|Home | About | Journals | Submit | Contact Us | Français|
Intimate partner violence (IPV) is defined as “any behavior within an intimate relationship that causes physical, psychological or sexual harm to those in the relationship.”1 Its role as a major public health issue affecting both men and women, regardless of social class, sexual orientation, or racial/ethnic group, was reinforced by the recently released population-based 2010 National Intimate Partner and Sexual Violence Survey.2 Results indicate that women have a lifetime IPV prevalence rate of 29%, with 1 in 7 injured by an intimate partner. With a lifetime prevalence rate of 10% (1 in 25 injured as a result of IPV),2 men are significantly less likely than women to experience physical or sexual IPV.3 While less is known about the long-term impact of IPV among male victims, both male and female victims of IPV experience high rates of adverse physical, social, emotional, and mental health outcomes.2-5 Being a victim of IPV is strongly associated with a host of behavioral risks such as sexually transmitted diseases, human immunodeficiency virus infection, and the use of tobacco, alcohol, and drugs and is linked to all other forms of violence.1 As a significant contributor to health, social, and economic disparities, IPV jeopardizes the fabric of families and transcends all levels of socioeconomic status. Due to high rates of health care utilization by both victims and perpetrators of IPV,6,7 health care settings have abundant opportunities for early identification, intervention, and secondary prevention of IPV—but have consistently failed in these endeavors.
A recent report, Clinical Preventive Services for Women: Closing the Gaps, that was published through the Institute of Medicine (IOM) recommends that IPV screening become part of routine preventive services for women.8 The IOM report is the latest in almost 20 years of similar clinical recommendations and guidelines targeted at improving IPV screening and response in health care settings. In 1992, the American Medical Association recommended screening all women for IPV,9 and other professional medical organizations widely endorsed these recommendations.10-13 However, the professional practice guidelines preceded both the evidence of effectiveness and the considerable implementation challenges. Studies that tried to assess the effectiveness of IPV screening quickly found that physicians were unable or unwilling to integrate IPV screening into their practice.14 Substantial barriers were identified that span the spectrum of physicians' knowledge, attitudes, beliefs, and behaviors.15-20 Even system-level interventions with extensive physician education demonstrated marked improvements in screening without an increase in IPV identification.21
During this time, the movement for the use of evidence to guide clinical practice was also developing. With the realization that many of our current treatment recommendations lacked high-quality evidence of effectiveness, the level of enthusiasm for IPV screening began to decrease. The inescapable fact is that, in spite of the well-acknowledged high prevalence of IPV exposure and knowledge of its adverse health consequences, these recommendations have been difficult to implement, and general discouragement has followed a chain of evidence-based reviews that have found no evidence that screening for IPV in health care settings reduces subsequent IPV or improves women's health.22-26 Concerns about taking limited physician time for screening that is not clearly associated with improved outcomes are understandable. However, recent evidence from several randomized clinical trials and new developments in the use of health information technology (HIT) have the potential to dramatically shift the balance in favor of routine IPV screening. Moreover, changes in the financing and restructuring of the way clinical care is provided and quality is measured are shifting the scope of health care system responsibility and opening up new opportunities for routine IPV screening in health care settings in ways that do not take physician time.
Before the 2009 study conducted by MacMillan et al,27 there were no large randomized controlled trials addressing the impact of IPV screening in health care settings. The MacMillan trial recruited 6743 women, aged 18 to 64 years, from 26 emergency departments and primary care sites. Participants were randomized into 2 groups: they either completed a validated IPV screen before seeing the physician or were placed into a “nonscreened” group. However, both groups were screened for IPV at the end of the visit and were given referrals to appropriate resources. The participants were then followed up for 18 months and were monitored for primary outcomes of exposure to future abuse and quality of life during that time. Secondary outcomes included depression, posttraumatic stress disorder, alcohol and drug abuse, global mental and physical health, and use of health and social services. One of the study goals was to specifically assess for any harms that were related to screening. During the initial clinic visit, 44% of currently abused women who were screened before seeing the physician and 8% of the control group (abused women who were screened at the end of their visit) discussed IPV with their physician. Women in both groups had reductions in IPV recurrence and improvements in quality-of-life scores. In addition, depression, posttraumatic stress disorder symptoms, and alcohol problems all improved; however, there were no statistically significant differences between the 2 groups. Unfortunately, loss to follow-up was more than 40% in both groups, and the women who were at higher risk and had greater abuse at baseline were more likely to be lost to follow-up. This loss to follow-up rate led the authors to speculate that it might be impossible to obtain adequate long-term follow-up with abused women. The significant investment made by the Canadian government to conduct this study resulted in the key finding that there was no evidence of harm related to screening.27
I would contend that the main finding of the MacMillan trial was that there is no support for physician intervention, since that was the only real difference between the groups. In contrast, the MacMillan study does offer observational support for the positive impact of systematic IPV screening and referral since all women in the trial were screened, and for ethical reasons, all abused women received referrals. In fact, there was no real “usual care” group, since actual usual care does not involve either IPV screening or referral. While this observational finding would not be enough to tip the balance of evidence in favor of routine IPV screening for the next United States Preventive Services Task Force review, there is new evidence cited by the IOM that may do this.
Specifically, the strongest evidence cited by the IOM comes from the recently completed National Institutes of Health-DC Initiative to Reduce Infant Mortality in Minority Populations, which compared perinatal counseling to usual care.28-30 The study enrolled 1044 pregnant African American women at 6 prenatal care sites in the District of Columbia and monitored them for 1 year. Screening for IPV and other psychosocial risks was done using a computer-based risk assessment. Abused women were randomly assigned to receive perinatal counseling before and after the birth of their baby. Social work intervention was delivered during routine prenatal and postpartum care by specifically trained social workers. Counseling for IPV emphasized danger assessment, safety behaviors, and information on community resources. Comorbid risks such as smoking and depression were also addressed. Results were very promising, and the high retention rate of 91% of enrolled abused women indicates confidence in the results. At follow-up, women in the intervention group had significantly fewer recurrent episodes of IPV during pregnancy and post partum (adjusted odds ratio, 0.48; 95% confidence interval, 0.29-0.80). Women in the intervention group also had better birth outcomes, including fewer preterm neonates (≤33 weeks) (1.5% vs 6.6%; P=.03); lower rates of very low-birth-weight neonates (<1500 g) (0.8% vs 4.6%; P=.52); and increased mean gestational age (38.2 weeks for the intervention group compared to 36.9 weeks for the control group; P=.016). The trial also found that depression, smoking, and environmental smoke exposure decreased in the intervention group.28-30 The authors hypothesized that the integrated nature of the intervention that addressed depression along with IPV and other psychological/behavioral risks may have greatly contributed to the intervention's success.
With the new knowledge that routine screening in conjunction with ongoing social work counseling is effective in reducing prenatal IPV and comorbid adverse health behaviors and is associated with improved pregnancy and birth outcomes, the weight of the evidence is finally tipping toward routine IPV screening. The caveat is that this type of intervention has only been shown to be effective among young urban women engaged during the perinatal period in a close ongoing relationship with a medical facility that has high-quality IPV-trained social work resources. Therefore, more work needs to be done before we can make evidence-based recommendations for routine IPV screening of women in all health care settings.
Another factor that is impacting the cost-benefit ratio and improving the feasibility of routine IPV screening is the dramatic increase in the use of electronic health records and personal health records. Personal health records in particular will lower the cost of routine IPV screening and ensure that no one who is at risk “falls between the cracks” because of its ability to offer patient-completed personal health risk assessments combined with online educational materials and personalized referrals to local resources. This new HIT has the potential to increase the confidentiality of the IPV screening process and allow health systems to ensure that every patient who completes the risk assessment and screens positive for IPV will also get appropriate referrals, regardless of whether they choose to share their IPV screening results with their physician. There is also the very real possibility of developing and testing HIT-based IPV counseling interventions using tailored messages. Prototypes of these sorts of electronic interventions are already being developed and tested for a variety of behavioral risks, from substance abuse to depression. Innovative violence prevention projects have been tested with teens and have been found to be a safe, acceptable, and potentially effective means of intervention.31,32 While perhaps it is not the preferred method of counseling, HIT-based interventions have a number of advantages, including compensating for inadequate social work resources, minimal time expenditure by practitioners, adaptability to low-literacy and non-English-speaking patients (especially if combined with audio), ability to provide individualized feedback, high treatment fidelity, and easy dissemination across a wide variety of clinical and nonclinical settings.
Nor would we need to restrict IPV screening to women patients, as there is no evidence that computer screening is harmful for either women or men.33 While we should continue to design and conduct good studies and gather rigorously collected data to assess the evidence for or against our ever-evolving clinical practices, HIT allows us to begin to routinely screen both men and women for both victimization and perpetration of violence and abuse. The IPV that is disclosed in health care settings (as opposed to criminal justice settings) can be assessed for level of risk and safety and addressed as an unhealthy relationship behavior, and patients can be referred or directly linked to the appropriate level of resources or treatment. For patients who want their physician to know about their psychosocial risks and would prefer to talk to a person, the use of HIT that allows for self-disclosure of IPV has the potential to minimize the information-gathering aspect of a health care visit and maximizes the time for physicians to spend responding to the patient's self-identified health needs.
Both cost and quality considerations should be included in decisions about IPV screening. Max et al34 estimated the 2004 costs of IPV in the United States to be more than $8.3 billion, indicating that potential savings from efforts to reduce IPV could be substantial. A great deal of these costs are attributable to the greater health care expenditures associated with IPV.7,35 After adjusting for comorbidity, Ulrich et al36 found that abused women had between 1.6- and 2.3-fold increases in total utilization and health care costs. Kothari and Rhodes37 reported that police-identified women victims of IPV had 3 times the annual rate of emergency department use compared to a population-based sample of women in the same age category, and the vast majority of visits were for medical complaints, not injury. In capitated managed care organizations, researchers have identified that patients who have experienced abuse have annual medical costs that are much higher than those of never-abused patients, with those most recently abused having the highest past-year costs.4 Adjusting for demographic characteristics and other causes of health care utilization, Bonomi et al4 found that, compared with never-abused women, total adjusted annual health care costs are 42% higher for women with ongoing physical abuse. Even women with nonphysical abuse have 33% higher total annual health care costs compared with never-abused women.6
The recent health reform legislation, the Patient Protection and Affordable Care Act (ACA), ambitiously strives not only to reduce costs but also to improve the quality of the US health care system.38 The ACA has the potential to affect financial incentives in ways that can modify health care delivery systems and incentivize health system accountability for population-based and preventive health care, as opposed to just sickness care. Although it is important to consider the possibility that the law will be changed in ways that support the status quo, the core concepts of the ACA are admirable. While it would be naïve to think that the system as a whole will reform overnight, the ACA includes some incentives for delivery system reform, including the creation of accountable health care organization pilots, bundling of payment for services, creation of medical homes, hospital value-based purchasing, capitation, and quality measures that favor patient-preferred health outcomes. In addition, the ACA includes increasing funding for comparative effectiveness and implementation research. The act also establishes an independent nongovernmental group that is seeking stakeholder input for refining a mission devoted to patient-centered outcomes research.39 Given the broad-based support for screening and addressing IPV and among patients,18,22 it is highly likely that any stakeholder-informed agenda seeking to improve population-based health will support health care interventions for violence and co-occurring psychosocial risk factors. However, at this point it is not yet clear how the movement toward population-based care and patient-centered outcomes will be operationalized and how the incentives can be ultimately aligned to help hospitals and health systems determine how to deliver better health care in ways that reduce costs and improve the health of communities.
The evidence in support of routine IPV screening, the feasibility of doing so, and the political will for addressing violence and related risk factors are rapidly evolving. Even though we are not there yet, after more than 20 years of encouraging busy physicians to screen for IPV, innovations in HIT and changes in the organization and delivery of health care have the potential to provide the system-level support needed to make this a reality.