Interventions for Intimate Partner Violence Against Women: Clinical Applications

Sufficient evidence regarding the effectiveness of physician screening, counseling, and referral for intimate partner violence is lacking. However, even in the absence of sufficient evidence, many medical organizations recommend that physicians make efforts to identify and refer patients who have experienced abuse. This article presents 3 cases that illustrate the interaction between patients experiencing violence in intimate relationships and physicians and other members of the health care team. Suggestions for care are based on guidelines and evidence where available. Resources to assist physicians with referral to appropriate services also are provided.

In this issue, Wathen and MacMillan¹ present an evidence-based review of interventions for violence against women. Interventions that have been evaluated include screening, referral programs for battered women, court-mandated programs for men who perpetrate partner violence, social programs, and educational interventions. None of the studies have evaluated, in a comparative or a controlled design, the effectiveness of screening for violence in the primary care setting in terms of outcomes for abused women, that is, none of the screening studies met all inclusion criteria for the review conducted by Wathen and MacMillan. Studies of other interventions, including counseling for abused women, effectiveness of shelter stays, and advocacy and counseling following shelter stays, have demonstrated fair, insufficient, or unclear evidence for their effectiveness. Given the lack of sufficient evidence about effective interventions by primary care clinicians, what should physicians do to address intimate partner violence (IPV) in patients?

One of the most important questions is whether primary care physicians should screen all women for IPV. Theoretically, screening is a low-risk, low-cost procedure with reliable screening tools.² If physicians ask about abuse, women often will disclose the problem,³ and exploratory qualitative research indicates that inquiry about abuse may be of benefit to abused women.⁴ However, because physician time is valuable, time spent on screening or counseling for nonevidence-based conditions can take time away from other physician tasks of proven benefit. To date, no rigorous studies demonstrate an improvement in outcomes from screening for domestic violence in the primary care setting.⁵

A 60-year-old white woman from an affluent upper middle-class suburb came to the emergency department (ED) complaining of chest pain. Her medical record revealed multiple hospitalizations to rule out myocardial infarction and several series of cardiac investigations including a recent cardiac catheterization. The ED physician did not routinely screen women for abuse but when the clinical diagnosis was confusing, she would ask more detailed questions about the woman's social history, including a standard question about potential abuse. In this case, she uncovered years of previously undisclosed emotional and physical abuse by the spouse. The patient was not willing to contact domestic violence support services but gave permission to discuss the abuse with her primary care physician. The patient wanted her primary care physician to know about the violence but she had been too ashamed to volunteer the information. Her primary care physician expressed surprise when the ED physician told him about the abuse history.

According to a random representative survey of approximately 2500 women conducted for the Commonwealth Fund in 1993, only 8.4% of women who reported experiencing abuse had ever told a physician, and less than 50% had told anyone.⁶ According to several studies, the majority of women experiencing IPV present to both EDs and primary care settings with noninjury-related complaints.^7,8 In these settings, a history of IPV is rarely identified.^9,10 Case scenario 1 is an example of a case finding in which the physician asked direct questions about violence because of a confusing clinical picture. The case finding approach requires the physician to have a low threshold for asking questions about IPV.¹¹ Many medical organizations suggest use of a case finding or use of aggressive case finding with appropriate referrals when domestic violence is detected.^12-14 However, some organizations recommend routine screening for IPV based on the prevalence and health burden of IPV and the feasibility of detection with routine screening.^15,16 In addition, the identification of an abuse history may influence the assessment and treatment of presenting health concerns.^17,18 Indeed, evidence suggests that failing to detect a patient's abuse history may be associated with the use of unnecessary investigations and interventions.^19,20

A variety of direct, gender-neutral questions can be useful for addressing the issue of IPV. Some examples are, "Has your partner ever hit you, shoved you, or otherwise physically hurt you?" or "Are you in a relationship with anyone who has hurt or threatened you?" If the initial question is answered affirmatively, the physician can follow up with questions regarding the nature of any injuries. In addition, questions regarding issues of power and control are helpful to elucidate the nature of the abuse. Example questions include "Is your partner very jealous or controlling?"; "Does your partner keep you away from family and friends?"; "Can you come and go as you please?"; and "Has your partner ever made you have sex when you didn't want to?" Common sense regarding safety and confidentiality dictates that these questions be asked in a private setting.

A 41-year-old Hispanic woman was seen by her family physician in a large multispecialty clinic. She had been evaluated multiple times in the previous 6 months for asthma and symptoms diagnosed as anxiety or panic disorder. Her physician had twice referred her to a psychiatrist but she had failed to follow-up. On this visit, the physician had a medical student in the office. The student learned that a marital fight had triggered the woman's asthma. On questioning by the family physician, the patient admitted to emotional abuse but denied physical abuse. The physician did not have enough time to adequately assess her situation or counsel her but asked the clinic's part-time social worker to see her. The social worker reported there was significant physical abuse as well but said the woman was not willing to go into a domestic violence shelter. She had children at home and no independent financial support. The social worker encouraged the woman to talk to her extended family and referred her to a community-based IPV support group.

When clinicians discover IPV, what should they do? Qualitative evidence from interviews and focus groups with abuse survivors suggests that even brief discussions with a physician, conducted in a concerned and nonjudgmental fashion, can help to change the way abused women view their situations, even if they do not disclose the abuse.^21,22 Women have reported that a discussion with a physician who acknowledged the abuse and validated their self-worth was a turning point in the process of extrication from an abusive relationship.²³ However, the process of changing any health behavior takes time and may happen gradually or not at all.²⁴ The primary care physician may help initiate thoughts of change by discussing the abuse, but patients may take weeks, months, or years before they contact community resources or initiate changes in their lives; and they may never take any such action. In the end, physicians may never know whether they were able to help the patient.

While primary care physicians frequently provide brief advice for other health risks, such as alcohol use,²⁵ many do not feel comfortable counseling women about abuse and some may wonder if their efforts are worthwhile.²⁶ Intensive screening programs with on-site advocacy and counseling have been shown to help identify abused women and link them to resources.²⁷ Wathen and MacMillan¹ identified a study in which domestic violence advocacy and counseling following a shelter stay were effective in improving abused women's quality of life.²⁸ In addition, civil orders of protection have been found to be effective in decreasing incidents of violence.²⁹ However, abused women will present in various stages of denial or readiness to change their situation.³⁰ Like the woman in case scenario 2, many women may understand their situation and yet choose to remain in the relationship for complex reasons, not the least of which is an increased risk of becoming a homicide victim in the 6 months following a separation from a violent relationship.³¹ Physicians who try to encourage women to make changes may find this frustrating.³² Ideally, they could rely on trained social workers or domestic violence advocates for help with assessment, counseling, and referral to supportive legal or counseling resources,³³ but such trained personnel often are not available.

If physicians learn about abuse, they should be prepared to acknowledge it and make appropriate referrals. In addition, physicians should be aware of state reporting or other requirements. A link to these resources is provided in Box 1 along with other Web-based domestic violence resources to assist physicians in identifying resources in their community.

Box Section Ref ID

A 26-year-old black man presented to an urgent care clinic for the third time with "back and neck pain." At this visit, he was screened for possible IPV and depression as part of a computer self-administered health risk assessment, which ideally could be given to all clinic patients.³⁴ The treating physician received an alert on the risk summary sheet generated by the computer indicating that the patient was possibly involved in partner abuse and that he had symptoms of depression. A chart review showed that, in addition to evaluating the patient for chronic sciatica, another physician had previously documented the patient's depression and difficulty controlling his temper. The treating physician prescribed a nonsteroid medication and referred the patient to a specialist to further evaluate his back and neck pain. The physician also documented a difficult home situation and a strained relationship with his current wife and referred the patient to a community mental health program, which included a batterers treatment program.

Several studies that have reported that men screened for IPV had similar rates of partner violence as women.^35,36 Others have concluded that many of the men disclosing being abused were abusers as well.^37,38 Given the cyclic nature of battering and that the majority of partner abuse will never be reported to police,³⁹ primary care physicians need to know what to do if a male patient either directly or indirectly discloses conflict with his partner. These disclosures may occur when asking about a source of stress. In addition, some patients may recognize they have a problem with controlling their anger.⁴⁰

Typically, perpetrators of violence may portray themselves as the recipients of abuse or present the situation as one of mutual combat.⁴¹ These patients, or their partners, may ask physicians about programs for marital problems or anger management. Wathen and MacMillan¹ identified an important intervention study for men who assaulted their wives, the San Diego Navy Experiment,⁴² that did not find any difference between treatment and control groups for prevalence of reabuse but did find that on-going surveillance by a commanding officer resulted in low rates of recidivism. A randomized clinical trial by O'Farrell and Fals-Stewart has shown a reduction of IPV as a result of couples behavioral therapy for alcoholism.⁴³ However, primary care counseling and referral of men who experience and perpetrate IPV has not been evaluated for either safety or effectiveness, and there are no current recommendations about screening male patients for partner abuse.⁴⁴

If violent conflict is voluntarily disclosed, the physician should treat the violent behavior in a concerned but matter-of-fact manner, similar to other unhealthy behaviors. This includes assessing the pattern of abuse, risk of lethality, and patient willingness to change. It is important that the physician does not collude in blaming the spouse or partner but holds the perpetrator responsible for their own behavior.⁴⁰ The physician can focus patient education on the adverse consequences of family violence, for everyone involved, and make appropriate referrals to community resources.

The 3 cases presented highlight the inevitability that physicians will be caring for patients who may be experiencing violence in their relationships. What are the clinical implications when evidence is lacking? Many medical organizations offer clinical guidelines that are based on both a review of the evidence and the consensus of expert opinion.⁴⁵ Consensus opinions can take into account both clinical context and professional experience and may be useful when research is lacking.⁴⁶ Guidelines on domestic violence have been published by the American Academy of Family Physicians (1996),¹⁶ the American Academy of Pediatrics (1998),¹³ the American College of Emergency Physicians (1990-1991),⁴⁷ the American College of Obstetricians and Gynecologists (1995),¹² the American Medical Association (1992),¹⁵ and Joint Commission on Accreditation of Healthcare Organizations (1993).⁴⁸ Most of these recommend a concerted effort at case-finding and a structured approach to documentation and referral.Box 2 contains a list of most of these medical organizations and their current practice guidelines as well as a information about the level of evidence for interventions from the Canadian Task Force on Preventative Health Care (2001)⁴⁹ and the US Preventive Services Task Force (1996).¹⁴ These guidelines were developed as an acknowledgment that domestic violence is a highly prevalent condition^50,51 with significant morbidity,⁵² mortality,⁵³ and potential for the intergenerational transmission of violence.^54,55

Box Section Ref ID

The accompanying review points out that much work remains to be done to establish whether physician screening, counseling, and referral in the primary care setting results in a decrease in violence against women. Domestic violence services need to undergo rigorous evaluation to assess their effect on a variety of short and long-term safety and quality of life outcomes for battered women and their children. For perpetrators of partner violence, research should focus on evaluating the effectivenss and safety of interventions not only for court-mandated batterers but also for those willing to disclose problems with violence to physicians and other members of the health care team. Nonetheless, physicians will continue to see both women and men who are abused by, or perpetrating, partner violence. Their identification may influence the evaluation of presenting complaints as well as the outcomes of care. Until there are more studies that meet evidence-based criteria, physicians should continue following clinical guidelines, increasing efforts to identify women in violent relationships, and linking them to appropriate community resources.

Corresponding Author: Karin V. Rhodes, MD, Section of Emergency Medicine, Department of Medicine, The University of Chicago, 5841 S Maryland Ave, MC 5068, Chicago, IL 60637 (e-mail: krhodes@medicine.bsd.uchicago.edu).

Funding/Support: This work was funded by the Agency for Healthcare Research and Quality, grant No. HS11096-02

Disclaimer: Dr Levinson, who edits the Scientific Review and Clinical Applications column, was not involved in the final review or editorial decision to accept this work for publication.

We encourage authors to submit papers to "Scientific Review and Clinical Applications." Please contact Wendy Levinson, MD, Contributing Editor, JAMA; phone: 312-464-5204; fax: 312-464-5824; e-mail: wendy.levinson@utoronto.ca.