Intimate partner violence screening in the emergency departmentJD Daugherty, DE Houry
Department of Emergency Medicine, Emory University, USA
Background and Aims: Every year between 1.5 and 4 million women are abused by a partner in the United States and many abused women turn to the Emergency Department (ED) as their first source of care. Even though the vast majority of patients would feel comfortable disclosing intimate partner violence (IPV) to their physician, identification and referral is inconsistent . Aims: The aim of this paper was to discuss prevalence statistics of IPV, current screening recommendations and practices in ED settings, and future directions to improve the screening and identifying of victims of IPV that present to the ED. Material and Methods: The authors conducted a Medline search for articles discussing IPV screening in the ED. Results: Intimate partner violence results in approximately 1,300 deaths and 2,000,000 injuries annually among women and up to a third of ED patients have a history of IPV. Despite patients' reported willingness to disclose this information, identification of IPV by healthcare practitioners remains very low, with some estimates ranging between 4-10%. Conclusions: If we do not identify victims of IPV in the ED, this may result in continuation of the abuse, routine returns to the ED for treatment of injuries, and possibly even death.
Keywords: Emergency department screening, intimate partner violence
Over the last several decades intimate partner violence (IPV), commonly referred to as domestic violence, has emerged as a serious public health problem. Intimate partner violence is defined as threatened, attempted, or completed physical or sexual violence, intimidation, or emotional abuse by a current or former intimate partner.  Every year between 1.5 and 4 million women are abused by a partner in the United States. , Intimate partner violence results in approximately 1,300 deaths and 2,000,000 injuries annually among women. ,, It is associated with several serious health conditions such as HIV, ,, mental health issues, ,, chronic pain  and substance abuse, , in addition to resultant physical injury from abusive acts.  Furthermore, people who report having experienced IPV anytime in their life are more likely to report current adverse health conditions and health risk behaviors,  and are more likely to be victimized in the future. , Although IPV is present in every segment of the population, regardless of race, income level, gender, sexual orientation, or age, there are some groups that experience a higher prevalence than others. For instance, Black and low-income women are more likely than White or high-income women to report instances of IPV.  Importantly, IPV appears to be especially prevalent in the Emergency Department (ED) setting. Victims of IPV may not have access to healthcare elsewhere and may present to the ED for an injury or the health and mental sequelae of abuse.  Furthermore, the ED often represents the IPV victim's first introduction to the healthcare setting. Ernst and colleagues reported that 33% of the female ED patients they studied had been a victim of physical IPV at some point in their lifetime and 22% had experienced non-physical IPV previously.  In a review of current studies taking place in EDs, Anglin et al. , found that 1-7% of all female patients present to the ED on account of an acute episode of physical abuse by their partner.  Perhaps most disturbingly, a review of cases of women murdered by a partner found that 44% had visited the ED within two years of their death. Of these, 93% had at least one visit to the ED related to an injury. 
Currently, the optimal protocol for screening in the ED is still not known.  Over 10 years ago, the United States Preventive Services Task Force (USPSTF) first examined the question of IPV screening in general medical settings. They subsequently assigned screening a "C" recommendation meaning they could not make a clear determination of its effectiveness  and in 2004 an "I" signifying there was insufficient evidence to make any recommendation for or against it.  Despite this, the Joint Commission on the Accreditation of Healthcare Organizations (JCAHO) has mandated since 1992 that all hospitals have policies and procedures for the early identification and referral of victims of abuse.  However, many are not in compliance. A study conducted in California found that the majority of hospitals in that state were not compliant with this recommendation.  If we do not identify victims of IPV in the ED, this may result in continuation of the abuse, routine returns to the ED for treatment of injuries, and possibly even death . 
Screening for IPV in the ED is occurring, but it is not done consistently. Many healthcare providers admit that they lack the confidence that they can adequately screen for IPV in the ED setting  and that they are not always aware of how to refer women with IPV to local resources. Whatever the case, many female victims of IPV report that they are not asked about their abuse  even though most, up to 89%, would feel comfortable disclosing IPV to their physician. ,,
Despite patients' reported willingness to disclose this information, identification of IPV by healthcare practitioners remains very low, with some estimates ranging between 4-10%. , Similarly, the presence of IPV is rarely noted in patients' medical records. Roberts et al. , reported that out of over 300 patients, only 50% of those that self-disclosed IPV in the last 24 h had this fact documented in their medical record.  Likewise, Houry and colleagues looked at a cross-sectional sample of patients presenting for assault in an urban Level 1 trauma center and found that data was missing from the medical record in the majority of cases. In 67% of records the identity of the assailant was not noted and in 79% of cases there was no mention of the location of the assault.  Although universal screening for IPV has not proven itself to be completely effective, screening creates the opportunity to radically increase the rate of IPV detection and referral to outside organizations.  However, to accomplish this, healthcare workers must consistently implement a screening protocol as well as document the presence of IPV in the patient's medical record.
What is known about increasing the effectiveness of IPV screening in the ED? We know that many times victims will not offer the information voluntarily; they must be asked directly if they are experiencing IPV at home. , We also know that IPV victims present with certain medical conditions more than others without an IPV history. For instance, IPV victims are more likely to have repeated, frequent use of the ED, present with vague complaints of pain, have a medical history rife with "accidents", disclose suicidal ideation, have alcohol and/or drug abuse issues, , depression  or have symptoms of post-traumatic stress disorder (PTSD)  In addition, certain injury patterns have been highly associated with IPV. For example, studies have shown that between 22-35% of all facial trauma victims have been abused or assaulted , and abused women are more likely to be injured in the head, neck, thorax, and abdomen than were women injured by other mechanisms.  Therefore it may be useful to look for a cluster of symptoms in conjunction with the screening tool to determine if IPV is present.  Despite this, it is critical to screen all women that present to the ED, not just those that present with complaints of injury.
Screening for IPV in the ED is not always an easy task. There are difficulties from both the patient and physician aspects. Perhaps, most importantly, not all patients are willing to disclose if they are experiencing violence in their intimate relationships,  particularly if there are mandatory reporting laws in his/her state.  However, Houry et al., found that only a small minority of patients (12%) would be less likely to seek care in the ED because of the mandatory reporting laws and 27% reported that they would actually be more likely to seek medical care because of these laws.  Advocates of mandatory reporting laws, which require a physician to notify law enforcement if an injury from IPV is noted in a patient, believe that these laws lift the burden of reporting off the victim to an outside party and start the process of legal ramifications for the perpetrator.  Currently, only seven states have laws that specifically mandate reporting of injuries that result from IPV. 
When looking at screening and reporting, we need to ask if it is placing the patient in further danger. Many victims fear the very real possibility of retaliatory violence if he/she discloses IPV to an outsider. , A patient also often deals with feelings of shame, embarrassment, fear that he/she won't be believed, loyalty to the abusive partner, and intimidation by the perpetrator. [18, Another concern with screening is those patients with lower severity of illness or injury are more likely to be screened for IPV than those that are sicker or have a more serious injury.  However, IPV victims have more medical and mental health issues than those who are not victims. , Patients who present to the ED specifically for injuries related to an episode of IPV may be missed for exactly this reason.
On the health practitioners' side, physicians sometimes only screen for IPV if an injury is clearly visible. ,, However, neither demographic nor health factors can accurately predict whether a patient is suffering from IPV victimization - all races, ages, socioeconomic and health statuses are affected.  Healthcare practitioners often have a lack of comfort with the issue of IPV. Many physicians cite a fear of offending and of powerlessness as reasons for not routinely screening women for IPV. , Additionally, physicians feel uncomfortable screening a patient for IPV if they are not sure what to do with them if the patient screens positive and discloses IPV. Social services are often unavailable at night when many IPV victims present to the ED.  Healthcare providers must have resources available for referral if patients do disclose victimization; otherwise identifying a victim of IPV but not providing her with resources or ensuring her safety could be harmful.  Thus it becomes important to screen for IPV at multiple levels during a patient's stay in the ED. Both nurses and physicians should be involved in IPV screening from triage to discharge as a way to increase the number of patients willing to disclose incidents of IPV.
There are ways to ameliorate these provider-directed issues, the primary method through education. However, as the Institute of Medicine noted in its recent report, the education of physicians about IPV is sporadic, variable, and inconsistent with limited evaluation data on the long-term effects of training.  Most research that looks at educational programs for healthcare providers finds improvement with knowledge, confidence, and ability to identify IPV victims.  Studies have demonstrated that physicians who receive adequate training on the identification of IPV are more likely to screen for it than those that do not receive adequate training.  Approximately 80% of US medical students believe they are receiving adequate training on IPV  but only one-fifth of students would qualify the training as being extensive.  Additionally, there is also evidence that the information gained from IPV education programs is not retained; therefore booster courses should be implemented in medical school or provider trainings. 
Despite many physicians feeling less than comfortable with the idea of screening for IPV, it is being done successfully in some locations. What instruments are being used? How is it conducted? To be effective in an ED setting, an IPV screening instrument must be short and direct; validated IPV victimization screeners are much too long to be practical for use in an ED.  An optimal instrument at this point is not known  and most investigators that study IPV develop their own instruments from validated measures. Many combine questions from validated screeners such as the Conflict Tactics Scale or the Universal Violence Prevention Screening Protocol. However, standardization is desired and more work is needed to determine which questions are most effective for identifying true victims of IPV.
There are mixed findings from several recent research studies. After requiring nurses to screen for IPV and having them complete an educational session on IPV, less than one-third of patients were screened.  A promising idea was developed by Olson et al . The research team restructured the medical chart to add a stamp with the question "is the patient a victim of intimate partner violence?" The proportion of cases identified was 1.8 times higher than in control months, just from stamping that one question on the chart. 
This is certainly another simple, effective idea that should be further investigated. Long surveys are not practical in the ED and a screener is just as accurate at determining victimization compared to a longer interview.  In addition, Houry and colleagues found that a positive response on an IPV screener was predictive of future physical and non-physical violence in women who disclosed IPV compared to those who did not have a history of IPV. 
It is clear that progress has been made in screening ED patients for IPV. More educational programs for students and physicians are in place, more studies are being conducted, and screener refinement is occurring. However, there are definite gaps in research as well as future directions that should be taken. First and foremost, there is a need for long-term screening programs and evaluations. Most extant research is for a finite period of time, typically one to six months. It will be beneficial to have prospective studies whose duration is more than a year. This will give investigators a better picture of day-to-day life in a hospital ED. We also need to study the feasibility and efficacy of making IPV screening a part of every patient's medical history instead of just having a survey available for the purpose of a study. As Olson et al. , stated, sometimes policy change is more effective than attempting to change physician behavior.  A brief verbal screen of just one or two questions is time-efficient and may be enough to identify victims of IPV. , MacMillan, in her review of IPV screening approaches, found that a face-to-face approach to screening is the least-preferred by patients while written screeners, such as the Woman Abuse Screening Tool, provided the least missing data.  These findings need to be incorporated into screening programs that are implemented in healthcare settings.
Another option for screening that should be scientifically pursued is the possibility of using computer technology. Rhodes et al., conducted a study that encouraged patients to complete a computer-generated questionnaire about health risks and present the findings to their physician.  This method produced significantly higher rates of disclosure than historical controls. Houry and her research team have conducted numerous studies using similar technology in ED waiting rooms which have produced promising results in terms of patients' safety. One specific study has vast implications for screening in the ED. Houry and colleagues looked at a convenience sample of patients that came into the waiting room of an urban Level 1 trauma center. Out of the 2,134 participants that had been in a relationship in the past year, 548 (25.7%) screened positive for IPV. No safety issues, such as calling security or a partner's interference with the screening, occurred during the ED visit for any patient who disclosed IPV. Of the 281 IPV victims followed up one week later, none reported any injuries or increased IPV resulting from participating in the study. A full 95% of IPV victims stated they benefited from project participation. Finally, 35% reported they had contacted community resources during the three-month follow-up period.  In sum, this large prospective screening study demonstrated that participants are not at risk during the screening process nor afterwards for retaliatory violence. Of course, there are logistical and financial concerns when considering the possibility of utilizing computers to screen patients for IPV. Issues of space and funding would need to be decided before such a program could begin.
In addition, studies looking at patient outcomes after screening need to be conducted. There are no identifiable studies that evaluate the effectiveness of ED-based screening of patients for IPV in decreasing the morbidity or mortality of IPV long term.  Similarly, researchers need to look at the efficacy of this screening in improving health status and decreasing the incidence of violence. Houry and colleagues looked at three-month outcomes after IPV screening such as development of safety plans and contact with resources.  Although their findings were extremely promising, there was no control group and the follow-up period was short.
Relatedly, Anglin et al., recommend that we look into tying screening programs to advocacy interventions as IPV victims that receive such an intervention are more likely to use a shelter and access community counseling services. For instance, Harris et al., describe a program that involved crisis counselors coming to the ED to work with IPV victims to provide referrals, fill out protection orders, and begin the process of reporting to the police.  Krasnoff and Moscati conducted a three-tiered advocacy intervention that included universal screening, an on-site IPV advocacy intervention, and telephone-based counseling by an IPV case manager. Of the 528 women identified as IPV victims for this program, 84% agreed to speak to the advocate, and 258 (54% of those seen by the advocate) accepted case management follow-up. After the case management process, 127 women reported that they no longer believed they were at risk of violence from their abuser.  These types of programs would also alleviate the burden of physicians when trying to determine how to handle IPV victims.
There are several additional avenues that need to be explored within the issue of IPV screening. There is a need for research that assesses IPV in men and teens.  While it is true that adult women make up the majority of IPV victims in the ED, men have not routinely been screened for IPV victimization and we may be missing the opportunity for an intervention. Additionally, violence in teenage relationships is a problem that is receiving greater attention and one that deserves to be further studied.  Finally, research is needed to determine whether universal or targeted screening programs are more effective for identifying IPV as well to determine additional factors that increase a patient's willingness to disclose, for instance age, race, and gender of the person conducting the screening.
Screening for IPV in the ED is not a new topic; recommendations for doing so and studies looking at the results of these programs have been around for decades. However, many questions still remain unanswered. Which tool is best? Should all patients be screened or only those that present with a certain constellation of symptoms? Do the risks outweigh the benefits? However, screening is low-cost and patients will not disclose IPV unless directly asked. Intimate partner violence is a medical, social, and public health problem and without provider screening and identification of IPV victims, we will not be able to assist those victims most at risk of future injuries and sequelae from abuse.