Epidemiologic Reviews Advance Access published January 5, 2016 Epidemiologic Reviews © The Author 2016. Published by Oxford University Press on behalf of the Johns Hopkins Bloomberg School of Public Health. All rights reserved. For permissions, please e-mail: [email protected].

DOI: 10.1093/epirev/mxv007

Risks and Targeted Interventions: Firearms in Intimate Partner Violence

April M. Zeoli*, Rebecca Malinski, and Brandon Turchan * Correspondence to Dr. April M. Zeoli, School of Criminal Justice, Michigan State University, Room 560, Baker Hall, 655 Auditorium Road, East Lansing, MI 48824 (e-mail: [email protected]).

The use of firearms in intimate partner violence (IPV) is widely recognized as an important public health threat. However, what we know about the risks of firearm access on IPV outcomes is limited. Therefore, we conducted a systematic review to determine the state of knowledge on 1) the risks of firearm access and use in IPV and 2) the effectiveness of interventions designed specifically to reduce firearm violence in intimate relationships. Only studies published in English in peer-reviewed journals from 1990 through 2014 were included. Results of the review suggest that, when violent intimates have access to firearms, IPV increases in severity and deadliness; however, increases in severity may not be due to firearm use. Additionally, statutes prohibiting persons under domestic violence restraining orders from accessing firearms are associated with reductions in intimate partner homicide, but certain provisions of these laws and their enforcement may impact their effectiveness. Future research should focus on elucidating the link between firearm access and increased IPV severity and on investigating whether and which specific provisions of domestic violence restraining order laws impact the laws’ effectiveness. Additionally, more evaluations of initiatives designed to improve the enforcement of domestic violence restraining order firearm prohibitions are needed. domestic violence; firearms; homicide; interventions; policy

Abbreviations: DVRO, domestic violence restraining order; IPH, intimate partner homicide; IPV, intimate partner violence.

against women involved firearms compared with 5% against men (10). Moreover, a study of domestic violence restraining order (DVRO) applications found that women were significantly more likely to report being harmed or threatened with firearms compared with men (11). Much of this research occurs within the United States. The geographical disparity in research on firearm access and use in IPV and firearm-related interventions for IPH is due, in part, to the high rate of firearm IPH in the United States. For example, firearms are used more frequently in intimate partner femicide in the United States than in many other nations, including South Africa (30%) (12), Sweden (20%) (13), and Portugal (45%) (14). This review serves multiple important functions. It will synthesize what is known from peer-reviewed literature about the risks of firearm access and use in IPV, as well as the impact and enforcement of interventions for firearm use in IPV. Research methods will be critiqued and existing gaps in the literature identified. Finally, we will suggest avenues for future research, policy, and practice.

INTRODUCTION

In 2013, 50% of the approximately 1,270 intimate partner homicides (IPHs) in the United States were committed with firearms (1), the lowest percentage since at least 1980 (2–6). Additionally, firearms were used in 3.4% (or 32,900 annually) of nonfatal intimate partner violence (IPV) victimizations from 2003 through 2012 (7). Although this is a critical area of study, knowledge about the risks of firearm access and use in IPV, as well as how best to intervene, is limited. There are, however, 2 main factors that frame the body of research on firearm use in IPV: sex and geography. Women comprise the majority of victims of IPH and IPV, and men the majority of perpetrators (1–9). The use of firearms in IPH and IPV also differs by sex. In 2013, 53% of female victims of IPH were killed with firearms (1). The next most commonly used weapons were knives at 19% (1). Weapon use in male IPH victimizations is not so skewed: 40% were killed with firearms and 38% with knives (1). Additionally, 11% of partner-perpetrated aggravated assaults 1

Downloaded from http://epirev.oxfordjournals.org/ at University of California, San Diego on January 29, 2016

Accepted for publication August 10, 2015.

2 Zeoli et al.

METHODS Search strategy

Inclusion criteria

We were interested in peer-reviewed research that used quantitative methods to examine 1) the risk or severity of IPV given firearm access or use and 2) the impact of interventions specifically designed to reduce firearm use in IPV on either violent outcomes or perpetrator access to firearms. A study must have examined at least one of the following victim-offender relationships to qualify for our review: current or former legal or common-law spouses and current or former dating partners, including homosexual relationships. Studies that aggregated intimate partner relationships with other relationships, such as nonintimate family, were excluded. Similarly, we excluded studies that measured firearm use in combination with other weapons (e.g., “knife or gun”) because of an inability to isolate firearms. Additionally, we excluded studies that solely examined specific subtypes of IPH, such as those that considered only homicidesuicide. During our search, article titles were first examined for relevance. If a title suggested that the study examined IPV outcomes, firearm access or use in IPV, or interventions for firearm violence, the abstract was read. If the abstract indicated that the study would meet our inclusion criteria, or if it was unclear on the basis of the information gleaned from the abstract that the study could be confidently excluded, the article was read. Each article was independently evaluated for inclusion by at least 2 authors. When there was disagreement, all 3 authors discussed the article until consensus was reached. Collectively, our database search yielded a total of 16,504 article titles; however, there was inevitably considerable overlap of results between search terms and across databases. We read 719 abstracts according to the perceived relevancy of the article’s title. Finally, 218 articles were read, and 19 were included in the review. References from all 19 articles were reviewed, and 15 cited works were selected for further examination, none of which met the inclusion criteria. In total, our

Data extraction

Two authors (A.M.Z. and B.T.) extracted pertinent data from the articles, including the intimate relationships considered, the composition of the study sample, how firearm access or use was operationalized, IPV outcomes measured, and details of interventions. We also gathered information on results of statistical tests of associations between firearms measures and interventions and IPV outcomes. RESULTS Risk studies

Table 1 provides descriptions of the 12 risk studies. Three studies examined whether firearm access was associated with the occurrence of nonfatal IPV victimization (15–17). In 5 studies, researchers examined people in relationships with existing IPV to probe whether perpetrators’ access to firearms was associated with the severity of nonfatal IPV (18–20), the use of firearm-related threats (21), or the risk of future IPV (22). Two case-control studies investigated whether perpetrator firearm access or use was associated with IPH victimization among partner-abused women (23, 24). Finally, 1 study tested whether state-level rates of firearm ownership were associated with IPH rates (25), and 1 longitudinal study followed women who purchased handguns over a 5-year period to determine if they were at an increased risk of homicide victimization (26). With exception, the majority of these studies were crosssectional, and therefore no information was gained on the temporal order between firearm access and abuse. It must not be assumed that firearm access was always concurrent with abuse. The majority of risk studies also relied on self-reported data. Importantly, this includes whether women’s intimate partners had access to firearms, which some women may not know or may be mistaken about. Additionally, women whose partners have not made firearm threats may be less likely to know whether their partner has firearm access, which could lead to an underestimation of firearm access among this group. Despite this limitation, self-report methods often obtain the most complete data on abuse-related factors, as cases of IPV may go undetected by official agencies. Risk of nonfatal IPV. There is no compelling evidence that firearm access alone increases the risk of nonfatal IPV victimization. Two studies surveyed general samples of women regarding whether firearms were kept in their homes and whether they experienced IPV victimization (15, 17). Because neither of these studies measured cohabitation or directly measured intimate partners’ access, it is unknown to what extent firearms in women’s homes were accessible to their partners, or if intimate partners possessed firearms not kept in the respondents’ homes. Results suggest that women who keep firearms in their homes are not at increased risk of nonfatal IPV victimization, but no inferences about the association between intimate partner firearm access and nonfatal IPV victimization can be made.

Downloaded from http://epirev.oxfordjournals.org/ at University of California, San Diego on January 29, 2016

We used a 2-phased approach to locate studies that quantitatively examined either the risks associated with firearm access and use in IPV or the impacts of interventions specific to firearm use in IPV. First, we completed a systematic literature search to gather original peer-reviewed research published in English between January 1990 and December 2014 that met our inclusion criteria. Two authors (R.M. and B.T.) conducted independent searches using the databases Web of Science, PubMed, ProQuest Criminal Justice, and ProQuest National Criminal Justice Reference Service Abstracts Database. Our initial search terms of “domestic violence,” “intimate partner violence,” “intimate partner homicide,” “firearm,” and “gun” were used in combination. Keywords from relevant articles found in the initial search were used in subsequent searches, including “policy,” “restraining order,” “protective order,” “risk factors,” and “batterer.” In the second phase of our literature search, we evaluated the cited works of eligible studies from the first phase for additional relevant studies.

search yielded 12 studies on the risk or severity of IPV as a result of firearm access or use and 7 studies on the impact of interventions designed to reduce firearm use in IPV.

Table 1. Characteristics of Research That Assessed the Risk of Intimate Partner Violence or Injury Outcomes Given Access to Firearms or Firearm Use, 1990–2014 First Author, Year (Reference No.)

Study Population Firearm Exposures Date

No.

Data Source

Campbell, 2003 (23)

1994–2000

Coker, 2000 (15)

February 1,401 1997– December 1998

Women recruited at 2 family practice clinics.

Folkes, 2013 (18)

Prior to 1,421 December 1997

Men who Police reports. committed Criminal physical records. assault or issued credible death threat against IP that resulted in police reports.

Self-reports.

Women who Self-reports. had been physically or sexually assaulted, stalked, or threatened with physical violence within 2 years of study.

Modified Conflict Tactics Scale, Danger Assessment

Findings

Location

Current or former “romantically or sexually” involved heterosexual relationship within the past 2 years

11 US cities

Victim had sole Victimization (homicide, access to firearm in home. nonfatal physical Perpetrator had a abuse) firearm in home or owning a firearm. Firearm use in IPH or worst incident

65% of perpetrators of IPH and 24% of perpetrators of nonfatal IPV had access to firearms. Perpetrator access to firearm increased the risk of homicide (aOR = 5.38; P < 0.001). Abuser use of a gun in worst incident was associated with an increased risk of IPH (aOR = 41.38; P < 0.001).

Columbia, South Carolina

Presence of guns in Victimization the participant’s (physical and household sexual IPV, physical but no sexual IPV, battered or emotional abuse without physical or sexual violence)

26.0% of the sample reported guns being present in their current home. Having a gun in the home was not significantly related to any of the types of IPV examined.

Index of Spouse Current or former Abuse, intimate (sexual) Women’s heterosexual Experience relationships that with lasted at least Battering 3 months Scale, shortened Abuse Assessment Screen

Modified Conflict Tactics Scale, Canadian Violence Against Women Survey, Danger Assessment

Type of Violence

Relationships Studied

Current or former Ontario, married Canada heterosexual couples. Current or former heterosexual cohabiting partners

Firearm access Perpetration 8% of the sample owned a (owning a firearm, (nonfatal physical firearm; 21 men made a license or abuse or death threat involving a firearm. registration for threat, Firearm access associated firearm, history of severity of assault with severity of policeweapons-related measured reported assault (r = 0.15; charges, using or according to the P < 0.001), but not due to threatening victim Revised Conflict firearm use in the assault with firearm). Tactics Scale) (n = 3). Firearm use or Firearm access was also threat with firearm associated with women’s fear of future assaults (OR = 2.15; P < 0.01), but was not associated with recidivism as measured by later police reports or criminal records of physical assault.

Current or former “romantically or sexually” involved heterosexual relationship within the past 2 years

Perpetrator having a Victimization firearm in home or (controlling owning a firearm behaviors)

11 US cities

24% of sample reported that partner had access to a firearm. 29% of those who used at least 1 controlling behavior and 15% of those who used none had access to firearms (OR = 2.24; P < 0.04). Firearm access was not different for those who used at least 3 compared with 0 controlling behaviors.

Table continues

3

Downloaded from http://epirev.oxfordjournals.org/ at University of California, San Diego on January 29, 2016

331

Cases: women Proxy reports with histories for fatal of IP abuse victims. killed by Self-reports perpetrator. for nonfatal Controls: victims. women abused by intimate partners within 2 years of study.

Instruments

Firearms in Intimate Partner Violence

1999–2000 Frye, 2006 (19)a

220 cases; 343 controls

Population

First Author, Year (Reference No.)

Study Population Firearm Exposures Date

No.

Population

Data Source

Instruments

Type of Violence

Findings

Relationships Studied

Location

Current or former “romantically or sexually” involved heterosexual relationship within the past 2 years

11 US cities

Perpetrator having a Victimization firearm in home or (homicide, owning a firearm nonfatal physical abuse)

United States

Use of firearm or Victimization 25% of partners owned a threatened use of (nonfatal actual gun. Threat with or use of firearm. or threatened a gun was predictive of Firearm ownership physical or sexual threatened or actual violence) physical or sexual violence at the 1-month follow-up (RRR = 1.93, 95% CI: 0.79, 4.75) (not significantly predictive, but predictive with a point estimate of over 1.33, the threshold for prediction in the study).

Glass, 1994–2000 2008 (24)a

23 cases; 53 controls

Cases: women Proxy reports Modified aged 18–20 Conflict for fatal years victims. Tactics murdered by Self-reports for Scale, partner. nonfatal Danger Controls: victims. Assessment women aged 18–20 years abused by partners within 2 years of study. Additional comparison group: women aged ≥21 years murdered by partner.

Glass, 2008 (22)

84

Female victims Self-reports. of physical or sexual IPV.

Danger Current or former Assessment, female same-sex risk factor relationships items developed by authors

McFarlane, 1998 (20)

199

Ethnically Self-reports. stratified cohort of pregnant women at public health clinics who were abused by their partners within the past 12 months.

Abuse Screen, Current or former United States Index of married Spouse heterosexual Abuse, couples. Danger Current or former Assessment, heterosexual dating Severity of partners Violence Against Women Scales, Relationship Inventory

Does [perpetrator] own or have access to a gun? Gun location

56% of perpetrators of IPH and 4% of perpetrators of nonfatal IPV had access to firearms (χ2 = 25.6; P < 0.01). No significant difference between perpetrator access to firearms between IPH victims aged 18–20 years and ≥21 years.

Victimization 41% of abused women (nonfatal physical reported their partner had abuse, severity of access to a firearm. Of abuse) these, 17% reported firearm was “on him,” 39% “in the house,” 16% “with a friend or relative,” 6% “in his vehicle,” 2% “in storage,” and 20% “unknown.” Partner access to firearms was significantly associated with more severe abuse on multiple scales. Firearms being kept “on” perpetrators was associated with increasing severity of IPV based on higher mean scores on the ISA (45.7 vs. 33.2; P = 0.037) and DA (7.9 vs. 6.1; P = 0.035) scales, but this did not hold once α levels were adjusted.

Table continues

4 Zeoli et al.

Table 1. Continued

Downloaded from http://epirev.oxfordjournals.org/ at University of California, San Diego on January 29, 2016

Table 1. Continued First Author, Year (Reference No.)

Study Population Firearm Exposures Date

Roberts, 2009 (25)

1985–2004

Rothman, 2005 (21)

1999–2003

Shuman, 2008 (16)

No.

Population

Populations of American counties.

8,529

Data Source

Instruments

FBI-SHR. Police Foundation Report.

Men enrolled in Self-reports. Massachusetts certified batterer intervention programs.

Relationships Studied

Not specified. SHR includes relationships of current or former heterosexual married (or common law) couples; current dating partners; same-sex relationships Batterer intervention program initial appointment interview

Type of Violence

Findings

Location

United States

Rates of firearm ownership by state in 1996

Attendance at batterer Massachusetts Perpetrator firearm intervention ownership in the program for IPV last 3 years (defined by circumstances)

Victimization (intimate partner homicide, intimate partner homicide with firearms)

57% of IPHs involved firearms. An increase in firearm ownership statewide of 1% significantly increased county-level IPHs (IRR = 10.78; P < 0.001) and firearm IPHs (IRR = 6.26; P < 0.001).

Perpetration 7% of sample owned a (firearm-involved firearm. threats (e.g., 13% of firearm owners and cleaned, loaded, 5% of non-firearm owners or held gun during reported using knives or argument with firearms to threaten their partner) partner. 12% of firearm owners made at least 1 firearm-related threat (46% of these made 2 or more threats). Firearm ownership in the last 3 years was significantly associated with threatening partners with guns (aOR = 7.8).

Items Current or former developed by married authors heterosexual couples. Current or former heterosexual dating partner

Rural Georgia

Whether partner owned a gun

Victimization 41% of IPV victims and 49% (nonfatal physical of nonabused women abuse) reported that their partners owned firearms (not significant in bivariate analysis). The gun ownership variable was excluded from multivariate analysis because of its strong correlation with abuse, causing “mathematical estimation problems” in the model.

BRFSS

Current or former spouse. Current or former dating partner. Date (includes same-sex relationships)

Montana, New York, and Ohio

Firearm in the respondent’s home

Victimization 23% of respondents in (nonfatal physical Montana reported having abuse within firearms in the home; 1 year of study) 15.4% in New York, and 8.5% in Ohio. There was no measurable relationship between firearms in the home and the occurrence of IPV.

Intimate partner (unspecified)

California

Handgun purchases Victimization Of 11 women killed, 6 were (homicide, killed by intimate partners. including intimate 5 of those homicides were partner homicide) committed with firearms. Handgun purchase was associated with increased risk of intimate partner homicide (SMR = 2.30) and overall homicide (SMR = 1.45).

Vest, 2002 (17)

Montana: 1998– 1999; New York: 1994; Ohio: 1996– 1997

10,461 Women who (weighted filled out the sample BRFSS in size) Montana, New York, and Ohio.

Wintemute, 2003 (26)

1991–1996

28,181

Self-reports.

Women who California purchased Department handguns in of Justice the first year reports of of the study handgun in California. purchases. Coroner’s reports. FBI- SHR.

5

Abbreviations: aOR, adjusted odds ratio; BRFSS, Behavioral Risk Factor Surveillance System; CI, confidence interval; DA, Danger Assessment; FBI, Federal Bureau of Investigation; IP, intimate partner; IPH, intimate partner homicide; IPV, intimate partner violence; IRR, incidence rate ratio; ISA, Index of Spouse Abuse; OR, odds ratio; RRR, relative risk ratio; SHR, Supplementary Homicide Reports; SMR, standardized mortality ratio. a Data are from the 11-city study conducted by Campbell et al. (23).

Firearms in Intimate Partner Violence

188 cases;. Cases: women Self-reports. recruited 360 controls from IPV shelters. Controls: nonabused women recruited at health clinics involved in an intimate relationship within the year prior to recruitment.

Downloaded from http://epirev.oxfordjournals.org/ at University of California, San Diego on January 29, 2016

6 Zeoli et al.

likely to make firearm-related threats against intimate partners than did those who did not own firearms. This study relied on self-reports from IPV perpetrators for information on both firearm ownership and threats. Although one might expect perpetrators to underreport on both of these variables, such underreporting would likely bias the estimates toward null findings; however, that did not appear to happen here. It is perhaps unsurprising that those who owned firearms were more likely to make firearm threats than those who did not own them, particularly given that 2 of the 4 indicators of firearm-related threats used in the study required the presence of a firearm. The finding that approximately 1% of the offenders who did not own firearms displayed them, however, has implications on our understanding of the problem and possible solutions. This finding may be related to the mismatched recall periods between firearm ownership (past 3 years) and firearm threats (lifetime): Offenders may have owned firearms at the time of the threat. Two additional possibilities are that they displayed firearms owned by their victims or owned unregistered guns that they were less willing to report (21). Research by Glass et al. (22) is notable for 2 reasons: 1) They tested whether firearm threats or use at a baseline interview was associated with threatened or actual use of violence at a 1-month follow-up interview; and 2) participants were women in abusive same-sex relationships. Given the recruitment strategy of advertising the study in key locations, such as newspapers with a lesbian, gay, bisexual, and transgender focus and domestic violence service agencies, it is difficult to assess the generalizability of the study results to the larger population of women in abusive same-sex relationships. Use of or threat with firearms by current or former intimate partners within the previous year was predictive of future actual or threatened IPV during the follow-up period. The goal of this study was to revise the Danger Assessment instrument for same-sex populations. The item measuring partners’ firearm threats was retained in the revised instrument; however, partners’ firearm ownership was excluded because it was not predictive of future IPV. Risk of fatal IPV. Four studies found that firearm exposure was associated with the risk of IPH. Campbell et al. (23) conducted a case-control study of female victims of IPH and nonfatal IPV, assessing what factors were associated with fatal violence. Proxy informants provided data in the place of IPH victims, while victims of nonfatal IPV completed selfreport interviews. Firearm access by victim and perpetrator was measured separately. The victim’s sole access was defined as having a gun in her home and living apart from her abuser, while abuser access was defined as a perpetrator’s owning a firearm or having one in his home. Cases and controls were equally likely to have sole access to firearms, suggesting that the victim’s sole access is not protective against IPH. In contrast, perpetrator firearm access was associated with a 5-times-greater risk of IPH, much of this risk due to the use of firearms in the worst abusive incident. Glass et al. (24) used a subset of the data from Campbell et al.’s case-control study to explore the link between perpetrator firearm access and risk of IPH in young adult populations. Estimates suggested that women aged 18–20 years were at greater risk of IPH when their partners had firearm access. For these studies by Campbell et al. and Glass et al.,

Downloaded from http://epirev.oxfordjournals.org/ at University of California, San Diego on January 29, 2016

In contrast, in a case-control study comparing abused women recruited from victim service agencies and nonabused women recruited from medical clinics, participants were asked whether their intimate partners owned firearms (16). Although a greater percentage of controls than cases reported that their intimate partners owned firearms, a bivariate analysis found no significant difference in intimate partner firearm ownership between cases and controls. This variable was not tested in a multivariate model because of collinearity with a variable representing whether a partner carried a weapon, which significantly increased the risk of IPV (E. Waltermaurer, The Benjamin Center for Public Policy Initiatives at SUNY New Paltz, personal communication, 2015). No conclusions can be made, however, as to whether the weapon carried was a firearm, as the carrying of any weapon, including knives, would have prompted a positive response from participants. Nonfatal violence in abusive relationships. Among those in abusive relationships, the evidence suggests that perpetrators’ access to firearms is associated with the severity of nonfatal IPV committed. In a sample of partner-victimized pregnant women, an intimate partner’s firearm access was significantly correlated with his inflicting more severe IPV (20). IPV severity was measured with multiple self-report scales, one of which, the Danger Assessment, included an item on the presence of a firearm in the home (27). Therefore, a perpetrator with firearm access has an increased Danger Assessment score, indicating greater abuse severity, simply because of his firearm access. Although this is a reason to view study results with caution, the increase in IPV severity associated with firearm access held when measured by the Index of Spouse Abuse (28) and the Severity of Violence Against Women Scales (29), neither of which measures firearm ownership. A second study based on self-reported data from partnerabused women also found an association between firearm access and psychological abuse (19). Perpetrator firearm access was positively related to the use of 1 or more psychologically abusive controlling behaviors compared with no controlling behaviors. Interestingly, there was no significant difference in firearm access between those who used 3 or more controlling behaviors and those who used none. The severity of police-reported physical assaults was also found to be greater when male IPV perpetrators involved in the Canadian criminal justice system had access to firearms (18). Firearm access was determined by any mention in police reports or criminal records of a perpetrator’s firearm access, including having a license or registration for a firearm or threatening a victim with a firearm. The authors acknowledged that this measure likely undercounted firearm access, as police may not always record such evidence in their reports. Still, access was positively correlated with assault severity, but not because of firearm use, which was rare. Instead, this relationship was due to the use of other weapons in the assaults. Additionally, there was no difference in the likelihood of committing future police-reported reassaults between those with firearm access and those without. However, because not all assaults are reported to the police (7), it is unknown whether firearm access was associated with future violence. Rothman et al. (21) found that IPV perpetrators enrolled in batterer treatment programs who owned firearms were more

Firearms in Intimate Partner Violence

Intervention studies

Each intervention study was conducted in the United States (Table 2). Four studies focused on the impact, and 3 studies focused on the implementation of state statutes specifically designed to prohibit access to and, by extension, use of firearms by IPV perpetrators. The statute receiving the most research attention prohibits those currently under certain DVROs from purchasing or possessing firearms. A second statute applies similar prohibitions to individuals convicted of domestic violence misdemeanor offenses. Although these statutes are found at state and federal levels, state-level statutes tend to prohibit firearm access under more circumstances (31). The third type of IPV-focused firearms statute under study is found only at the state level and authorizes law enforcement officers to confiscate firearms from the scene of IPV under specific circumstances. DVRO firearms restriction statutes. Two studies suggest that DVRO firearm restriction statutes are associated with significant decreases in rates of IPH (31, 32), while a third study found no association (33). Both Vigdor and Mercy (31) and Zeoli and Webster (32) conducted interrupted time-series analyses to examine the relationship between the statutes and IPH. Vigdor and Mercy’s analysis examined 46 states

from 1982 through 2002, during which time 24 states enacted DVRO firearm prohibition statutes (31). Zeoli and Webster (32) examined 46 of the largest US cities, representing 27 states, from 1979 through 2003, during which time 15 of the states enacted the statutes. Homicide data were obtained from the Federal Bureau of Investigation’s Supplementary Homicide Reports for both studies. DVRO firearm prohibitions were associated with an 8%–19% decrease in total IPH and a 9%–25% reduction in firearm IPH. Importantly, in further analyses, Vigdor and Mercy (31) found that states that prohibited firearm purchase experienced a 10% decrease in IPH, but states that prohibited only possession experienced no measurable reduction. Additionally, they tested whether the DVRO laws’ estimated association with IPH differed in states with a “high” or “low” ability to detect IPV-related firearm restrictions in the background check system that is queried by licensed firearms dealers to identify restricted individuals before sale (31). The DVRO firearm prohibition was only associated with a reduction in IPH in states with a high ability to detect disqualifications in the background check system. Bridges et al. (33), however, found no association of DVRO prohibition statutes with IPH. Their analysis, however, has major methodological limitations, and results must be viewed with caution. First, they conducted a cross-sectional examination of the association of DVRO prohibition statutes and 6 other IPV-related statutes with IPH with 47 states using an averaged IPH rate for the years 1995 through 1999. Therefore, the analysis has few observations in which to find significant variation by statute. Second, according to the data of Bridges et al., 13 states enacted DVRO firearm prohibition statutes from 1996 through 1999, but how the presence or absence of those statutes is coded given the cross-sectional time period of 1995 through 1999 is not addressed in the paper. Additionally, it must be noted that the years when states enacted the IPV-specific firearm laws used by these authors differ from those presented by Vigdor and Mercy (31). The association between DVRO firearm prohibition statutes and nonfatal IPV victimization is informed by 1 study. Dugan (34) measured IPV victimization using data drawn from the 1992 through 1998 National Crime Victimization Surveys, which survey a nationally representative sample of households. Using an individual-level analysis, Dugan conducted a logistic regression testing whether being located in a state with the DVRO firearm prohibition statute was associated with whether a household reported IPV. The results suggested that the odds of nonfatal IPV in dating relationships were 14% lower in states with the statute, but there was no measurable association on violence among current or former spouses. These 4 studies on the impacts of policy on fatal and nonfatal IPV share 2 important limitations. First, although each of the studies controls for additional selected policies and factors, none of them is exhaustive in controlling for policies and/or factors that may reduce fatal and nonfatal IPV. It is possible that there are omitted factors that produce the resulting statistical associations. Second, none of the studies included measures of policy implementation. It is therefore unknown whether the policies under study were fully or even partially implemented and how many people they affected. If they were not implemented or affected only a

Downloaded from http://epirev.oxfordjournals.org/ at University of California, San Diego on January 29, 2016

the use of proxy informants may have produced a different quality of data than the control group self-reported data. It is possible that proxies were less aware of offender firearm access, which could lead to an underestimation of its association with IPH, or that proxies overestimated firearm access given the seriousness of the abuse, which would suggest an overestimation of the association. An ecological time-series analysis found that state-level firearm ownership rates positively correlated with both countylevel IPH rates and firearm IPH rates (25); however, caution should be taken when interpreting these findings because of the study’s main limitation. Roberts (25) reports that state-level data on firearm ownership were from only 1 year, the midpoint of the 20-year study period, making it unlikely that changes in ownership rates over time were accurately represented. Additionally, the reported source of firearm ownership data (30) does not actually contain the specified data; thus, we were not able to assess the potential limitations of the data source. Wintemute et al. (26) took a novel approach to studying female firearm access in a longitudinal study that tested whether women who purchased handguns in California in 1991 experienced an increased risk of homicide in the ensuing 5 years. Following their purchases, women were at a higher risk for IPH and homicide victimization overall, but not for homicide victimization by a nonintimate partner. However, there are limitations to the conclusions that can be drawn from this study. First, handgun purchasers did not necessarily keep their firearms for the duration of the study period. Some may have sold their firearms after 1 year while others may have maintained ownership all 5 years; thus, the level of risk associated with owning a handgun was not evenly distributed across years or across the study population. Additionally, inferences made regarding the increased risk of IPH must be careful to consider that the perpetrators’ access to firearms, either their intimate partners’ handguns or their own, was not assessed.

7

8 Zeoli et al.

Table 2. Characteristics of Research That Examined the Effectiveness of Interventions Targeting Firearm Use in Intimate Partner Violence, 1990–2014 First Author, Year (Reference No.)

Intervention Description

Date

Location

Population

Outcome Measures and Data Source

Key Findings

Author Recommendations

Additional research is IPH rates for those aged DVRO firearm prohibition needed on DVRO statutes were inversely ≥15 years. firearm prohibition laws related to family homicide Family homicide rates for to determine 1) whether rates in point biserial all ages. they impact nonfatal IPV correlations (r = −0.32; Average of IPH and family and 2) whether giving P = 0.014) and the homicide rates. judges discretion versus average of IPH and family Data source: SHR. mandating judges to homicide (r = −0.2; IP relationships: apply firearm P = 0.025), but not related current or former prohibitions to DVRO to IPH rates. heterosexual married respondents affects In multiple regressions (or common law) outcomes. testing all 7 DV statutes, couples; current dating no statute was partners; same-sex significantly associated relationships. with any type of homicide.

Bridges, 2008 (33)

1995–1999 State DVRO firearm prohibition statutes. State DVM firearm prohibition statutes. Other interventions controlled for 5 additional statutes related to conditions under which DV is a felony offense and sentencing for DV convictions.

47 US states

Populations of states

Dugan, 2003 (34)

January 1992– State DVRO firearm June 1998 prohibition statutes. Other interventions controlled for: 5 additional state DV statutes related to DVROs.

United States

States should adopt laws DVRO firearm prohibitions Three groupings of All respondents that protect victims, but significantly related to household (aged ≥12 years) care must be taken by family (OR = 0.87; victimizations with the to the NCVS. local officials to properly P < 0.05) and boy/ following relationships: All violent incidents implement and enforce girlfriend violence 1) nonintimate family; reported to these laws. (OR = 0.86; P < 0.01), but 2) current or former NCVS not significantly related to spouses; and 3) spousal violence. current or former boy/ DVRO firearm prohibition girlfriend. was unrelated to whether Two other dependent police were informed or variables: police arrest was made for all informed and arrest for violent incidents. all violent incidents. Data source: NCVS. IP relationships: current or former married couples and current or former dating partners Table continues

Downloaded from http://epirev.oxfordjournals.org/ at University of California, San Diego on January 29, 2016

Table 2. Continued First Author, Year (Reference No.)

Intervention Description

Date

Location

1982–2002 State DVRO firearm prohibition statutes. State DVM firearm prohibition statutes. State DV firearm confiscation statutes. Other interventions controlled for several other firearm laws that may limit one’s ability to obtain a firearm without a background check and the speed and frequency with which one can purchase firearms.

46 US states

Vittes, 2008 (35)

California The purchase restriction of May 2003– California’s DVRO firearm November prohibition statute. 2005 (DVROs issued). 1998–2005 (handgun purchase applications)

Populations of states

Outcome Measures and Data Source

Key Findings

Author Recommendations

Recommend passage and For homicide victims of at DVRO firearm prohibition greater implementation laws were associated with least 10 years, total of DVRO firearm a significant decrease in IPHs and IPHs with prohibition laws, as well all IPHs (IRR = 0.92; female victims; total as quicker entry of P < 0.05), all IPHs with IPHs with firearms and firearms disqualifications female victims IPHs involving firearms into background check (IRR = 0.92; P < 0.01), all with female victims. systems. firearm IPHs (IRR = 0.91; Data source: SHR. Recommend registries of P < 0.01), and firearm IP relationships: current or firearm owners to assist IPHs with female victims former heterosexual police officers in (IRR = 0.9; P < 0.01). married (or common enforcement of DVRO States that have purchasing law) couples; current firearm possession restrictions in their DVRO dating partners; restrictions. laws had significant same-sex relationships reductions of 10%–13% in DVM firearm prohibitions and DV firearm all 4 IPH types tested confiscations laws may (P < 0.05), whereas states have benefitted with possession only individuals and families, DVRO laws did not have but not had a reductions. measurable impact at States in which IPV-related the state level. purchase disqualifications are entered into a computerized background check system experienced decreases of 9%–12% in all 4 IPH types tested (P < 0.05). DV misdemeanor firearm prohibitions and police confiscation at scene laws were not significantly associated with IPH or firearm IPH.

California residents Applications and Lack of entry of DVROs into Persons under DVROs approvals for handgun the background check applied to purchase fewer purchases for 4 system likely allowed handguns than persons categories of disqualified persons to before DVROs were applicants: DVRO obtain handguns. issued or after DVROs respondents in the time Legislation passed in expired. before DVROs were California that makes the Roughly 40% of applications put into place, while the court responsible for for handgun purchases of DVROs were active, entry of disqualification those under DVROs were and after the DVROs into the background approved after a expired; and California check system within background check. residents in general. 1 day may mitigate the Data sources: California’s problem. DVRO System and Dealer’s Record of Sale log. IP relationships: not specified in the text

Firearms in Intimate Partner Violence

Vigdor, 2006 (31)

Population

Table continues

9

Downloaded from http://epirev.oxfordjournals.org/ at University of California, San Diego on January 29, 2016

First Author, Year (Reference No.)

Intervention Description

Date

2002–2003

Location

New York City. Los Angeles

Population

Outcome Measures and Data Source

Key Findings

Author Recommendations

Whether judges ordered 82 (15%) of the 542 542 women who DVRO firearm prohibition interviewees who had abusers to surrender obtained DVROs statutes must be better DVROs reported their firearms. knew whether implemented and abusers owning a firearm. their partners had Whether firearms were enforced. Judges ordered removal Instead of providing judges removed from abusers. firearms and in 26% of cases in which Data source: DVRO completed a with discretion, judges abusers had firearms; petitioner interviews. 9-month (on should be required to however, they failed to IP relationships: not average) order firearm surrender order removal in 52% of specified in the text follow-up for those under DVROs. the 37 cases in which interview Victims and advocates women requested should “hold law removal. enforcement 12% of victims reported accountable if they do abusers forfeiting firearms not act to ensure that or having them seized. defendants have been Abusers’ surrender of disarmed” (p. 97). firearms was associated with victim reports of the judge ordering firearm removal (χ2 = 6.71; P = 0.035).

Webster, 2010 (36)

The possession restriction of DVRO firearm prohibition statutes in California and New York.

Wintemute, 2014 (37)

2007–2010 Initiatives to confiscate firearms from those prohibited from possession by DVROs in 2 counties. Initiative used California Automated Firearms System and other California Department of Justice databases, as well as DVRO petitioner interviews, and the DVRO application form to determine whether the DVRO respondent possessed firearms.

In San Mateo County, 23% Recovery of firearms is Whether firearms were San Mateo Restraining order possible and facilitated (n = 119) of firearms were confiscated from those respondents and Butte by databases that recovered from identified that evidence believed to counties, identify DVRO DVRO respondents. suggested possessed possess firearms California respondents who have In Butte County, 51% firearms. firearms. Search (n = 45) of firearms were Data source: Law warrants may facilitate recovered from identified enforcement records. recovery from those who DVRO respondents. IP relationships: not deny possession. specified in the text

Zeoli, 2010 (32)

1979–2003 State DVRO firearm prohibition statutes. State DVM firearm prohibition statutes. State DV firearm confiscation statutes. Other interventions controlled for police staffing levels and warrantless and mandatory arrest statutes for DVRO violations.

46 large US Populations of cities cities

DVRO firearm prohibition For victims aged at least DVRO firearm prohibition statutes may save lives. laws were associated with 15 years: total IPHs Future research should decrease in IPH and IPHs by firearms. investigate what factors (IRR = 0.81; P = 0.044) Data source: SHR. or processes may impact and firearm IPH IP relationships: current or the effectiveness of this (IRR = 0.75; P = 0.02). former heterosexual policy in communities, so DV misdemeanor firearm married (or common that these processes prohibitions and police law) couples; current may be replicated in confiscation at scene laws dating partners; other communities. were not significantly same-sex relationships associated with IPH or firearm IPH.

Abbreviations: DV, domestic violence; DVM, domestic violence misdemeanor; DVRO, domestic violence restraining order; IP, intimate partner; IPH, intimate partner homicide; IPV, intimate partner violence; IRR, incidence rate ratio; NCVS, National Crime Victimization Survey; OR, odds ratio; SHR, Supplementary Homicide Reports.

10 Zeoli et al.

Table 2. Continued

Downloaded from http://epirev.oxfordjournals.org/ at University of California, San Diego on January 29, 2016

Firearms in Intimate Partner Violence 11

Domestic violence misdemeanor firearms restrictions statutes. Although domestic violence misdemeanor firearm re-

striction statutes operate on the same principle as the DVRO firearm restriction statutes, that of removing access to firearms for IPV perpetrators, researchers have found no association of

these laws with IPH (31–33). Although these laws simply may not affect IPH, there may be alternative explanations as to why no association was found. For example, it may be because many states do not have specific domestic violence misdemeanor crimes, and people convicted of misdemeanor violence against intimate partners are not identified as having committed a disqualifying crime (31). In comparison, DVROs that have firearm prohibitions are more easily identified and may be more likely to be entered into the background check system than disqualifying domestic violence misdemeanor convictions. Firearm confiscation at the scene of IPV. Two studies examined whether statutes allowing the confiscation of firearms from the scene of IPV was associated with IPH (31, 32), neither of which found an association. It is possible that no significant association was found because of the variation in laws between states. For example, some states specify that only firearms used in the IPV incident may be confiscated, even if other firearms are present at the scene (38). One might not expect a policy that leaves all but 1 firearm in the home to prevent homicide. DISCUSSION

Results from this review suggest that female victims of IPV whose partners have access to firearms may be in lethal danger, but there is evidence to suggest that policy interventions in the United States reduce this risk. The results of this review also suggest that when violent intimate partners have access to firearms, IPV increases in severity. Although the increase in fatal violence is due to firearm use, the link between a violent partner’s firearm access and abuse severity is less clear. What is clear is that, because firearms are used in IPV and are associated with IPH, effective interventions are needed. Research on interventions has focused on legislation that prohibits abusers from accessing firearms and suggests that statutes that apply firearm restrictions to DVRO respondents reduce rates of IPH. Compellingly, there are ways to improve the enforcement of these statutes to potentially protect more victims. Implications of risk studies

Three main observations from our review of the literature on risks of firearm access are worth noting for their implications for policy and research. First, results suggest that IPV perpetrators’ access to firearms is associated with increased IPV severity (18–24). However, given that researchers did not establish the temporal order between firearm access and abuse and the findings that increased severity was not always due to firearm use (18, 19), the reasons behind this relationship remain unclear. Among IPV perpetrators, the risks associated with firearm access may be related to offender subtype (39). Those subtypes characterized by more severe violence and violence against nonfamily members may be more likely to possess and use firearms in IPV. The included research suggests that offenders who possess firearms are more likely to use controlling behaviors against partners (19) and to perpetrate more severe abuse (18, 20). There is also some support in the literature outside this review that batterers who have characteristics

Downloaded from http://epirev.oxfordjournals.org/ at University of California, San Diego on January 29, 2016

small number of people, the possibility that the association between the policies and IPH and IPV is due to confounding becomes greater. DVRO firearm prohibition statute implementation. Three studies evaluated implementation of these statutes (35–37). The research indicates that, while DVRO firearm restriction statutes are implemented with some success, there are opportunities for improvement. Vittes and Sorenson (35) assessed the implementation of the firearm purchase prohibition by linking data from California’s Domestic Violence Restraining Order System with California’s Dealer’s Record of Sale for handguns to determine the rates at which those under DVROs, most of whom would be prohibited from firearm purchase under California law, apply and are approved for handgun purchase. Although the denial rate for those currently under DVROs was higher than the denial rate for Californians in general, “nearly half of the purchase applications they filed were approved after a background check” (35, p. 9). The generalizability of these findings to other states may be limited. California requires background checks for private sales of handguns, which many states do not. The rate at which those under disqualifying DVROs are able to purchase firearms in states without background checks for private sales is presumably higher than what this study found. To examine the courts’ enforcement of DVRO firearm possession restrictions, Webster et al. (36) conducted interviews with 542 DVRO petitioners who knew whether their abusers owned firearms. Women reported that judges ordered firearms surrendered in approximately half of the cases in which they requested it, and that men were dispossessed of firearms in 24% of the cases in which surrender was ordered. As the authors point out, the use of DVRO petitioners as informants may limit the accuracy of information on both firearm restrictions and surrender, of which women may not be fully aware. Finally, Wintemute et al. (37) analyzed initiatives in 2 California counties to facilitate law enforcement officers’ removal of firearms from those prohibited from possession by DVROs. To identify armed DVRO respondents, detectives 1) examined DVRO applications, on which petitioners could include information about firearms; 2) searched California’s Automated Firearms System, which contains records of firearms purchases, and other relevant state databases; and 3) interviewed the DVRO petitioner, if necessary. At the time the DVRO was served, or after if it was served by a private party, law enforcement officers explained firearms restrictions to DVRO respondents, indicating that respondents could either immediately surrender their firearms or sell them to a retailer. Firearms were recovered from 23% of the respondents identified as having them in San Mateo County and from 51% in Butte County. Two explanations were put forth to explain low rates of firearm recovery. First, not all DVROs were served, and second. law enforcement lacked the authority to search respondents identified as firearm owners who denied having firearms. Importantly, there were no reports of law enforcement’s encountering severe violence during attempted recoveries.

12 Zeoli et al.

Implications of intervention studies

Intervention studies located by our review focused solely on state-level firearm legislation and its implementation in the United States. Although these studies differed greatly in methodologies, the 2 ecological time-series analyses of the association between DVRO firearm prohibition statutes and IPH and the individual-level analysis of the statutes’ association with IPV indicated that these statutes are associated with reductions in violence (31–34). The statutes studied were enacted during a time in which many laws designed to reduce IPV were passed and access to victim services increased; therefore, it is possible that these factors confounded the estimated relationship between these statutes and IPV outcomes. The consistency of findings from the reported studies, coupled with the range of factors and laws controlled for in each, however, increases confidence in the validity of the estimated associations, but future research should strive to address this limitation. None of the studies that examined impacts of the statutes under study, however, measured the implementation of the laws. Research on the DVRO firearm prohibition statutes suggests that implementation is not widespread (35, 36), which may suggest that the statutes would have little impact. However, findings from Vigdor and Mercy (31) actually underscore this point. DVRO firearm prohibition statutes were not associated with reductions in IPH when they restricted only possession, a restriction that requires a great deal of action from judges and law enforcement to implement. Additionally, no association was found in states that failed to put disqualifying information in background check systems, making implementation incomplete. The DVRO firearm prohibition statute, then, was associated with reductions in IPH only in those conditions in which it was most easily implemented (31).

Multiple avenues exist to increase the coverage and implementation of DVRO firearm prohibitions. For example, legislators in states that do not have these laws or restrict only possession should enact them with purchase restrictions (31). Legislators may also consider prioritizing and systematizing the immediate entry of disqualifying DVROs and convictions into relevant background check systems to better enforce purchase restrictions (31, 35, 41). Procedures must also be developed to ensure that DVRO firearm possession restrictions are enforced. Within recent years, multiple states have enacted provisions specifying that prohibited abusers must surrender their firearms (42); now, jurisdictions are facing the challenge of recovering them. Part of this challenge is identifying who has firearms. Initiatives to increase firearm recovery in 2 California counties relied, to much success, on multiple data sources to identify those who possessed firearms (37). Additional efforts to develop and evaluate effective firearm recovery practices are needed, particularly in states that, unlike California, do not have firearm registries. However, research also showed that judges failed to order firearm removal in many DVRO cases in which victims requested it (36). This point is echoed in qualitative research that found judges’ views on both gun control and IPV influenced their willingness to order firearm surrender in DVRO cases (43). The use of specialized domestic courts, in which court personnel have a greater understanding of IPV and associated risks, may help to overcome lack of implementation of the law at the court level (43). Questions remain regarding whether specific provisions within state DVRO firearm prohibition statutes influence the impact of these laws. In addition to variations in whether states prohibit firearm purchase as well as possession, states differ on whether their statutes cover ex parte DVROs and what relationships are protected under the law. The federal law does not apply to ex parte DVROs or dating partners who have not had a child or lived together, although some state statutes do. However, in 2013, dating partners accounted for roughly 48% of IPH perpetrators in the United States, and firearms were the most commonly used weapons in these homicides (1). Additionally, researchers found that applicants for ex parte DVROs who had not married, lived with, or had a child with the DVRO defendant, and therefore would not qualify for the federal firearm ban, mentioned firearms in their applications as often as those who did qualify for the federal firearm ban (11). This suggests that dating partners and ex parte DVROs are important to cover under the law. Although the impact of these laws and their provisions must be studied further, there are also discrepancies in published research on when these statutes were enacted. It is unclear why there are variations between Vigdor and Mercy (31) and Bridges et al. (33) in the years of enactment for state DVRO and domestic violence misdemeanor firearm prohibition statutes. For example, Bridges et al. (33) list New Jersey as gaining a DVRO firearm prohibition statute in 2003, while Vigdor and Mercy (31) list the New Jersey law’s year of passage as 1991. In response to this discrepancy, we conducted our own legal research and found that the law was passed as part of New Jersey’s “Prevention of Domestic Violence Act of 1991,” P.L. 1991, c.261. While a full investigation into all of the statutes’ years of enactment was beyond the

Downloaded from http://epirev.oxfordjournals.org/ at University of California, San Diego on January 29, 2016

associated with generally violent/antisocial IPV offenders (39), such as engaging in more severe violence, violence with weapons, and threatening to harm or kill the IPV victim or others, are more likely to have access to firearms (11, 40). However, other factors indicating general violence, such as prior criminal history, attempted homicide, police involvement, severe violence such as sexual assault, and problems such as gambling and substance abuse, may not be associated with gun access (11, 18, 21). Researchers should seek to better understand firearm possession and use in the context of IPV offender subtypes. Third, the use of ambiguous measurements of which intimate partner has firearm access complicates the interpretation of results. Two studies examined whether women who had firearms in their homes were at an increased risk of IPV victimization (15, 17). However, without knowing whether a woman cohabits with her partner, it is unknown if the presence of firearms in her home indicates that her intimate partner has access, or she has sole access, to them. Conversely, the absence of firearms in a woman’s home does not indicate that her intimate partner lacks access. We recommend that researchers take the approach of Campbell et al. (23) by measuring firearm ownership, firearm presence in the household, and cohabitation, which allowed them to reasonably isolate which partners had access to firearms.

Firearms in Intimate Partner Violence 13

scope of this review, for those states we examined at random, our findings were consistent with those of Vigdor and Mercy (31). The field would benefit from a verified data source on enactment of these laws, as well as provisions of these laws, to facilitate the investigation of their impacts. Moving forward

Closing

This review is limited in that it includes neither evaluations of broad interventions for firearm violence that may impact IPV nor evaluations of interventions for IPV that are not specific to firearms but may affect firearm use by intimate partners. However, our literature search uncovered only 2 such studies, both examining the impact of more general firearm laws on IPH (25, 48). Second, despite the fact that we did not limit our search parameters to research from the United States,

ACKNOWLEDGMENTS

Author affiliations: School of Criminal Justice, Michigan State University, East Lansing, Michigan (April M. Zeoli, Rebecca Malinski); and School of Criminal Justice, Rutgers University, Camden, New Jersey (Brandon Turchan). Conflict of interest: none declared.

REFERENCES 1. Federal Bureau of Investigation, US Department of Justice. Uniform Crime Reporting Program Data: Supplementary Homicide Reports, 2013. Ann Arbor, MI: Inter-University Consortium for Political and Social Research; 2015. (ICPSR36124-v1). 2. Cooper A, Smith EL. Homicide Trends in the United States, 1980–2008. Washington, DC: US Department of Justice; 2011. 3. Federal Bureau of Investigation, US Department of Justice. Uniform Crime Reporting Program Data: Supplementary Homicide Reports, 2009. Ann Arbor, MI: Inter-University Consortium for Political and Social Research; 2011. (ICPSR30767-v1). 4. Federal Bureau of Investigation, US Department of Justice. Uniform Crime Reporting Program Data: Supplementary Homicide Reports, 2010. Ann Arbor, MI: Inter-University Consortium for Political and Social Research; 2012. (ICPSR33527-v1). 5. Federal Bureau of Investigation, US Department of Justice. Uniform Crime Reporting Program Data: Supplementary Homicide Reports, 2011. Ann Arbor, MI: Inter-University Consortium for Political and Social Research; 2013. (ICPSR34588-v1). 6. Federal Bureau of Investigation, US Department of Justice. Uniform Crime Reporting Program Data: Supplementary Homicide Reports, 2012. Ann Arbor, MI: Inter-University

Downloaded from http://epirev.oxfordjournals.org/ at University of California, San Diego on January 29, 2016

Expanding the scope of future research and improving surveillance systems would enhance our knowledge about the risks and prevention of firearm access in IPV. For example, researchers should investigate the likelihood of criminal justice system involvement among perpetrators of IPV who have access to or use firearms. We did not find any studies that investigated firearm access as a risk for criminal justice system involvement or being under a DVRO. Findings from the National Crime Victimization Survey suggest that more than 60% of violent crimes committed with firearms are reported to the police (44). However, data also suggested that the most common reasons why people do not report crimes to the police are because they view them as private matters or do not believe them to be serious (44), 2 reasons victims of IPV, specifically, do not report violence to the authorities (45). Additionally, only half of the women killed by their intimate partners utilized the criminal justice system for concerns about ongoing IPV and stalking in the year before their murders (46). Because firearm prohibition statutes apply only to those involved in the justice system in specific ways, there is likely an important group of perpetrators untouched by these interventions. Future research should explore the size and characteristics of this group, determining possible opportunities for intervention. Given the gravity of the problem and the need for high quality research, violence surveillance systems should be improved and updated to include additional relevant variables. For example, the Supplementary Homicide Reports, which provided data for each of the studies of policy impact on IPH in this review (31–33), are a useful database of homicides throughout the United States, but they have many limitations. For example, jurisdictions often fail to report homicides to the voluntary system, and there are frequently missing data for reported homicides (47). Additionally, because the Supplementary Homicide Report’s victim-offender relationship variable does not include former dating partners as a category, we cannot test the effect of the firearm laws on the full problem of IPH in the United States. The Centers for Disease Control and Prevention’s National Violent Death Reporting System overcomes this limitation by including former dating partners among relationships measured but does not yet have nationwide coverage.

only 1 study in this review occurred outside the United States (18). Our exclusion of articles not published in English likely contributed to an underrepresentation of evidence on firearms and IPV internationally. Finally, because we included only peer-reviewed journal articles, research that appeared only in governmental reports or other non-peer-reviewed sources was not located. We know of 1 study that was excluded because it was published only as a report (49); however, this study leads to many of the same conclusions derived from the included research. Although there is a great deal of research still to be conducted to answer questions about firearm risk in IPV and the impact of IPV firearm-specific interventions on IPV outcomes, existing research converges around 3 main points: 1) firearm access by violent intimates is associated with increased risks of severe and fatal violence; 2) firearm prohibitions for those currently under DVROs are associated with reductions in IPH; and 3) more can be done to increase the enforcement of DVRO firearm purchase and possession restrictions. By expanding our knowledge of firearms in IPV and translating that knowledge into policy and practice, we will continue to reduce the burden of firearm use in intimate partner violence.

14 Zeoli et al.

7. 8.

9.

10.

12. 13.

14. 15.

16. 17. 18.

19.

20. 21. 22. 23.

24.

25. Roberts DW. Intimate partner homicide: relationships to alcohol and firearms. J Contemp Crim Justice. 2009;25(1):67–88. 26. Wintemute GJ, Wright MA, Drake CM. Increased risk of intimate partner homicide among California women who purchase handguns. Ann Emerg Med. 2003;41(2):281–283. 27. Campbell JC. Nursing assessment for risk of homicide with battered women. ANS Adv Nurs Sci. 1986;8(4):36–51. 28. Hudson W, McIntosh S. The Index of Spouse Abuse: two quantifiable dimensions. J Marriage Fam. 1981;7(2):873–888. 29. Marshall L. Development of the severity of violence against women scales. J Fam Violence. 1992;7(2):103–121. 30. Cook PJ, Ludwig J. Guns in America: Results of a Comprehensive National Survey on Firearms Ownership and Use—Summary Report. Washington, DC: The Police Foundation; 1996. 31. Vigdor ER, Mercy JA. Do laws restricting access to firearms by domestic violence offenders prevent intimate partner homicide? Eval Rev. 2006;30(3):313–346. 32. Zeoli AM, Webster DW. Effects of domestic violence policies, alcohol taxes and police staffing levels on intimate partner homicide in large US cities. Inj Prev. 2010;16(2): 90–95. 33. Bridges FS, Tatum KM, Kunselman JC. Domestic violence statutes and rates of intimate partner and family homicide: a research note. Crim Justice Policy Rev. 2008;19(1):117–130. 34. Dugan L. Domestic violence legislation: exploring its impact on the likelihood of domestic violence, police involvement, and arrest. Criminol Public Policy. 2003;2(2):283–312. 35. Vittes KA, Sorenson SB. Keeping guns out of the hands of abusers: handgun purchases and restraining orders. Am J Public Health. 2008;98(5):828–831. 36. Webster DW, Frattaroli S, Vernick JS, et al. Women with protective orders report failure to remove firearms from their abusive partners: results from an exploratory study. J Womens Health (Larchmt). 2010;19(1):93–98. 37. Wintemute GJ, Frattaroli S, Claire BE, et al. Identifying armed respondents to domestic violence restraining orders and recovering their firearms: process evaluation of an initiative in California. Am J Public Health. 2014;104(2):e113–e118. 38. Frattaroli S, Vernick JS. Separating batterers and guns: a review and analysis of gun removal laws in 50 states. Eval Rev. 2006; 30(3):296–312. 39. Holtzworth-Munroe A, Stuart GL. Typologies of male batterers: three subtypes and the differences among them. Psychol Bull. 1994;116(3):476–497. 40. Sorenson SB, Wiebe DJ. Weapons in the lives of battered women. Am J Public Health. 2004;94(8):1412–1417. 41. Zeoli AM, Frattaroli S. Evidence for optimism: policies to limit batterers’ access to guns. In: Webster DW, Vernick JS, eds. Reducing Gun Violence in America: Informing Policy with Evidence and Analysis. Baltimore, MD: Johns Hopkins University Press; 2013:53–63. 42. Zeoli AM, Frattaroli S. Enactment of legislation to protect victims of domestic violence. In: Webster DW, Vernick JS, eds. Updated Evidence and Policy Developments on Reducing Gun Violence in America. Baltimore, MD: Johns Hopkins University Press; 2014:10–12. 43. Frattaroli S, Teret SP. Understanding and informing policy implementation: a case study of the domestic violence provisions of the Maryland Gun Violence Act. Eval Rev. 2006; 30(3):347–360. 44. Hart TC, Rennison CM. Reporting Crime to the Police, 1992–2000. Washington, DC: US Department of Justice; 2003. 45. Fanslow JL, Robinson EM. Help-seeking behaviors and reasons for help seeking reported by a representative sample of

Downloaded from http://epirev.oxfordjournals.org/ at University of California, San Diego on January 29, 2016

11.

Consortium for Political and Social Research; 2014. (ICPSR35023-v1). Truman JL, Morgan RE. Nonfatal Domestic Violence, 2003– 2012. Washington, DC: US Department of Justice; 2014. Hamby S. Self-report measures that do not produce gender parity in intimate partner violence: a multi-study investigation [ published online ahead of print October 27, 2014]. Psychol Violence. (doi:10.1037/a0038207). Black MC, Basile KC, Breiding MJ, et al. The National Intimate Partner and Sexual Violence Survey (NISVS): 2010 Summary Report. Atlanta, GA: National Center for Injury Prevention and Control, Centers for Disease Control and Prevention; 2011. Addington LA, Perumean-Chaney SE. Fatal and non-fatal intimate partner violence: what separates the men from the women for victimizations reported to police? Homicide Stud. 2014;18(2):196–220. Vittes KA, Sorenson SB. Are temporary restraining orders more likely to be issued when applications mention firearms? Eval Rev. 2006;30(3):266–282. Abrahams N, Jewkes R, Martin LJ, et al. Mortality of women from intimate partner violence in South Africa: a national epidemiological study. Violence Vict. 2009;24(4):546–556. Belfrage H, Rying M. Characteristics of spousal homicide perpetrators: a study of all cases of spousal homicide in Sweden 1990–1999. Crim Behav Ment Health. 2004;14(2): 121–133. Pereira AR, Vieira DN, Magalhães T. Fatal intimate partner violence against women in Portugal: a forensic medical national study. J Forensic Leg Med. 2013;20(8):1099–1107. Coker AL, Smith PH, McKeown RE, et al. Frequency and correlates of intimate partner violence by type: physical, sexual, and psychological battering. Am J Public Health. 2000;90(4): 553–559. Shuman RD Jr, McCauley J, Waltermaurer E, et al. Understanding intimate partner violence against women in the rural South. Violence Vict. 2008;23(3):390–405. Vest JR, Catlin TK, Chen JJ, et al. Multistate analysis of factors associated with intimate partner violence. Am J Prev Med. 2002;22(3):156–164. Folkes SEF, Hilton NZ, Harris GT. Weapon use increases the severity of domestic violence but neither weapon use nor firearm access increases the risk or severity of recidivism. J Interpers Violence. 2013;28(6):1143–1156. Frye V, Manganello JA, Campbell JC, et al. The distribution of and factors associated with intimate terrorism and situational couple violence among a population-based sample of urban women in the United States. J Interpers Violence. 2006;21(10): 1286–1313. McFarlane J, Soeken K, Campbell J, et al. Severity of abuse to pregnant women and associated gun access of the perpetrator. Public Health Nurs. 1998;15(3):201–206. Rothman EF, Hemenway D, Miller M, et al. Batterers’ use of guns to threaten intimate partners. J Am Med Womens Assoc. 2005;60(1):62–68. Glass N, Perrin N, Hanson G, et al. Risk for reassault in abusive female same-sex relationships. Am J Public Health. 2008; 98(6):1021–1027. Campbell JC, Webster DW, Koziol-McLain J, et al. Risk factors for femicide in abusive relationships: results from a multisite case control study. Am J Public Health. 2003;93(7): 1089–1097. Glass N, Laughon K, Rutto C, et al. Young adult intimate partner femicide—an exploratory study. Homicide Stud. 2008; 12(2):177–187.

Firearms in Intimate Partner Violence 15

women victims of intimate partner violence in New Zealand. J Interpers Violence. 2010;25(5):929–951. 46. McFarlane JM, Campbell JC, Wilt S, et al. Stalking and intimate partner femicide. Homicide Stud. 1999;3(4):300–316. 47. Fox JA, Swatt ML. Multiple imputation of the Supplementary Homicide Reports, 1976–2005. J Quant Criminol. 2009;25(1): 51–77.

48. McPhedran S, Mauser G. Lethal firearm-related violence against Canadian women: Did tightening gun laws have an impact on women’s health and safety? Violence Vict. 2013;28(5):875–883. 49. Moracco KE, Clark KA, Espersen C, et al. Preventing Firearms Violence Among Victims of Intimate Partner Violence: An Evaluation of a New North Carolina Law. Washington, DC: US Department of Justice; 2006.

Downloaded from http://epirev.oxfordjournals.org/ at University of California, San Diego on January 29, 2016

Risks and Targeted Interventions: Firearms in Intimate Partner Violence.

The use of firearms in intimate partner violence (IPV) is widely recognized as an important public health threat. However, what we know about the risk...
566B Sizes 0 Downloads 8 Views