Law Enforcement Officer Knowledge of, Attitudes Toward, and Willingness to Use Extreme Risk Protection Orders

Key Points Question How familiar are law enforcement officers (LEOs) with extreme risk protection orders (ERPOs), and in what circumstances do they consider ERPOs to be appropriate? Findings This survey study of 283 active-duty LEOs in states with ERPO laws found that knowledge of ERPOs was high, but willingness to use ERPOs varied. Across all provided scenarios, LEOs with ERPO training or experience were substantially more likely to support the use of ERPOs than those without. Meaning The findings of this study suggest that training LEOs on ERPO laws may improve their implementation.


Introduction
Extreme risk protection order (ERPO) laws temporarily prohibit individuals judged to be at high risk of harming themselves or others from purchasing or possessing firearms.Currently, 21 states and the District of Columbia have ERPO laws, with most enacted after the school shooting in Parkland, Florida, in February 2018.The ERPO laws seem to be effective at preventing firearm suicide [1][2][3] and possibly mass shootings, [4][5][6] but slow uptake has limited their impact. 7,8plementation and use of ERPOs largely depends on law enforcement officers (LEOs), who are permissible petitioners (ie, the individuals requesting ERPOs) in all ERPO states but Vermont and are the only permissible petitioner in 4 states. 91][12] However, lack of awareness, limited training, and local politics can all serve as barriers to officers using ERPOs. 13Prior surveys suggest that knowledge of these laws is limited among the general population and certain petitioners, with approximately two-thirds of California adults and more than 70% of Maryland physicians (who are permitted petitioners in that state) at the Johns Hopkins Hospital in Baltimore never having heard of the law. 14,15Training for LEOs on ERPOs might help raise awareness and provide guidance for implementation, but little is known about how common such training is.However, even if LEOs are aware of and trained to use their state's ERPO law, implementation may be stymied by local or personal politics.A number of sheriffs, most notably in Colorado, declared their counties "Second Amendment Sanctuaries," indicating a refusal to petition for ERPOs (although approximately 25% of these counties issued ERPOs in Colorado within the law's first year 12 ).While most people support ERPOs after learning about the law, 15 including firearm owners, 16 key informant interviews previously suggested that officers are likely to have personal political reservations about using ERPOs. 13To the extent that politics are preventing individuals from having meaningful access to ERPOs, there is inequity in the law's implementation (ie, if local law enforcement refuses to petition for or serve an ERPO on political grounds, individuals living within their jurisdiction are faced with a substantial barrier to using ERPOs that others do not face).
Because LEOs play an essential role in the implementation of ERPO laws, we sought to learn more about officer knowledge of, attitudes toward, and willingness to use this law.Prior key informant interviews exploring ERPO implementation have included LEOs, but these samples are small and typically include officers from limited geographic locations. 2,3,13Building on this earlier work, this study presents results from a survey of LEOs in the District of Columbia and all 19 states with an ERPO law in 2021 that explored officer knowledge, training, and opinions about ERPO laws as well as willingness to use ERPOs across a range of scenarios.To our knowledge, this is the first study of its kind.Findings may directly inform LEO implementation of ERPO laws and may be of interest to those working in law enforcement, policy, and violence prevention.

Design
This was a cross-sectional online survey of LEOs in all ERPO states and the District of Columbia, designed by 2 of us (V.A.P. and A.C.) and administered by Qualtrics from April 5 to August 30, 2021.
Qualtrics recruits participants through nonprobability methods including opt-in research panels, social media outreach, and specialized recruitment. 17Incentives may include airline miles, gift cards, or charitable donations, among other things.Survey participants were self-reported active-duty, sworn LEOs in a state or district with an ERPO law.This was originally a California-specific study but was opened to officers in other ERPO states beginning July 12, 2021, as Qualtrics was unable to recruit the target number of participants in California alone.This study was approved by the University of California, Davis Institutional Review Board.Participants read an informed consent statement and clicked on accept to indicate consent before the survey's initiation.We followed the

Measures
Nearly all survey questions were close-ended and, when appropriate, response options were presented with Likert scales.After screening for eligibility, participants were asked to report their age, educational level, gender, and races and ethnicity.These demographic features were collected to facilitate our characterization of the study participants.They also were asked to indicate their political ideology and political party preference.Following this, we asked about their job: the year that they became an officer, their current job title and assignments, and characteristics of their agency.
Next, we asked participants about their familiarity with and opinions about ERPO laws.We asked whether they had ever heard of ERPOs, gun violence restraining orders, or red flag laws, and how familiar they were with these laws.We then provided a brief description of ERPO laws and asked participants to indicate their agreement with various statements about the law, such as "[ERPO]s are likely to reduce firearm violence."We also asked which groups of people the participants thought should be permitted to petition for ERPOs.
Following this, we asked about training and experience with ERPOs.Specifically, we asked whether the participant or their department had "been directly involved" with an ERPO.We also asked whether they received training on their state's ERPO law, how knowledgeable they felt about when and how to petition for an ERPO, and how confident they were that they could effectively serve an ERPO.
Finally, we presented the participants with 5 brief scenarios and, following each, asked about (1)   the extent to which they agreed that "the police officers responding to this situation should petition the court to issue a[n ERPO]" and ( 2) what other action the responding officer should take.
Participants could choose any combination of the following actions: make an arrest, seek a domestic violence restraining order, seek psychiatric assessment/hospitalization (eg, 5150), and take some other action; alternatively, they could select none of these.We also solicited free-text open-ended responses allowing participants to specify what action they would take if "take some other action" was selected and to explain why they agreed or disagreed with petitioning for an ERPO.The scenarios covered a range of potential ERPO situations, including intimate partner violence (IPV), suicide, homicide, and mass shootings (eTable 1 in Supplement 1 presents the full text of all scenarios).We created 2 versions of each scenario, varying a single feature of the case in each, and randomized the version presented to participants.Differences between versions of the scenarios included gender, directionality of the threat (self vs other), risk factor history, firearm type, and race.
The full text of the survey instrument is included in the eMethods in Supplement 1.

Statistical Analysis
Our analysis of close-ended questions included standard descriptive statistics and χ 2 tests (α = .05) to make comparisons across groups, including across randomly allocated versions of each scenario.
Testing was 2-sided and unpaired.We assessed randomization by testing for differences in the demographic characteristics (age, educational level, race and ethnicity, state of residence, and political ideology) of participants randomized to different scenario versions.These analyses were done in R, version 4.

(R Foundation for Statistical Computing).
To analyze the open-ended questions following each scenario, we developed qualitative codebooks through an inductive, iterative review of responses, with care given to the consideration of deviant responses. 19Two coders were trained on each codebook who then independently coded a sample of 30 responses. 20We calculated Krippendorff α in Stata, version 18 (StataCorp LLC) to determine intercoder agreement for each code. 21We considered codes with α Ն0.80 to have good agreement, indicating that our coding frame was sufficiently specified to support its use across coders. 22 Most participants (81.3%) were somewhat or very familiar with ERPO laws (Table 3).While 56.2% of participants received ERPO training, over three-quarters reported that they were somewhat or very knowledgeable about deciding when an ERPO is appropriate (86.2%) and about the procedures involved in petitioning for an ERPO (75.6%).Just over half (52.7%) reported that their department had been directly involved in an ERPO, and 40.6% reported direct personal involvement.When asked about who should be allowed to petition for an ERPO, most (79.2%)selected law enforcement, whereas fewer than half selected mental health professionals (47.3%), family members (43.5%), and health care professionals (41.7%) (eTable 3 in Supplement 1).
The opinions of LEOs on ERPO laws were favorable overall but varied substantially by political ideology, with self-identified conservative participants having less favorable opinions of the law and being less willing to use it than self-identified liberal participants (Figure 1; eTable 4 in Supplement 1).
For example, 91.6% of liberal participants at least somewhat agreed that "ERPOs are likely to reduce firearm violence" and 92.8% agreed that "other states should adopt these policies"; in contrast, just 52.1% of conservative participants at least somewhat agreed that ERPOs are likely to reduce firearm violence" and 59.7% agreed that "other states should adopt these policies."Approximately one-quarter of conservative (26.9%) and moderate (24.5%) participants at least somewhat disagreed that they would "petition for an ERPO under the right circumstances and with appropriate training"; across all participants, 21.9% at least somewhat disagreed (eTable 4 in Supplement 1).
Participants randomized to different versions of each scenario were well balanced on demographic characteristics.Most officers at least somewhat agreed with petitioning for an ERPO across all specific scenarios (Figure 2A; eTable 5A in Supplement 1).Support for an ERPO tended to be higher in cases involving IPV (71.4%-78.6%)and lower in cases involving suicidality (54.2%-73.3%)and non-IPV interpersonal violence (56.0%-65.8%).None of the randomized differences within scenarios resulted in significant differences in agreement.Participants were asked to optionally explain why they agreed or disagreed with petitioning for an ERPO after each scenario (the proportions providing explanations are presented in eTable 6A in Supplement 1).The most common reason given by those who strongly agreed with an ERPO was that there was a threat of violence (eTable 6B in Supplement 1).Among those who somewhat agreed or disagreed, the most common explanation was the need for more information.Those who strongly disagreed most commonly indicated that the ERPO was unneeded (eg, the threat was not credible).

JAMA Network Open | Health Policy
Beyond ERPOs, we explored what other actions the participants would take in each scenario, including making an arrest, seeking a domestic violence restraining order (DVRO), and seeking a psychiatric evaluation.Most participants indicated they would take some other action in addition to  or instead of an ERPO (eTable 7A in Supplement 1).In the IPV scenario in which a woman was threatened by her husband, most supported arrest (55.1%) and a DVRO (51.7%); however, the most popular other action was a DVRO (72.4%) among participants who agreed with an ERPO and arrest (50.0%) among those who disagreed (eTable 7B in Supplement 1).In the IPV scenario in which a man was threatened by his husband, most LEOs supported a DVRO (70.4%).Most supported a psychiatric evaluation in all 3 suicide scenarios (58.9%-70.2%).There was not a majority agreement about other actions that should be taken in the non-IPV interpersonal violence scenarios, and the most popular action varied by whether the participant supported petitioning for an ERPO.

Discussion
This of 283 active-duty LEOs working in states with ERPO laws found that participants' selfreported knowledge of ERPOs was high, that a little more than half had received ERPO training, and that a large minority had direct personal experience with the law.The level of ERPO knowledge, training, and use in this sample was unexpectedly high given that ERPO use is quite rare. 7This could suggest that ERPO use is greater than previous research reports, that participants conflated ERPOs with other protective orders, or that officers who were more familiar with ERPOs participated in the survey more often than those who were less familiar.This sample is not representative of all activeduty officers in states with ERPO laws, and our findings with regard to knowledge, training, and use likely do not generalize to that population.Nevertheless, we found 2 very interesting patterns in the data, neither of which should be influenced by this limitation, that can directly inform ERPO implementation efforts.
First, self-reported political ideology was significantly associated with officers' opinions about and general willingness to use ERPOs, with conservative participants being more skeptical of the law and less willing to use it than liberal participants.Additionally, more than one-quarter of conservative participants disagreed that they would petition for an ERPO "under the right circumstances and with appropriate training" (compared with 8.4% of liberal participants).This may suggest an ideologically

Somewhat agree Strongly agree Response
A total of 251 of 283 participants were included; 32 participants were excluded due to unknown political ideology.For each opinion, the P value from a χ 2 test for independence by political ideology was <.01.motivated refusal to use a law that, for the sake of equity, should be meaningfully available to everyone living in an ERPO state.These results bolster prior qualitative findings suggesting that personal and local politics may be a barrier to LEO uptake of ERPOs. 13Other research has similarly found that conservative law enforcement personnel are less likely to support firearm regulations. 24 Training may also help dispel some of the misconceptions officers seem to have about when ERPOs can be used.For example, about 1 in 3 participants who strongly disagreed with using an ERPO in the school shooting scenario justified their position by noting that no illegal behavior was reported (eTable 8 in Supplement 1).Smaller proportions justified disagreeing with ERPOs in other scenarios because the threat did not involve firearms or firearms were not brandished.However, ERPO laws require neither that the respondent commit a crime nor that a firearm is used or owned (in fact, preventing the purchase of firearms is a key component of ERPOs).Similarly, a small number of officers noted concerns about ERPOs violating due process or the Second Amendment.Training could clarify that ERPO laws are widely considered and adjudicated to be constitutional. 26,27e results also suggest that training on how to respond to crises could be improved in general (note the variation in eTable 7A in Supplement 1) and that ERPO training could be improved, particularly with respect to clarifying the role that ERPOs can play in suicide prevention.[3] However, LEOs were the least likely to agree with using an ERPO in the scenarios.This reticence is also reflected in how ERPOs are being used in California, where only 15% of ERPOs in the first 3 years were issued for suicide threats alone. 4Extreme risk protection orders may be underused for suicide prevention in Washington and Colorado as well. 10,12 addition to training, procedural and structural changes may support the creation of a robust LEO ERPO implementation effort.This includes forming partnerships between police departments and city attorney's offices, who can represent LEO petitioners; this removes the burden of filing petitions and attending court hearings from LEOs, who previously identified these things, as well as discomfort with self-representation in court, as barriers. 13Creating dedicated ERPO units in larger police departments can also support implementation and reduce risk of injury by having officers who are specially trained in firearm removal serve the orders.Finally, if feasible, regional ERPO teams covering many jurisdictions could promote implementation in areas served by smaller police departments that may not have the resources or demand for specialized ERPO units.

Limitations
This study's findings should be considered in light of its limitations.As is true of all forms of survey research, this study is potentially subject to unmeasured error.Status as an active-duty LEO was selfreported; however, given that Qualtrics uses double-opt in market research panels with screening questions to identify hard-to-reach groups, we have confidence that our participants belonged to our target population.Additionally, data are from a nonrepresentative sample and LEOs from California are overrepresented; findings may not generalize to the broader population of LEOs working in states with an ERPO law.Future studies using representative samples are warranted.Furthermore,

Figure 1 .
Figure 1.Extreme Risk Protection Order (ERPO) Opinions by Political Ideology Law Enforcement Officer Use of Extreme Risk Protection Orders JAMA Network Open.2023;6(10):e2338455. doi:10.1001/jamanetworkopen.2023.38455(Reprinted) October 19, 2023 2/12 Downloaded From: https://jamanetwork.com/ on 10/22/2023 American Association for Public Opinion Research (AAPOR) reporting guideline for survey research. 18 23r items below this threshold, the coders met to reconcile disagreements before independently coding an additional 30 responses.Agreement was reached through an open discussion of coding decisions, a process that may have required adjustments to the codebook (eg, merging, redefining, or expanding codes).23Allcodes reached α Ն0.80 after a second round of coding and were then single coded.Questions with fewer responses were fully double coded, with disagreements reconciled in the manner described above.
After reaching this threshold, responses were single coded.Most codes exceeded this JAMA Network Open | Health Policy Law Enforcement Officer Use of Extreme Risk Protection Orders JAMA Network Open.2023;6(10):e2338455. doi:10.1001/jamanetworkopen.2023.38455(Reprinted) October 19, 2023 3/12 Downloaded From: https://jamanetwork.com/ on 10/22/2023 threshold after the first round of coding.(85.2%) and non-Hispanic White officers (71.4%) (Table 1).Most participants lived in California (53.7%), followed by New York (14.5%) and Florida (10.6%).Five ERPO states were not represented in the analytic sample (Hawaii, Maryland, New Mexico, Rhode Island, and Vermont).A plurality (42.0%) of participants self-identified as politically conservative.Participants had been working in law enforcement for a median of 13 (IQR, 6-20) years (Table 2).A total of 59.4% worked as officers or deputies, 24.4% worked in positions of leadership (ie, supervisory roles), and 16.3% worked in other roles.At the time of the survey, 64.0% of participants were assigned to patrol/enforcement, 29.7% were assigned to specialized units (eg, special weapons and tactical teams), and 19.8% were engaged in community policing.A total of 10.2% worked in training and 10.6% had other assignments.Participants worked for agencies of varying sizes and urbanicity.
Law Enforcement Officer Use of Extreme Risk Protection Orders However, there were substantial differences in agreement between officers with and without ERPO training or direct personal experience with ERPOs: those with training or experience were 14 to 25 percentage points more likely to at least somewhat agree with petitioning for an ERPO across scenarios (Figure2B; eTable 5B in Supplement 1).The differences were particularly notable in the suicide and homicide scenarios, as most of those with training or experience agreed with seeking an ERPO (suicide, 62.8%; homicide, 67.1%) and most of those without training or experience disagreed JAMA Network Open.2023;6(10):e2338455. doi:10.1001/jamanetworkopen.2023.38455(Reprinted) October 19, 2023 4/12 Downloaded From: https://jamanetwork.com/ on 10/22/2023

Table 2 .
Law Enforcement Employment Details a These positions comprise supervisory roles including supervisory investigator, sergeant, lieutenant, captain, major, commander, deputy chief, and chief.bCategories are not mutually exclusive; 27.6% of officers indicated working in at least 2 areas.
,25Figure 2. Agreement With Petitioning for an Extreme Risk Protection Order (ERPO) in Hypothetical ScenariosSecond, we found that ERPO training or direct personal experience was associated with officers' agreement with using ERPOs across a range of scenarios.In all scenarios, officers with training or experience were more likely to agree with using an ERPO than those without training and experience-usually by about 20 percentage points.This suggests both the necessity and value of ERPO training, as it seems to convey to officers the potential utility of ERPOs in preventing firearm violence of all kinds.
JAMA Network Open.2023;6(10):e2338455. doi:10.1001/jamanetworkopen.2023.38455(Reprinted) October 19, 2023 8/12 Downloaded From: https://jamanetwork.com/ on 10/22/2023 findings suggest possible sampling bias, potentially with LEOs more familiar with ERPOs disproportionately opting into the survey compared with those less familiar.This impacts the generalizability of the total proportions reporting knowledge of, training on, and experience with the law but is unlikely to affect our main findings.Finally, we are lacking information about the quality and extent of participants' ERPO training.Identifying heterogeneity in training, as well as best practices, are areas ripe for research.27.Blocher J, Charles JD. Firearms, extreme risk, and legal design: "red flag" laws and due process.Virginia Law review Association.October 19, 2020.Accessed September 8, 2023.https://virginialawreview.org/wp-content/ uploads/2020/12/BlocherCharles_Book.pdfDemographic Characteristics by Completion of Survey eTable 3. Opinions on ERPO Petitioners eTable 4. ERPO Opinions Overall and by Political Ideology eTable 5. Agreement with Petitioning for an ERPOs in Hypothetical Scenarios eTable 6. Reasons For and Against Petitioning for an ERPO by Scenario eTable 7. Suggested Further Actions eTable 8. Indications of Misconceptions About ERPOs Laws in Qualitative Explanations of Why/Why Not an ERPO Should Be Used in Specific Case Scenarios eMethods.Survey Instrument