Variation in Extreme Risk Protection Order Use Among Urban,Suburban,and Rural ZIP Codes in Maryland: A Descriptive Study

Abstract

Extreme Risk Protection Orders (ERPOs) are a tool used to temporarily remove access to firearms from those who may pose a risk of violence to themselves or others. Little research on ERPOs has examined variation in use of the law between urban and rural jurisdictions. Data abstracted from Maryland ERPO casefiles between October 2018 and June 2020 were paired with ZIP code-level Rural Area Commuting Area codes and American Community Survey population data. Rate of ERPO use was tabulated for 3 urbanicity categories (urban, suburban, and rural). Chi-squared and 1-way ANOVA tests were used to determine between-urbanicity variation in respondent characteristics, precipitating threat, petitioner type, and court outcomes. Univariate logistic regression supplemented these analyses when a statistically significant difference was evident. The observed rate of ERPO use was 1% higher in rural areas and 66% higher in suburban areas than urban areas. Racial demographics of respondents appeared to reflect underlying distributions of race in each urbanicity category. The types of threats precipitating ERPOs were overall similar, although suicide attempts were more common in non-urban petitions and interpersonal violent threats were most common in urban petitions. Family members and minors were more frequent targets of violence in the events precipitating rural petitions compared with non-rural petitions. Law enforcement officers were over twice as likely to petition in rural areas than in suburban or urban areas. Interim hearings occurred more frequently in suburban and rural areas than urban areas (although this result was not statistically significant for rural petitions). ERPOs are used in urban, suburban, and rural areas to address largely similar forms of violence. While the rate of ERPO use is higher in less urban areas, civilians comprise a smaller proportion of rural petitioners than urban or suburban petitions. Implications for implementation practice and further research are discussed.

Keywords

firearm extreme risk protection order suicide violence policy rural

Introduction

In 2023, over 46 000 people in the United States lost their lives to gun violence, amounting to an age-adjusted rate of 13.7 deaths per 100 000.¹ Researchers and practitioners have urged policymakers to conceptualize and address this crisis not merely as a criminal justice issue, but as a matter of public health.² One promising public health approach to the gun violence epidemic is extreme risk protection orders (ERPOs). ERPOs are an evidence-based tool for temporarily removing firearms from those who may pose an imminent threat to themselves and/or others. Both firearm suicide and interpersonal firearm violence are often precipitated by identifiable warning signs that may be apparent to law enforcement, family members, clinicians, and others who have contact with the person at risk.^3
-5 As of March 13, 2025, 21 states, the District of Columbia, and the U.S. Virgin Islands have implemented ERPO laws.

Maryland is 1 of 7 jurisdictions where healthcare professionals are authorized ERPO petitioners; family members and intimate partners are also authorized to petition in Maryland.^6,7 Across all implementing states, the vast majority of ERPO petitions tend to be granted, with law enforcement petitions generally granted at a higher rate than civilian petitions in states where both groups are authorized to petition.^8,9

Research suggests that ERPOs are a promising tool for preventing gun violence. Among states that have implemented ERPO laws, ERPOs are being used to intervene in threatened mass shootings,¹⁰ suicidal crises,^11,12 and interpersonal violence.^13,14 The potential of ERPOs for addressing gun suicide has been particularly well-documented. In Oregon’s first 15 months of implementation, 73% of ERPO respondents had a history of suicide attempts.¹⁴ Swanson et al also found evidence that ERPOs reduce gun suicides based on predicted suicides in absence of the law, with an estimated 1 suicide prevented for every 13 to 18 ERPOs filed where suicide was a dangerous behavior described in the petition.¹⁵

Despite the potential benefits of ERPO laws, recent state-level implementation experiences suggest a number of barriers to uptake. Law enforcement officers have described the process of filing ERPOs as time-intensive,¹⁶ and civilian petitioners have likewise reported being confused by the process.¹⁷ Perhaps unsurprisingly, the proportion of civilian petitioners for ERPOs has been limited, even in states where several categories of civilians are authorized to petition. In California, 96.1% of petitioners between 2016 and 2018 were law enforcement officers,¹² and in King County, Washington from 2017 to 2018, 97.3% of petitions were law enforcement-initiated (although just over half of these petitions were initiated because family members or friends contacted law enforcement to alert them to a threat).¹³

As researchers and practitioners explore approaches to improving ERPO implementation, an important question is whether patterns in ERPO implementation differ between urban, rural, and suburban areas, both by count and rate of ERPO petitions (accounting for differences in underlying population distribution). Preliminary evidence suggests that petitions may not be equally dispersed: Rooney et al, studying petitions in the state of Washington, found that 89% of petitions were in urban ZIP codes¹⁸ and a study of Oregon’s law indicated that petitions were concentrated in more populous regions.¹⁴ In the first 4 years of California’s law (2016-2019), nearly one-third of all petitions came from a single urban county (San Diego County), while 14 counties did not issue any.¹⁶ While these findings suggest a potential concentration of petitions in urban areas, to our knowledge no research has yet investigated whether characteristics of ERPO casefiles, or overall use of ERPO relative to population, differ by urbanicity.

Potential differences in ERPO use by urbanicity have important public health implications given the toll of gun violence on rural communities. In 2023, over 8000 people in non-metro areas lost their lives to gun violence, amounting to an overall rate of 17.57 deaths per 100 000—76.4% higher than the metro rate of 13.11 per 100 000.¹⁹ The rate of gun deaths in non-metro areas increased 13% between 2018 and 2023, hitting its highest in 2021.¹⁹ These trends appear to be driven largely by gun suicides: in 2023, the rate of gun suicide in rural areas was 63% higher than in urban areas.²⁰ Additionally, some evidence suggests that suburban and rural areas have a higher total number of shootings targeted at schools compared to urban areas.^21,22 Many factors may contribute to the higher rate of gun suicide—and gun violence in general—in rural communities. People living in rural areas are significantly more likely to own guns than those in urban and suburban areas and are also more likely to own more than 1 gun.^23,24 The population in rural areas is also disproportionately white, a population among which gun suicide is more prevalent.²⁵

ERPOs may be an important tool for addressing gun violence in rural communities, particularly given their promise in addressing firearm suicide.²⁶ However, despite the clear need, there are reasons to believe that rural implementation of ERPO may be uniquely complicated. In interviews with justice system professionals involved with ERPO implementation in California, Pear et al found evidence that attitudes toward firearms and gun laws in rural communities may make it difficult to improve uptake of ERPOs.¹⁶ Limited law enforcement capacity in rural areas may also limit effectiveness, particularly because such a large proportion of ERPO petitions are filed by law enforcement. Rural law enforcement agencies are often underfunded, understaffed,²⁷ and may lack capacity to store removed guns and serve orders in a timely manner.²⁸ Research on domestic violence restraining orders (DVROs), a policy upon which ERPOs were based, has indicated that rural victims may face a greater burden to obtain an order compared to urban victims, including monetary cost and lack of personnel to navigate the filing process; in contrast, in urban areas, court personnel and law enforcement office staff are more likely to perform and assist tasks related to filing the DVRO.²⁹ Considering the existing evidence on the burden of gun violence in rural communities and the potential barriers to gun removal in these areas, this research seeks to (1) explore and characterize the nature of ERPO use in rural communities and (2) determine whether urbanicity may play a role in ERPO implementation experiences, including trends in use and court outcomes.

Methods

Data

This analysis uses data abstracted from all ERPO casefiles in the state of Maryland from October 2018 to June 2020. There were 1347 total petitions in Maryland during this period. This data was abstracted by trained researchers as part of a larger study of ERPO use in California, Colorado, Connecticut, Florida, Maryland, and Washington.³⁰ Variables included in this analysis were:

Characteristics of respondents, including respondent age, race, and sex, as well as whether they possessed a gun.

Precipitating event(s) that led to the petition, including threats and acts of violence, type of violence, targets of threats and acts of violence, and presence of a DVRO in the casefile.

Type(s) of ERPO present in the casefile, including interim, ex parte, and final orders. In Maryland, interim orders are pursued when a district court is closed (eg, after hours or on a weekend/holiday). The petition is reviewed by a district court commissioner, who determines whether to issue an interim ERPO until courts re-open and a judge can review. Ex parte (temporary) ERPOs are pursued when the district court is open and a petitioner believes that someone poses an imminent risk to themselves and/or others. If granted by a judge, this authorizes law enforcement to temporarily remove guns possessed by the respondent and prohibit them from purchasing more. Following either of these pathways, the respondent has the opportunity to participate in a final hearing within 3 business days (interim to ex parte) or 7 days (ex parte to final) to determine whether a final ERPO will be granted for up to a year.

Court outcomes, including denial, dismissal, stipulation, and granting. In an ERPO hearing, denial generally indicates that a judge found the petition lacked merit after a hearing. Dismissal means that a hearing did not or could not occur; reasons for this can include that a petitioner was not authorized to petition or did not appear in court (although the latter has also been occasionally cited as a factor in denial). Stipulation occurs when a respondent elects at the interim or ex parte stage to waive their final hearing and accept a final ERPO.

Rural Urban Commuting Area (RUCA) codes at the Zip Code Tabulation Area (ZCTA) level were obtained from USDA data. RUCA codes include urbanized areas, metropolitan areas, micropolitan areas, small towns, and rural areas, refined by the commuting flow of residents as a proxy for connection to more urbanized surrounding areas.³¹ RUCA codes were joined to the Maryland ERPO data based on respondent ZIP codes. To generate population estimates for each urbanicity category, American Community Survey (ACS) population data was joined with RUCA codes and summed by urbanicity category.³² Petitions were classified as urban (RUCA code 1, Census-designated urbanized areas of 50 000 people or more) suburban (RUCA code 2-3, metropolitan areas with 10-30% of residents commuting to urban areas), and rural (RUCA codes 4-10, or micropolitan and smaller areas). This suburban designation captures metropolitan areas geographically outside of urbanized areas but with high commuting patterns into those areas.

Records were excluded if they did not fall within October 2018 to June 2020, did not include ZIP code data, or contained a ZIP code outside of the state of Maryland.

Analysis

Chi-squared tests and 1-way ANOVA were used to test for statistically significant differences between urbanicity categories on a range of characteristics abstracted from petitions, including respondent demographics, type of threat that precipitated a petition, petitioner type, and court outcomes. When results indicated that at least 1 urbanicity category was statistically significantly different from the others on a given variable, univariate logistic regression of that variable on urbanicity category was used to determine odds ratios. These regressions were unadjusted and did not include potential confounders.

Results

Use by ZIP Code Classification

Out of n = 1347 petitions, n = 1236 (91.8%) had sufficient data to be matched and were included in the analysis. Reasons for nonmatching included incomplete ZIP code data and ZIP codes that did not match Maryland ZCTAs. Of the 1243 petitions analyzed, 961 petitions were in urban ZCTAs (77.75%), 238 petitions were in suburban ZCTAs (19.26%), and 37 petitions were in rural ZCTAs (2.99%). Based on American Community Survey data for Maryland’s ZCTAs aggregated to RUCA-based urbanicity category, the average monthly rate of petitions across the study period was 0.90 per 100 000 in urban areas, 1.50 per 100 000 in suburban areas, and 0.91 per 100 000 in rural areas (Supplemental Table 1). Thus, the observed rural rate was slightly over 1% higher than the observed urban rate and the observed suburban rate was roughly 66% higher.

Demographic Characteristics of Respondents

Across all levels of urbanicity, ERPO respondents were disproportionately male. In urban areas, 86.2% of respondents were male; in suburban and rural areas, 91.6% and 95.6% of respondents were male respectively (Table 1). Suburban respondents were more likely to be male compared to urban respondents (P = .017, Table 3), while no statistically significant difference could be detected between urban and rural or suburban and rural respondents. The mean age of respondents fell between 39 and 44 years old in each of the 3 urbanicity categories; suburban respondents were statistically significantly older than urban and rural respondents (Table 1). Less than 3% of respondents in all areas were under the age of 18 (2.4% in urban areas, 1.3% in suburban areas, and 2.7% in rural areas, Table 1).

Table 1.

Respondent Characteristics, Petitioner Types, and Threats.

Characteristic	Urban	Suburban	Rural	Chi-squared p-value
Total number of petitions	961	238	37
Gender
Male	86.2% (828)	91.6% (218)	94.6% (35)	.089
Female	13.3% (128)	7.6% (18)	5.4% (2)
Unknown	0.1% (5)	0.2% (2)	0.0% (0)
Age, mean (SD)	40.09 (14.62)	43.50 (14.12)	39.38 (15.44)	.005
Under 18	2.4% (23)	3 (1.3%)	2.7% (1)	.562
Race
White	59.2% (569)	79.0% (188)	86.5% (32)	<.001
Black	37.7% (353)	18.5% (44)	13.5% (5)
American Indian/Alaska Native	0.2% (2)	0.0% (0)	0.0% (0)
Asian	1.5% (14)	0.0% (0)	0.0% (0)
Native Hawaiian/Pacific Islander	0.1% (1)	0.0% (0)	0.0% (0)
Hispanic	1.0% (10)	0.8% (2)	0.0% (0)
Other/unknown	1.3% (12)	4 (1.7%)	0.0% (0)
Possessed a firearm	64.6% (633)	182 (76.5%)	56.8% (21)	.001
Used violence in precipitating event	34.7% (338)	35.7% (85)	43.2% (16)	.56
Target of violence
Intimate partner	47.9% (158)	38.8% (33)	43.8% (7)	.32
Family member	9.7% (32)	18.8% (16)	37.5% (6)	.001
Minor	3.9% (13)	9.4% (8)	18.8% (3)	.009
LEO	1.2% (4)	0.0% (0)	6.3% (1)	.10
Hallucination	0.6% (2)	1.2% (1)	0.0% (0)	.81
Other/unclear	11.5% (38)	16.5% 14	18.8% (3)	.36
Threatened violence in precipitating event	70.2% (675)	60.9% (145)	59.5% (22)	.011
Target of threat
Intimate partner	45.2% (305)	40.0% (58)	45.5% (10)	.52
Family member	17.0% (115)	18.6% (27)	18.2% (4)	.90
Minor	8.6% (58)	6.9% (10)	9.1% (2)	.79
LEO	5.0% (34)	7.6% (11)	18.2% (4)	.021
Hallucination	0.7% (5)	1.4% (2)	0.0% (0)	.68
Other/unclear	40.4% (273)	37.2% (54)	36.4% (8)	.73
Mass shooting threat	14.9% (143)	7.1% (17)	13.5% (5)	.042
Suicide risk	45.6% (438)	48.3% (115)	43.2% (16)	.71
Type of suicide risk
Ideation	148 (33.8%) 148	48 (41.7%)	37.5% (6)	.28
Threat	249 (56.9%) 249	71 (61.7%)	62.5% (10)	.6
Plan	248 (56.5%) 248	73 (63.5%)	62.5% (10)	.39
Aborted attempt	45 (10.3%) 45	12 (10.4%)	0 (0.0%)	.4
Attempt	34 (7.8%) 34	16 (13.9%)	25.0% (4)	.013
Currently or previously under DVRO^a	105 (10.9%) 105	25 (10.5%)	10.8% (4)	.983
Petitioner type
Law enforcement	58.2% (559)	53.36% (127)	75.7% (28)	.033
Family member	6.9% (66)	8.8% (21)	5.4% (2)	.53
Intimate partner	30.6% (294)	34.9% (83)	18.9% (7)	.12
Healthcare professional	0.6% (6)	0.4% (1)	0 (0.0%)	.84
Unknown	3.8% (36)	2.5% (6)	0 (0.0%)	.33
Who alerted law enforcement?^b
Family member	6.8% (38)	10.2% (13)	Suppressed	.34
Intimate partner	13.8% (77)	12.6% (16)	Suppressed	.58
Friend	1.4% (8)	Suppressed	Suppressed	.56
Self	6.6% (37)	5.5% (7)	Suppressed	.90
Healthcare professional	2.2% (12)	5.5% (7)	Suppressed	.052
Employee/employer/coworker	2.5% (14)	Suppressed	Suppressed	.87
Other	10.0% (56)	15 (11.8%)	Suppressed	.83
Unknown	56.9% (318)	50.4% (64)	50.0% (14)	.35

The DVRO history of a respondent was likely not captured systematically and is an underestimate as this variable was only coded when a DVRO was present and/or documented in the casefile.

For this category, counts under 5 are suppressed to protect confidentiality.

Over half (59.2%) of urban respondents were white and 37.7% were Black. The distribution in suburban and rural areas skewed more toward white respondents: in suburban areas, 79.0% of respondents were white, 18.5% were Black, and 0.8% were Hispanic; and in rural areas, 86.5% of respondents were white and 14% were Black. No respondents in rural areas were reported to be of any race other than white or Black. Chi-squared testing indicated a statistically significant difference in the distribution of respondent race by urbanicity category (P < .001, Table 1) before adjusting for racial demographics in each of these categories. When comparing these distributions to the racial breakdown of each urbanicity level (Supplemental Table 2), it appears that the proportion of racial groups in suburban and rural petitions roughly reflect the underlying racial demographics of these areas. However, the proportions of white and Black respondents are still higher than might be expected based on population, reflecting the absence of respondents identified as American Indian/Alaska Native, Asian, and Native Hawaiian/Pacific Islander, and the fewer than 5 Hispanic respondents in suburban and rural areas.

Nearly two-thirds (64.6%) of ERPO petitions for respondents in urban areas indicated that the respondent possessed firearms at the time of the ERPO, as did 76.5% of suburban respondents and 57.8% of rural respondents (Table 1). Logistic regression revealed higher odds of possession among suburban respondents compared to urban respondents (P = .001) and rural respondents (P = .013, Table 3).

Type of Threat

The majority of petitions in all 3 urbanicity categories suggested the respondent had threatened violence toward others (which does not necessarily entail the use of violence). In urban areas, 70.2% of respondents were reported to have threatened violence toward others (either alone or in addition to another behavior, including using violence or displaying suicidal risk); 60.9% of suburban respondents and 59.5% of rural respondents had threatened violence (Table 1). The odds of a respondent threatening violence were 34% lower in suburban than urban areas (P = .006), but other differences were not statistically significant (Table 3).

Targets of both threatened and actual violence differed as well. Family members and minors were more frequent targets of the use of violence in both rural and suburban petitions compared to urban petitions (urban vs rural P = .002; urban vs suburban P = .021, Table 3). Law enforcement officers were a more frequent target of threats of violence in rural petitions compared to urban petitions (P = .021, Table 3).

45.6% of urban petitions, 48.3% of suburban petitions, and 43.2% of rural petitions indicated that the respondent was at risk of suicide. Of the reported suicide risk behaviors present at the precipitating event, there was one statistically significant difference: rural petitions had nearly 4 times the odds of mentioning suicide attempts when compared to urban petitions (P = .023), and suburban petitions had roughly twice the odds of mentioning suicide attempts compared to urban petitions (P = .044).

Petitioner Characteristics

All 3 urbanicity categories included petitions from law enforcement officers, family members, and intimate partners. While 6 healthcare professionals filed petitions in urban areas (comprising 0.6% of overall petitions), there was only 1 suburban petition from a healthcare professional and none in rural ZIP codes. However, healthcare professionals did still play a role in petition initiation even when they did not petition independently. Among the law enforcement-initiated petitions, 5.5% in suburban areas were reportedly a result of a healthcare professional alerting law enforcement to a potential threat, compared to 2.2% of law enforcement-initiated petitions that stemmed from healthcare professionals in urban area (Table 1)).

While law enforcement initiated just over half of all petitions in urban and suburban areas (58.2% and 53.4% respectively), over 3-quarters of all rural petitions (75.7%) were initiated by law enforcement (Table 1). Rural petitions thus had over twice the odds of originating with law enforcement compared to urban petitions (P = .038) and nearly 3 times compared with suburban petitions (P = .013; Table 3).

Court Outcomes

72% of urban casefiles, 81% of suburban casefiles, and 84% of rural casefiles included a request for an interim order. The odds of interim ERPO initiation were higher in suburban than urban areas (P = .007), but the higher odds of interim ERPO initiation in rural areas compared to urban areas were not statistically significant (P = .119; Table 3). Interim orders tended to be granted: 94% of urban and suburban interim orders and 100% of rural interim orders were granted (Table 2).

Table 2.

Hearing Outcomes.

	Urban	Suburban	Rural	Chi squared P-value
Interim hearing requested^a	71.9% (680)	80.6% (191)	83.8% (31)	.009
Interim outcome
Granted	93.8% (638)	94.2% (180)	100.0% (31)	.36^b
Denied	6.2% (42)	5.8% (11)	0.0% (0)	.36^b
Ex parte hearing requested^c	93.9% (887)	91.5% (216)	91.9% (34)	.41
Ex parte outcome
Granted	73.7% (651)	71.6% (154)	64.7% (22)	.44
Denied	13.6% (120)	13.0% (28)	8.8% (3)	.72
Dismissed	11.7% (103)	13.5% (29)	14.7% (5)	.68
Stipulated to final	1.0% (9)	1.9% (4)	11.8% (4)	<.001
Final hearing requested^d	72.8% (695)	74.6% (176)	78.4% (29)	.186
Final ERPO outcome
Granted^e	66.3% (461)	67.6% (119)	75.9% (22)	.55
Denied	13.7% (95)	9.1% (16)	10.4% (3)	.24
Dismissed	20.0% (139)	23.3% (41)	13.8% (4)	.41

Data on whether an interim hearing occurred was missing for 16 petitions (1 suburban and 15 urban). This row reports hearings requested among the subset for which data was available.

The outcome did not vary for rural petitions.

Data on whether an ex parte petition occurred was missing for 18 petitions (1 suburban and 17 urban). This row reports hearings requested among the subset for which data was available.

Data on whether a final hearing occurred was missing for 8 petitions (all urban).

This category includes final ERPOs that were stipulated to as well as those that were granted after a hearing.

At the ex parte stage, the majority of petitions were granted in all jurisdictions, with a slightly lower proportion granted in rural areas compared to suburban and urban areas. In rural areas, stipulating to a final order at the ex parte stage was a more common outcome (11.8% of ex parte hearings vs 1.9% and 1.0% in suburban and urban areas respectively, Table 2). Overall, rural respondents had over 12 times the odds of having their final order granted at the ex parte stage compared to urban respondents (P < .001), although this should be interpreted with caution given that this data was only available for 29 rural petitions (Table 3).

Table 3.

Univariate Odds Ratios for Selected Characteristics.

	Suburban				Rural
	Ref: urban		Ref: rural		Ref: urban		Ref: suburban
Characteristic	OR (CI)	P	OR (CI)	P	OR (CI)	P	OR (CI)	P
Male	1.75 (1.07, 2.87)	.026	0.62 (0.14, 2.78)	.535	2.81 (0.66, 11.82)	.159	1.61 (0.36, 7.17)	.535
White race	2.60 (1.85, 3.65)	<.001	0.60 (0.22, 1.63)	.318	4.31 (1.66, 11.71)	.003	1.66 (0.61, 4.48)	.318
Possessed gun	1.78 (1.28, 2.47)	.001	2.48 (1.21, 5.07)	.013	0.72 (0.37, 1.40)	.329	0.40 (0.20, 0.83)	.013
Threatened violence	0.66 (0.49, 0.89)	.006	1.06 (0.52, 2.15)	.865	0.62 (0.32, 1.21)	.164	0.94 (0.46, 1.91)	.865
Threatened law enforcement	1.55 (0.76, 3.13)	.225	0.37 (0.11, 1.28)	.117	4.19 (1.34, 13.06)	.014	2.71 (0.78, 9.41)	.225
Used violence against a minor	2.16 (1.12, 4.16)	.021	0.45 (0.11, 1.92)	.281	5.59 (1.91, 16.38)	.002	2.22 (0.52, 9.48)	.281
Used violence against family	2.53 (1.01, 6.33)	.047	0.38 (0.12, 1.22)	.105	5.63 (1.43, 22.20)	.014	2.59 (0.82, 8.16)	.105
Mass shooting threat	0.63 (0.43, 0.92)	.017	0.65 (0.27, 1.53)	.319	0.97 (0.44, 2.15)	.938	1.55 (0.66, 3.66)	.319
Suicide attempt	1.92 (1.02, 3.62)	.044	0.48 (0.14, 1.69)	.256	3.96 (1.21, 12.95)	.023	2.06 (0.59, 7.19)	.256
Law enforcement petitioner	0.82 (0.62,1.09)	.180	0.37 (0.21, 0.96)	.013	2.24 (1.04, 4.79)	.038	2.72 (1.23, 6.01)	.013
Interim hearing held	1.62 (1.14, 2.31)	.007	0.80 (0.32, 2.04)	.646	2.02 (0.83, 4.90)	.119	1.24 (0.49, 3.16)	.646
Stipulated to final order at ex parte hearing	1.84 (0.56, 6.04)	.314	0.14 (0.03, 0.59)	.008	12.95 (3.77, 44.42)	<.001	7.03 (1.67, 29.62)	.008

Discussion

While circumstances and court outcomes differ by urbanicity, Maryland’s ERPO law is being used across all areas of the State. ERPO use during the study period was much higher in suburban areas and slightly higher in rural areas compared to urban areas. This finding adds important nuance to research and commentary on attitudes toward ERPO legislation in rural areas¹⁶; while public hesitance around ERPOs in less urban areas should continue to be explored, Maryland’s experience demonstrates that people living in rural and suburban areas are using ERPOs in response to various forms of gun violence risk in their communities.

Preliminary analysis appears to suggest that the distributions of respondent race—including the significantly higher proportion of white respondents in suburban and rural areas—roughly correspond to the demographics of urban, suburban, and rural areas in Maryland. However, both Black and white Marylanders may be slightly overrepresented as ERPO respondents relative to their share of the overall population in each level of urbanicity. Correspondingly, the number of Hispanic, Asian, and American Indian/Alaska Native respondents appears to be disproportionately low across each urbanicity category relative to their population. This pattern is similar to findings on respondent race in King County, Washington¹³ and warrants further research to understand these different levels of uptake.

Respondents across all urbanicity categories tended to exhibit similar forms of gun violence risk behaviors. Given that rates of both all-cause suicide and gun suicide are especially high in rural areas,^19,20 it is notable that a greater proportion of petitions in rural areas were in response to threats of violence against others. However, rural petitions did more frequently mention suicide attempts compared to urban and suburban areas, despite remaining similar on other behaviors related to suicide (ideation, threats, plans, and aborted attempts). The greater representation of a more severe and potentially fatal risk behavior may suggest a need to emphasize the value of ERPO and/or other interventions early in the course of suicide risk, particularly in rural areas.

This research also demonstrates that ERPOs are being used to intervene when mass shootings are threatened in urban, rural, and suburban areas alike, paralleling what is known about the distribution of mass shootings targeted at schools.^21,22 This also adds to the current literature that documents EPRO use in response to mass shootings without attention to area type.¹⁰ Further research is warranted to understand experiences of mass shooting threats among rural, suburban, and urban areas, including factors influencing ERPO initiation in response.

Another finding that held across urbanicity was that intimate partners were a primary target of both reported use and threats of violence in the precipitating event, with at least 40% of petitions in each urbanicity category mentioning a threat of violence against an intimate partner. In Maryland, DVROs offer an existing avenue for firearm dispossession, as those subject to DVROs are not legally permitted to own firearms while the order is in effect. The presence of ERPO in cases of intimate partner violence may thus indicate ERPO is being used to fill a gap in implementation of DVRO-related firearm relinquishment, which has been documented in DVRO research in general³³ and in rural areas specifically.²⁸ This finding may also point to the role that ERPO plays in allowing law enforcement to petition for gun dispossession in cases of apparent intimate partner violence even if the target of that violence—the only authorized petitioner for a DVRO—declines to pursue a DVRO.

The targets of reported threats and/or use of violence were overall similar, with the exception that rural and suburban petitions were more often precipitated by the use of violence against family members and minors. Rural petitions also had higher odds of mentioning threats against law enforcement officers, which may be related to the relatively higher proportion of law enforcement petitioners in these areas.

Across all levels of urbanicity, law enforcement officers were the most common petitioners, paralleling existing research on ERPOs. However, Maryland has a notably higher overall proportion of civilian petitioners (just over 40%) compared to many other states with ERPO for which data is available.⁹ Rural areas had the highest proportion of law enforcement-initiated petitions, followed by urban and suburban areas, and rural petitions were overall twice as likely to originate from law enforcement compared to urban petitions, suggesting implementation of ERPO in rural settings may have been less effective at empowering civilian petitioners than law enforcement petitioners. Existing research on DVRO implementation offers some potential explanations for lower use in rural areas, including the geographic dispersal of rural areas, which may limit access to courts, as well as limited court personnel available to assist civilian petitioners.^28,29

It is also important to consider that civilians may be less willing to use ERPO in rural areas due to heightened mistrust of firearm removal processes as reflected in DVRO research²⁹ and preliminary research on California’s ERPO law.¹⁶ Further research is needed to elucidate whether similar factors are at play for rural civilian ERPO petitioners and would-be petitioners.

Across all jurisdictions, the majority of ERPOs were granted at all stages, with the highest rates in rural areas. There were notably more stipulations to final ERPOs at the ex parte stage in rural areas than in suburban or urban areas; however, these results should be interpreted with caution given the small number of rural petitions for which data was available and the fact that court outcomes may not have been systematically captured in casefiles.³⁰ In general, reasons that ERPO respondents stipulate are not well understood, but may potentially reflect an agreement that temporary firearm relinquishment would benefit them and/or those around them, particularly for those experiencing suicidality; the higher rate of both stipulation and precipitating suicide attempts in rural areas lends some credence to this hypothesis. More research is needed to determine how and why pathways to a final ERPO differ by urbanicity, and why respondents in general choose to stipulate.

These results also confirm the importance of disaggregating urbanicity beyond a dichotomous “urban” and “non-urban” classification in research on ERPOs, and potentially in gun violence research more broadly. While there were clear demographic similarities between respondents in suburban and rural areas that contrasted with urban respondents, there were also important differences between urban/suburban and rural categories when it came to the proportion of law enforcement petitioners and the trajectory of ex parte petitions. Thus, urbanicity—disaggregated and examined by category rather than as a binary classification—is an important factor to consider in further research on ERPOs and court outcomes related to gun violence.

Limitations

This study has several limitations. Firstly, this analysis relied on data available in abstracted ERPO casefiles, some of which may have been incomplete and/or inaccurate. Additionally, while ZIP code data abstracted from petitions allowed a more granular look at geographic trends than would have been possible with county-level aggregation, ZCTA data from the ACS may differ slightly from actual ZIP codes in terms of geographic boundaries, particularly because ZIP code designations can change from year to year while ZCTAs tend to be reassessed once every 10 years.³³

Only 5 Maryland ZIP codes listed on petitions could not be matched to a ZCTA, giving us confidence that the ZCTAs adequately captured most ZIP codes in the state. Furthermore, the ZIP code data in this study represented the respondent’s home address, not necessarily the location where a threat occurred. This may pose a validity issue if threats and/or hearings occurred in areas of differing urbanicity from the respondent’s home address. That said, this is the first analysis to provide a detailed analysis of ERPO use across area types using ZIP codes and ZCTA to define those areas.

This study was also limited by the number of ERPO petitions that occurred during the period examined, which reduced the statistical power to detect between-category differences. While there were 1236 total petitions in the final dataset, only 37 originated in areas designated as rural. This meant that even apparently large effect sizes between rural petitions and other categories examined often did not reach statistical significance. Many of the statistically significant differences that could be detected were between suburban and urban areas due to the larger sample of suburban petitions. Regardless, the trends in this dataset do suggest that further exploration with larger datasets is warranted. That this research was limited to Maryland may also limit generalizability, particularly because Maryland has a more expansive definition of eligible petitioners than many other states, and the distribution of ERPO uptake among rural, suburban, and urban areas is different from early reports of other states.^16,18 However, examining Maryland’s ERPO experience has also allowed this research to look in more detail at how petitioner types differ by urbanicity, which would not have been possible in states that restrict petitioner eligibility to law enforcement. Furthermore, the rate of ERPO use in rural areas (while translating into only several dozen petitions) also allows for some early insights into differences among area types. Finally, the statistical tests used in this research did not control for the influence of potentially confounding characteristics; chi-squared tests and univariate logistic regressions assess the relationship between urbanicity and individual variables of interest. Factors such as age, sex, race, and geographic proximity of courts likely play a role in ERPO use and outcomes, and should be adjusted for in future analyses with larger sample sizes. Gun violence incidence and population size are also important factors to include in future models. That said, the role of Maryland as an early statewide implementer holds insights for states where ERPO uptake is low and the picture of what more robust implementation would look like is unclear.

Conclusion

Despite some evidence suggesting unfavorable attitudes toward ERPOs in rural areas, this research suggests that uptake of ERPOs in Maryland is by no means limited to urban areas. In Maryland, ERPOs are being used in rural and suburban areas at higher rates than in urban areas, with the intent of addressing a wide range of potentially lethal threats including suicide, interpersonal violence, and mass shootings. While clear differences exist in ERPO use between areas of differing urbanicity, it is also apparent that across all regions in Maryland, ERPOs are being used by a wide range of people—from loved ones to law enforcement to clinicians—to remove access to lethal weapons before threats can escalate into deadly acts of violence. Implementation efforts must focus on ensuring that those in rural areas have equal opportunities and resources for petitioning the court to remove firearms during the times of heightened risk so that all can equally benefit from the lifesaving potential of ERPOs.

Supplemental Material

sj-docx-1-inq-10.1177_00469580251370930 – Supplemental material for Variation in Extreme Risk Protection Order Use Among Urban, Suburban, and Rural ZIP Codes in Maryland: A Descriptive Study

Supplemental material, sj-docx-1-inq-10.1177_00469580251370930 for Variation in Extreme Risk Protection Order Use Among Urban, Suburban, and Rural ZIP Codes in Maryland: A Descriptive Study by Mia Aassar, Elise Omaki, Lisa Geller, April M. Zeoli and Shannon Frattaroli in INQUIRY: The Journal of Health Care Organization, Provision, and Financing

Footnotes

Acknowledgements

The authors would like to than the National Collaborative on Gun Violence Research for supporting this work, as well as the coders of the original six-state ERPO study for creating the dataset that enabled this study.

ORCID iDs

Mia Aassar

Lisa Geller

Ethical Considerations

Ethics approval was not obtained because this study used de-identified data.

Author Contributions

MA conceptualized the study, conducted data analysis, and drafted the manuscript. SF oversaw study design and data analysis, contributed to interpretation of results, and provided substantive edits to the manuscript. EO provided feedback on analytical approach and interpretation of results, and provided substantive edits to the manuscript. LG and AZ provided substantive edits to the manuscript.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The ERPO casefile coding that created the data used for this study was supported by a grant from the National Collaborative on Gun Violence Research. The views expressed in this manuscript are the authors’ and do not necessarily reflect the view of the National Collaborative on Gun Violence Research.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The data used in this study are protected by statute in Maryland and are not available for public access.

Supplemental Material

Supplemental material for this article is available online.

References

Centers for Disease Control and Prevention, National Centers for Injury Prevention and Control. Web-based Injury Statistics Query and Reporting System (WISQARS). n.d. Accessed February 27 2025. www.cdc.gov/injury/wisqars

The Public Health Approach to Gun Violence Prevention . Educational Fund to Stop Gun Violence; Published November 2025.

Rudd

Berman

Joiner

Jr , et al Warning signs for suicide: theory, research, and clinical applications. Suicide Life Threat Behav. 2006;36(3):255-262. doi:10.1521/suli.2006.36.3.255

Miller

McCaw

Intimate partner violence. N Engl J Med. 2019;380(9):850-857. doi:10.1056/NEJMra1807166

Twelve Years of Mass Shootings in the United States . Everytown for Gun Safety; Published November 2020.

National ERPO Resource Center (Johns Hopkins Center for Gun Violence Solutions). n.d. Accessed April 6, 2024. https://erpo.org/

Maryland Code. Public Safety: Title 5, Firearms—Subtitle 6, Extreme Risk Protective Orders, § 5-601(e)(2)(i) (2018).

Rakshe

Valek

Teichman

Freeman

DeFrancesco

Carlson

KF.

Five years of extreme risk protection orders in Oregon: a descriptive analysis. Psychol Rep. Published online April 26, 2024. doi:10.1177/00332941241248599

Zeoli

Frattaroli

Aassar

Cooper

Omaki

Extreme risk protection orders in the United States: a systematic review of the research. Ann Rev Crim. 2025;8:485-504. doi:10.1146/annurev-criminol-111523-122602

10.

Zeoli

Frattaroli

Barnard

, et al Extreme risk protection orders in response to threats of multiple victim/mass shooting in six U.S. states: a descriptive study. Prev Med. 2022;165(Pt A):107304. doi:10.1016/j.ypmed.2022.107304

11.

Valek

Teichman

Rakshe

DeFrancesco

Carlson

KF.

Use of Oregon's extreme risk protection order law to address risk of firearm suicide. Inj Prev. Published online March 23, 2025. doi:10.1136/ip-2024-045581

12.

Pear

Pallin

Schleimer

, et al Gun violence restraining orders in California, 2016-2018: case details and respondent mortality. Inj Prev. 2022;28(5):465-471. doi:10.1136/injuryprev-2022-044544

13.

Frattaroli

Omaki

Molocznik

, et al Extreme risk protection orders in King County, Washington: the epidemiology of dangerous behaviors and an intervention response. Inj Epidemiol. 2020;7(1):44. doi:10.1186/s40621-020-00270-1

14.

Zeoli

Paruk

Branas

, et al Use of extreme risk protection orders to reduce gun violence in Oregon. Criminol Public Policy. 2021;20(2):243-261.

15.

Swanson

Zeoli

Frattaroli

, et al Suicide prevention effects of extreme risk protection order laws in four states. J Am Acad Psychiatry Law. 2024;52(3):327-337. doi:10.29158/JAAPL.240056-24

16.

Pear

Schleimer

Tomsich

, et al Implementation and perceived effectiveness of gun violence restraining orders in California: a qualitative evaluation. PLoS One. 2021;16(10):e0258547. doi:10.1371/journal.pone.0258547

17.

Prater

Rooney

Bowen

, et al Civilian petitioners and extreme risk protection orders in the state of Washington. Psychiatr Serv. 2022;73(11):1263-1269. doi:10.1176/appi.ps.202100636

18.

Rooney

Conrick

Bellenger

, et al Understanding the process, context, and characteristics of extreme risk protection orders: a statewide study. J Health Care Poor Underserved. 2021;32(4):2125-2142. doi:10.1353/hpu.2021.0186

19.

Centers for Disease Control and Prevention, National Centers for Injury Prevention and Control. Web-based Injury Statistics Query and Reporting System (WISQARS) [online]. n.d. Accessed February 24, 2025. www.cdc.gov/injury/wisqars

20.

Nowicki

JM.

K-12 Education: Characteristics of School Shootings. Report to Congressional Requesters. GAO-20-455. US Government Accountability Office. Published June 2020.

21.

Livingston

Rossheim

Hall

KS.

A descriptive analysis of school and school shooter characteristics and the severity of school shootings in the United States, 1999-2018. J Adolesc Health. 2019;64(6):797-799. doi:10.1016/j.jadohealth.2018.12.006

22.

Parker

Horowitz

Igielnik

Oliphant

Brown

America’s Complex Relationship with Guns. Pew Research Center. Published June 22, 2017. Accessed February 2, 2023. https://www.pewresearch.org/social-trends/2017/06/22/americas-complex-relationship-with-guns/

23.

Azrael

Hepburn

Hemenway

Miller

The stock and flow of U.S. firearms: results from the 2015 National firearms survey. RSF. 2017;3(5):38-57.

24.

Curtin

Brown

Jordan

ME.

Suicide rates for the three leading methods by race and ethnicity: United States, 2000-2020. NCHS Data Brief. 2022;450:1-8.

25.

Kivisto

Phalen

PL.

Effects of risk-based firearm seizure laws in Connecticut and Indiana on suicide rates, 1981-2015. Psychiatr Serv. 2018;69(8):855-862. doi:10.1176/appi.ps.201700250

26.

Office of Community Oriented Policing Services. Conversations with Rural Law Enforcement Leaders: Volume 1. Rural Law Enforcement. Office of Community Oriented Policing Services. Published March 2020. Accessed February 23, 2025. https://www.policinginstitute.org/publication/conversations-with-rural-law-enforcement-leaders-volume-i/

27.

Lynch

Logan

TK.

Implementing domestic violence gun confiscation policy in rural and urban communities: assessing the perceived risk, benefits, and barriers. J Interpers Violence. 2020;35(21-22):4913-4939. doi:10.1177/0886260517719081

28.

Logan

Shannon

Walker

Protective orders in rural and urban areas: a multiple perspective study. Violence Against Women. 2005;11(7):876-911. doi:10.1177/1077801205276985

29.

Zeoli

Molocznik

Paruk

, et al A multi-state evaluation of extreme risk protection orders: a research protocol. Inj Epidemiol. 2024;11(1):49. doi:10.1186/s40621-024-00535-z

30.

2010 Census Urban and Rural Classification and Urban Area Criteria. United States Census Bureau. 2010. Accessed April 24, 2023. https://www.census.gov/programs-surveys/geography/guidance/geo-areas/urban-rural/2010-urban-rural.html

31.

Rural-Urban Area Commuting Codes: Documentation. United States Department of Agriculture. n.d. Accessed April 24, 2023. https://www.ers.usda.gov/data-products/rural-urban-commuting-area-codes/documentation/

32.

Webster

Frattaroli

Vernick

O’Sullivan

Roehl

Campbell

JC.

Women with protective orders report failure to remove firearms from their abusive partners: results from an exploratory study. J Womens Health. 2010;19(1):93-98. doi:10.1089/jwh.2007.0530

33.

Grubesic

Matisziw

TC.

On the use of ZIP codes and ZIP code tabulation areas (ZCTAs) for the spatial analysis of epidemiological data. Int J Health Geogr. 2006;5:58. doi:10.1186/1476-072X-5-58

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.01 MB