Perpetration of violence against intimate partners: health care implications from global data

Intimate partner violence is a widespread global phenomenon engendering serious health problems for women and children.1^–3 Consequently, there have been many recommendations to identify and intervene with victims in the health care system. Although evidence of the efficacy of interventions for victims is still lacking, clinical trials are under way, and well-validated assessment strategies and protocols have been developed. In contrast, no protocols have yet been established for identifying perpetrators. This gap may be particularly important given that at least 20% of abusive males in one sample had been seen in the mental health system before killing their spouses and then themselves.4 Such patterns of intimate partner homicide–suicide have been documented in many countries, including South Africa,5 the setting of the study by Gupta and colleagues6 appearing in this issue of CMAJ.

This new study, based on data from the South Africa Stress and Health Study, is extremely important in helping us to identify the characteristics of perpetrators of intimate partner violence and reinforces the significance of exposure to childhood violence as a risk factor, as found in an earlier regional South African study.7 The study also confirms the high levels of violence reported previously by both perpetrators8 and victims9 in South Africa. However, the prevalence of intimate partner violence that Gupta and colleagues report is lower than in one of those prior studies,9 and the provincial differences described are not consistent with other South African studies on intimate partner violence or other crime.10 These differences may reflect how the fieldwork was done, but they may also indicate that adding a few questions about intimate partner violence to a survey designed primarily for other purposes yields less complete reporting than studies dedicated to intimate partner violence, as has been demonstrated previously.10 Thus, although studies like this one add substantially to our knowledge about the relations between intimate partner violence and other variables, they are less useful for determining prevalence. This limitation has important implications for Canada, which has not repeated its groundbreaking national survey of spousal assault, conducted in 1995,11 but rather has relied on general crime surveys to estimate the prevalence of the problem.

Measuring exposure to community and structural violence (e.g., war, refugee situations) is an important addition to studies of perpetration of intimate partner violence. However, the survey on which the present study was based did not measure direct experiences of community violence; instead, respondents were asked to estimate the level of crime in their own areas. A better study would measure not only direct exposure to such violence but also the associated mental health consequences. US research has indicated that unresolved post-traumatic stress disorder among veterans returning from the Vietnam conflict led to increased violence toward their wives and children.12 This finding has important implications for both Canada and the United States, whose veterans are returning from current conflicts overseas. In addition, further work is required among immigrant refugee populations to identify the need to treat post-traumatic stress disorder and to address the potential for intimate partner violence during such treatment. Gupta and colleagues may have missed an opportunity to explore the relations between perpetration of intimate partner violence and the mental health variables measured in the South Africa Stress and Health Study, which ideally would have included post-traumatic stress disorder. However, they do reinforce the importance of preventing intimate partner violence by intervening with those who have experienced abuse during childhood or who witnessed intimate partner violence in their childhood homes. The link between those exposures and perpetration of intimate partner violence in adulthood may relate, at least in part, to a failure to treat post-traumatic stress disorder resulting from the exposures.

The study by Gupta and colleagues was an interdisciplinary cross-national collaboration. Such collaborations are crucial for learning about health issues that are specific to low-and middle-income countries but that also have global relevance. For instance, men's use of physical and sexual violence toward their intimate partners is contributing to the HIV epidemic in South Africa and globally.13 International collaborations have also illustrated the importance of interventions addressing intimate partner violence from a community or structural perspective. In South Africa, one approach involves the provision of microcredit financing to women, along with an intervention aimed at changing community norms;14 another works directly with young men to change their attitudes toward and behaviour involving intimate partner violence, as well as to prevent HIV/AIDS.15

The most important implications of the study by Gupta and colleagues and other research on intimate partner violence is that the health care system must be involved in addressing this widespread problem. Medical and nursing curricula need to cover intimate partner violence and its health consequences, providing evidence-based clinical information about how to assess women for victimization and encouraging increased use of forensic sciences and exposure to clinical forensic practice. The curricula should include detailed information on traumatology and wound identification, to allow practitioners to appropriately identify and document injuries, as well as to develop the ability to offer opinions on their nature and causation.16 Clinicians also need to start developing assessment questions for use in identifying perpetrators, allowing for appropriate evaluation of psychometric properties but also attending to any potential threat to the safety of their partners that might result from such assessment.

@@ See related research paper by Gupta and colleagues, page 535

Key points

• Intimate partner violence is widespread and has serious health consequences.

• Well-validated assessment protocols are available to identify the victims of intimate partner violence in the health care system, but not for identifying the perpetrators of such violence.

• Medical and nursing curricula throughout the world should include information on the health consequences of intimate partner violence. These curricula should also cover the various modes of assessment, including the use of forensic science methods in identifying injuries.

• Future research should examine the links between direct exposure to community and structural (e.g., war) violence and the perpetration of intimate partner violence and should address all aspects of the intersection of gender-based violence (including intimate partner violence) and HIV/AIDS.