Given the potential for ED-based interventions to impact IPV due to the notion of such medical visits providing a “teachable moment”, it is imperative to better understand correlates of IPV among a comprehensive portrait of ED patients to inform future intervention development efforts. Findings from this study show that about 9% of male and female patients presenting to the ED for care for any reason (medical or injured) were involved in IPV in the prior year. These rates are on the lower end of the range found in prior ED studies (5–29%) (Cunningham et al., 2003; Dearwater et al., 1998; Houry, et al., 2008
; MacMillan et al., 2006
; Walton et al., 2007
). Further, the rates found in this study are consistent with data from community samples (e.g., 7.8% – 21.5%) (Shafer & Caetano, 1998
; Staus, 1990
). Variations in rates may reflect measurement differences (for review see Lindhorst & Tajima, 2008
), sampling differences (women only as compared to women and men), or definition differences (among the total sample as compared to those with a current or recent partner).
Females were significantly more likely to report IPV (including victimization and aggression) than males. This finding may reflect underreporting among males who fear retribution for admission. It is important to note also that violence severity and subsequent injury may differ by gender and requires further investigation. Socio-demographic factors (e.g., insurance, income, employment, education) were related modestly to IPV in expected directions. Surprisingly, African-Americans were more likely to report IPV than other races. The relationship between ethnicity and IPV is unclear; as noted by Grossman and Lundy (2007), several studies have reported negligible race differences after controlling for socioeconomic status.
Markers of IPV were remarkably similar for victimization and aggression, perhaps reflecting the reciprocal nature of IPV. Although aggression may be exhibited by both partners, it is likely that the motivation behind the violence (e.g., control vs. self-defense) differs by gender and should be further studied. Interesting exceptions were that participants who reported psychoactive prescription drug use and recent substance use treatment were more likely to report victimization than aggression. It may be that victims of violence self-medicate with prescription drugs as a coping strategy; however, this supposition requires further validation. Finally, substance use treatment episodes may provide an opportunity to assess and intervene with violence issues.
The notion that substance use is a potent marker of IPV is not new. However, this study provides a broad examination of alcohol, illicit drugs, and psychoactive prescription drug use, as well as substance use patterns and diagnoses of substance use disorders in a large sample of inner-city ED patients. Any substance use, including tobacco, alcohol, marijuana, cocaine and psychoactive prescription drug use, was related to IPV. Further, participants who report any substance abuse/dependence diagnoses, regardless of specific drug, were more likely to report IPV. This study makes an important contribution to the literature by examining substance use groups, acknowledging differences in participants based on whether they use a single substance, or are poly-substance users, and the specific combinations of substances that they use. Consistent with prior research on violence (Chermack & Blow, 2002
; Fals-Stewart, 2003
), findings show that participants who reported alcohol and cocaine use were the most likely to report IPV. Overall, both acute intoxication and social/contextual considerations may explain the relationship substance use and IPV. For example, laboratory research demonstrates increased aggression following acute alcohol use in particular, but also with cocaine use (Chermack & Giancola, 1997). Prior research also shows that women who abuse substances are likely to have a substance abusing partner (Grisso et al., 1999
; Gomberg 1993
; Kyriacou et al., 1999
), that substance use among men is associated with increase risk for IPV (Chermack, Walton, Fuller, Blow, 2001
), and that women's substance use is viewed by partners as an excuse to justify violence (Leonard, 2001
Although IPV was not related to overall physical health in this relatively young ED sample, current ED visit reason was significantly related to IPV, with injured patients being more likely to report IPV than medical patients, even when controlling for other factors. Finding regarding the relationship between IPV and depression is consistent with prior research from substance abuse treatment and ED samples showing that psychiatric distress is associated with IPV (15 Houry et al., 2007
, Lipsky et al., 2005
; Walton et al., 2007
). Participants reporting depression were twice as likely to have experienced IPV in the past year, controlling for other factors. Inclusion of this depression questions in routine screenings would indicate patients in need of additional assessment for IPV. Although the causal direction of the relationship between IPV and depression or poor mental health functioning can not be determined in this study, theoretically, it is possible that distress is both a precipitant and a result of IPV (Briere & Jordan, 2004
Despite mandates from the Joint Commission on Accreditation of Healthcare Organizations (JACHO) recommending screening all women for IPV, there are limitations to this in practice including lack of questioning of medical patients, and one in five injured women are not assessed for IPV (Bonsal et al., 2007
). Although the development of brief single- or multi-item measures has reduced time/feasibility concerns, barriers to routine IPV screening in the ED include methodology (e.g., asking in the presence of the abusing partner), concerns by patients (e.g., fear from patients regarding police involvement and retaliation), and concerns by staff (e.g., feeling inadequate to address IPV) (for a review see Ernst & Weiss, 2002
). A recent study alleviates some of these concerns, as IPV did not increase following ED screening; further, a third of women sought community resources within three months post ED screening (Houry et al., 2008
). Data from this study suggest that referral approaches should include information on substance use and mental health functioning, given the correlation between these factors and IPV. Referral menu approaches could be standardized for both men and women, to allow for gender differences in interest in services. For example, men are clearly not appropriate for referrals to domestic violence shelters, but may be amenable to other substance use or mental health services. It may be particularly important to address concomitant substance use given evidence that reduced substance use following treatment is related to reduced aggression (O'Farrell et al., 2003; O'Farrell et al., 2004; Walton et al., 2002
). A recent meta-analysis of brief alcohol interventions in the ED concluded these approaches effectively reduce alcohol-related consequences (Harvard et al., 2008). Alternatively, motivational enhancement approaches (Miller and Rollnick, 2002; Dunn, Deroo, Rivara, 2001
) could be used to increase referral compliance, given data shows that such approaches increase engagement and retention in substance use treatment (De Leon, 1996
; Joe et al., 1998
). ED-based brief interventions approaches for alcohol and other drug use should consider assessing and intervening for IPV.
Limitations of this study require acknowledgement and include the fact that IPV was measured using very brief computerized screening questions; more lengthy approaches may find higher rates of IPV. Next, given our setting, an inner city ED, findings may not generalize to ED's located in suburban or rural settings. Findings require replication with patients presenting during midnight shifts. Further, although no data was collected from significant others to corroborate IPV, validity of self-report is increased in research where no consequences occur as a result of admission (see Darke, 1998). Findings from this study may not generalize to rural or suburban settings. Additional studies should be conducted in communities with greater concentrations of people of Hispanic/Latino ethnicity. Future longitudinal studies are needed that employ timeline calendar techniques (Chermack & Blow, 2002
; Fals-Stewart, 2003
) to elucidate the role of acute substance consumption and IPV. Finally, future ED-based studies should examine gender differences in IPV based on motivation (control, defensive, result of escalatory processes) to better inform screening and intervention efforts.
Despite these limitations, to our knowledge this is the first detailed examination of demographic, health, and substance use correlates of IPV among a comprehensive sample of ED patients (medical and injured). Clinically, results further justify routine screening of all ED patients for IPV, especially patients with poor mental health or substance use. The use of brief screens administered by computer increase the translation of protocols for detecting patients in need of a variety of healthcare services (e.g., violence prevention, mental health, and substance use treatment). The interrelationship among mental health, substance use, and violence problems needs to be considered when developing referral guidelines and potential intervention approaches for IPV. Future studies should develop and test the effectiveness of brief interventions and referrals for IPV and related problems among both males and females in the ED.