomy has been selectively applied to violence against women "in order to protect male domination" (p. 39). Arguing that the social failure to intervene with wife abuse on the grounds of privacy is not separate from, but actually part of violence, Schneider proposes that the public/private dichotomy ought to be broken down, and the notion of privacy reconstructed in ways that make wife abuse a public problem requiring collective action.  26  The social response to violence against women in general, and wife abuse specifically has been limited, and seems largely ineffectual. The way in which violence has been theorized, the limited role of the state, and ideas that wife abuse is a private matter have contributed to the nature of the social response. The health care response has occurred with the larger social context, and is subject to the same influences. The Health Care Response to Violence Against Women Regrettably, despite the fact that women come to the formal health care system for assistance, the health care response has also been limited and largely ineffectual. The response to wife abuse within the health care system has been characterized by non-recognition of abuse, and by interventions which are often less than helpful. Women who are abused frequently seek assistance from the formal health care system. Studies of hospital emergency units have found that a high percentage of women who come to an emergency unit of a hospital do so because of injuries inflicted by their abuser or because of health care problems that result from living under the chronic stress of violence (Goldberg & Tomlanovich, 1984, McLeer & Anwar, 1989; Stark et al., 1979; Warshaw, 1993). The VAWS found that about one-fifth of women who are assaulted seek, medical attention because of the severity of injury (Rodgers, 1994), and Ratner (1995) found that women seek help from health care professionals primarily because of physical injuries. In addition to actively seeking help from the health care system, women who are abused are in contact with the health care system for problems related to the consequences of abuse and other health issues. While estimates of prevalence vary, there is consensus that many pregnant women are abused (Campbell, Oliver & Bullock, 1993; Campbell, Poland, Waller & Auger, 1992; McFarlane, 1993), with up to 23% of obstetrical patients estimated to be abused  27  (Warshaw, 1993). Recent Canadian statistics show that 21% of women abused by a marital partner were abused during pregnancy and 40% of these women said that the abuse began during pregnancy (Rodgers, 1994). Herman (1992) cites a variety of studies and concludes that 50-60% of psychiatric in-patients, and 40-60% of psychiatric out-patients report childhood histories of physical or sexual abuse or both. Violence against women has a significant impact on the health of women, and the health care system is one source of assistance to which women turn. The response, however, has been at best woefully inadequate, and at worst, a contributing factor to the perpetuation of violence. Health care professionals do not identify abuse despite being presented with obvious indicators, and when abuse is identified, the response is often detrimental. Not surprisingly, health care institutions and professionals are not usually counted as among the helpful responders by women who are battered or by service providers in the community. It is important to note that most research on health care responses to violence against women does not distinguish between the practice of nurses and the practice of other health care professionals. Despite increasing attention in nursing to violence against women, the practice of nurses has not yet been described. Therefore, the health care response will be discussed in general and then some possible similarities and differences in nursing practice will be discussed. Rates of Identification of Women Who Have Been Battered The rates of recognition of violence against women by health care professionals have been consistently found to be extremely low. Most of the research on rates of identification has been done in Emergency units, although as noted, there has been considerable research in obstetrical areas, and some in other areas such as primary care (e.g. Bullock, McFarlane, Bateman & Miller, 1988; McFarlane et al, 1991; McCauley et al., 1995).  28  Although it is commonly reported that 22-35% of women coming to Emergency have been abused, this summary is somewhat of an oversimplification. Comparison of the various studies is hampered by differences in samples and data collection methods. Table 2-1 summarizes a number of studies that have examined prevalence and recognition rates in the Emergency. All studies are from the United States, with the exception of Grunfeld, Ritmiller, MacKay, Cowan and Hotch (1994) from Canada, and Roberts, O'Toole, Lawrence, & Raphael (1993) and Roberts, O'Toole, Raphael, Lawrence & Ashby (1996) from Australia. As can be seen, five of these studies were conducted on samples of adult female trauma patients; three on all adult patients, two on all trauma patients, and one each on all women with vaginal bleeding and all women positively identified as physically abused. One of the critical differences between the studies that makes comparison difficult is that some identify women through research protocols, some through clinical protocols, some through clinical identification without protocols, and some through a combination of these. Another critical difference is that some studies include any event during a life time as abuse, whereas others only include current or recent abuse events. One of the earliest studies to identify the discrepancy between the number of women presenting to Emergency who are abused and the number of women identified as abused was by Stark et al., (1979) who estimated that the actual prevalence of abuse among female trauma patients was about 10 times higher than was recognized by physicians. Of the 481 charts of female trauma patients reviewed, they found that 9.6% were "positive" for abuse (injury stated as inflicted by male intimate or family member), and 15.2% were "probable" (person was kicked, beaten, hit, etc., by an unidentified but not unknown assailant) or "suggestive" (at least one injury inadequately explained by the medical history).  T3 cu T3 «  cd E  S <->  3  +2-»  ca ^  u  ca E cu <+o  E CD  2  +H  o  o  00  00 CN  T3  X  «  cn  ca cu  c  ^  00  E  c  ^ E »  cu  C  o  e  c*-<  o  ca  a.  C  CD  VO  co  a c  >  cu > ce  cn O  e  c (50 3  C  £ £  cu  CD  CN ^  ON  ON  CN CN  •a is E 5 u •' C u « « = E  O  CD  E  s .s 5 «  v  c  B « CCNO > £ B « c  o> cu  cn o >- o cx 3 £ O o ca O cu c y O N j3 r- ' os<u fa ce ^ o, e S ee > o 8 — ON 4) E o g CN a. Q.  o x:  o o .9 \°  o .o  CQ  c  .g O  ON  I  -2  >  ce  ^  c <u E o  O, CD  ^  ce  ^ g  CU  L-  >  0) T3 O  = .2  o cu  n. > ^  ?t  o CU  c  CD  cn  y  >•  ^  -4-»  •3 £  C  00  CD  c .2  ce  u o  CD  ce  in  — ce  ^  ce >£}  c .2 o 00 c2 Co c Xj  OH  3  C  cn 3  3 ~"  c  T  •3 £  E  3 ce  3  S-i  "ce  cEH  ce , U i-  c O N .2 uo ce CO a.  CD >  cu  CD CD-  o  3  00  s 5o O CD  IT)  g ce  o  ce CD  X.  CD  oo  CU  c 5  •g »  ce  T3  S E  iCD  C  c  x  CD  cn CD  0 _  CN  00  ce >  c  1  VO  c '5b  O 00  -3 Xi  in  CN  H C  o 3  cu CU  -o o. § » E ~  « 2 c uo  c  CD CD  ;s;  cn  f 2 £ uo  CO"  II1 1  CO  CN  CD  3  ce  ce  _«J  cu > 3 cn C  «  CD  ce  00  o  CN  -a a. CD  ^  Xi  D,  VO  ci  13 CD  CO  oo  c  'C  cn  O  „  u ce  cn  cS  CD — CD  e g o ? o ^  E '* 2 o JL °J SS 1L5  E  ^—» cn 3 X  00  !<§ •a •«  CO  CD  a-  ce  uo  7j  I/O  2n  ^1  2^. E -I _ VO o c  •o  ce la  T3  ce  ^  1  ce  CD  cu 3 <  CD CU  >-  T3 cu  ce C>3  y! u.  >  •a x: H  CD  ON  Q  ON  Q,  oo ON  oo ON  3  u  ce  £  c  o o  c CJ s-  >  Lav  C/3  ce  CD  *3  ce  ce  CD  CN  cn  ce  <U  C  CD  ON  XJ cu  >  cu io  cu  o  00  XI  O  CD  c ccu  2 >  c c  o  cn  CD CO  r  T3  ce  I** ^ *  ce  •c ce x:  IS  o  ce  CD  CD  i-  o cn  s 3  ce  00  CD LH  ce  a, ce c ce a w E ce x: 3  c  u.  E c o cC c^ ^  o  > cu  o x; o  c 2 ce a.  X II  cd  c«  ce  CD  _  .2  CD  CX c«  cu cu cu V3 t : co ce xi a, o  c  .2 '>  I  Q. ^4-  o o  cu 00 ce  art  Oil  o e c Ec S^ .2 'C 3  S O  CD  o  ec  CD  XJ  ido  ce o  o c  c ce  s  u  cu  ce  C  3  CN  <u cu  '-5 o  x  ^ ro  N^  c  CJl  o •—  >  CN  CQ  Oil _C  ? ce _ o = 3 ca -CJ a  cu  CU  E  Q.  * E o <u o a o "s ^ .S oo ce  £  CN  s-  •a B ^ ti c ca  CD  ca  X cn  -. ce  C/3  00  ON  ON  ON  oo  00  CO ON ON  CO ON ON  CD  C  2  c  3 --  o ON ON  ce  I/O ON ON  ce  -g ^  X)  in  ON ON  o s  cu O .  x>  o  2  t  o  u.  o Cci O  VO ON ON  vo ON ON  9-1 ce  QL  30  In the largest study to date Kurz and Stark (1988) also examined the discrepancy between clinical recognition and the prevalence of abuse determined through chart review. Using 3,676 randomly selected charts of women who presented to emergency with trauma, they identified 642 (19% of all female trauma patients) as abused because of evidence that was either "positive" or "probable". They also found that 40% of all injuries presented in abusive relationships (many charts contained multiple trauma episodes). Medical personnel identified only 15% of the "positive" group as "abuse" or "battering", and associated only 4.5% of abusive injuries with violence, and only 1.8% of the total 3,717 trauma episodes presented by abused women with "abuse" or "battering". Similarly, Warshaw (1993) found that of 52 "positive" cases of physical abuse, only one case was dealt with directly as abuse. Only two replicated studies were located. First, McLeer, Anwar, Herman and Maquiling (1989) replicated McLeer and Anwar's (1989) earlier study to examine the extent to which the effect of training health care personnel had persisted. Disappointingly, although the recognition rate rose from 5.6% to 30% of female trauma patients during the initial study, eight years later, the rate had dropped to 7.7%. The only other replication (Roberts, O'Toole & Raphael, et al, 1996) did not deal with clinical recognition rates, but obtained consistent prevalence rates among all trauma patients of 7% and 8.5% of men, and 23% and 23.9% of women identified as abused. Despite the limitations, and differences between studies, it is reasonable to conclude that up to 30% offemale trauma patients have been abused, and that rates of clinical identification, both with and without protocols, are much lower than the number of women abused. Reasons for Non-identification The reasons for non-identification of violence are complex and are not well understood as they have not been specifically studied. Some explanations tend to blame the woman, focusing  31  on women's reluctance to disclose rather than the conditions under which disclosure may be made, and emphasizing the characteristics of women who are abused as influential. However, the factors which influence a woman's decision to disclose abuse and the factors that enable a health care professional to recognize abuse appear to be intertwined. The response a woman anticipates is likely to influence her decision to confide in a health care provider. Herman (1992) argues that people who have been traumatized, including survivors of abuse and war, assume, until proven otherwise, that a person in a therapeutic role cannot bear to hear the truth about the trauma. Stanko (1988) argues that reporting of abuse is impeded by women's belief that police are reluctant to intervene in domestic affairs, and women's fear of reprisal. Stanko also concludes that "part of women's silence around physical and sexual abuse in the home... can be attributed to the barriers preventing women from speaking of their homes as anything but sanctuaries" (p. 88). Women's reluctance to disclose abuse to health care providers may be influenced by the same factors; women may think health care providers do not want to hear, and may think that abuse is a private matter and that nothing can or will be done. Studies illustrating the impact of training on recognition rates further support the idea that the perceived reluctance of women to disclose abuse may be due to the behavior of the health care professionals. McLeer and Anwar (1989) found that with protocols and training of health care personnel, the percentage of female trauma patients recognized as abused rose from 5.6% 30%. Similarly, McFarlane, Christoffel, Bateman, Miller and Bullock (1991) found that by conducting personal interviews, recognition of abused women increased from 7.3% to 29.3% of women coming to a planned parenthood clinic. The similarities and differences between social identities of health care professionals and abused women may be a significant factor in determining whether or not abuse will be  32  recognized. Hampton and Newberger (1988) found that with regard to child abuse, "class and race are the important factors defining the gradient between reported and unreported cases" (p. 217). Such biases may also influence recognition of wife abuse. Campbell, Pliska, Taylor, and Sheridan (1994) described battered women's perceptions of racism, classism and negative attitudes toward drug and alcohol problems by hospital Emergency personnel, and Greif and Elliot (1994) found that Emergency nurses expressed similar attitudes when surveyed. Racism may be a particularly critical factor in a woman's decision to disclose abuse. Dobash and Dobash (1992) note that seeking assistance from authorities by women of color is seen as unacceptable because the response by authorities (e.g., arrest) is seen as "a further act of racial oppression against men of color"(p. 52). They further analyze the double bind saying that "women of color are in fact being expected to bear the brunt of gender violence within a racial or ethnic group in order that the group itself or its violent members not be exposed to further racial oppression" (p. 53). Acknowledging low levels of recognition, nurses have called for universal screening in which every woman seen by a health care professional is asked questions regarding whether violence is a problem in her life (Bullock, et al., 1989; Furniss, 1993; Grunfeld, et al. 1994; Humphreys, 1993; King, 1993; Lazzaro & McFarlane, 1991; McFarlane, Greenberg, Weltage & Watson, 1995). The studies to date suggest that training and protocols can improve recognition rates, but the experience of McLeer, Anwar, Herman et al. (1989) suggest that the effects are not lasting. Further, the rates of clinical identification tend to be low in all studies, ranging from 1.8% of female trauma patients, to 6% of all women, with the exception of McFarlane, et al. (1995) who identified abuse in 38% of women presenting with vaginal bleeding.  33  The literature on identification suggests that the problem of identification is complex, but is at least in part due to the behavior of health care professionals. The reasons for nonidentification have been suggested to be related to the attitudes of health providers and the messages they convey, as well as to the willingness of women to disclose abuse. If the problem of identification is conceived as partly a problem of recognition by health care professionals, rather than only as a problem of disclosure on the part of the woman, then the responsibility for recognition might be partially shifted from the woman who has been battered to health care providers and institutions, and the behavior of the latter becomes a target for study and change. Detrimental Responses to Women Who Have Been Battered Even when women are identified as having been abused, the response by health care professionals and institutions is less than helpful. Kurz and Stark (1988) found that while 4% of women identified as "positive" for abuse were hospitalized (twice the rate for "non-abuse" injuries), 47% were simply returned home and only 12%> were referred to appropriate social services. Remarkably, 8% of the women identified as "positive" for abuse were r