Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Aug 25.
Published before final editing as: Psychol Crime Law. 2023 Jun 24:10.1080/1068316X.2023.2227309. doi: 10.1080/1068316X.2023.2227309

Risk and Contextual Factors Associated with Legal Intervention Injury and Hospital Outcomes Among Trauma Patients in Pennsylvania

Zhenyu Zhang 1, Samantha Vervoordt 1, Frank Hillary 1
PMCID: PMC12373391  NIHMSID: NIHMS1912094  PMID: 40857427

Abstract

Police violence has been described as an ongoing public health issue with profound impacts on individuals and communities. We leveraged a state-wide trauma registry to identify patients injured through legal intervention between 1989 to 2019 to examine individual characteristics associated with injuries, contextual factors during legal intervention, and clinical outcomes of injury to capture the nature of legal intervention injury. A total of 1098 patients injured through legal intervention were identified and 1095 patients from the PTOS dataset were matched for comparison with respect to injury and demographic factors. Patients injured through legal intervention were more likely to be male, identify as Black, had pre-existing psychiatric and substance use disorders (PECs), be intoxicated at admission, and have lower socioeconomic status. For White patients, PECs, intoxication, and lower SES were risk factors for worse clinical outcomes. Time of day and location of event also played a role in understanding this type of injury. Police violence disproportionately impacts racial minority groups as well as other disadvantaged and marginalized communities. An intersectional and social constructionist approach may help to understand contextual risk factors that can in turn aid in efforts to reduce law enforcement-related deaths and injuries.

Keywords: legal intervention injury, risk factors, racial disparities

Roughly 61.5 million U.S. residents aged 16 or older, about 24% of the U.S. population, had at least one interaction with police during 2018, which was 21% more than 2015 according to the most recent report on the Police-Public Contact Survey (PPCS) conducted by the United States Department of Justice (Harrell & Davis, 2020). The data on injuries and deaths caused as a result of legal intervention, defined as injuries caused by police or other law enforcement personnel in the course of performing legal actions, from the Centers of Disease Control and Prevention (CDC) show that 9,841 civilians died between 2001 to 2020, an average of 492 civilian deaths per year, with mortality estimated around 0.16 per 100,000 individuals in the United States (CDC, National Center for Injury Prevention and Control, 2022). Those that suffered fatal injuries following these encounters were predominantly males (95.74%) between the ages of 20 to 49 (77.15%) and were overwhelmingly injured by firearms (85.3%). There were also higher death rates among Black (0.3 per 100,000) and American Indian individuals (0.29 per 100,000) compared to White individuals (0.14 per 100,000) (data for 2020, United State Census Bureau, 2021)). In addition, an estimated 1,607,864 civilians received treatment in general U.S. hospital emergency rooms between 2001 and 2020 for non-fatal injuries resulting from legal intervention (CDC, 2022), with the age-adjusted incidence rates (a statistical measure that takes into account differences in age distribution between different populations) increasing from 21.9 in 2001 to 26.5 in 2020 per 100,000 U.S. population. Even with formal attempts to monitor the frequency of police-related fatalities in the United States (see Washington Post Police Shooting Database), epidemiologic data are still incomplete, and trends in nonfatal injuries through police violence, which outweigh deaths by a large margin, are poorly understood (Kaufman et al., 2017).

Based on findings from the four PPCS data collections from 2002–11 (Hyland et al., 2015), among those who experienced force through legal intervention during their most recent contact, most perceived both verbal (71.4%) and physical (75.0%) force as excessive, and 86.7% of these individuals reported this force as perceived police misconduct. Law enforcement policies and practices have been recognized as social determinants of health, and police violence has been cited as an ongoing public health issue with profound impacts on individuals and communities (Cooper, & Fullilove, 2016; Cooper et al., 2004; DeVylder, 2021).

Research has shown that Black, Native American, and Hispanic populations are more likely to be targeted by police using stop and frisk and are more likely to experience excessive force in their interactions with police when compared to their non-Hispanic White counterparts (Krieger et al., 2015; Miller et al., 2017; Mooney et al., 2018). Racialized police violence has increased mistrust in police and stresses among communities of color, which could further jeopardize the police-citizen cooperation in solving crimes and reducing community problems (Flexon et al., 2019).

The intersection of race and gender increases the disparity in police violence in some groups. Men, especially Black and Latino men, are more likely to have encounters with legal intervention personnel and injuries from those encounters (Edwards et al., 2018; Gilbert & Ray, 2016; Loder & Leiser, 2021), and sexual minorities are also at greater risks of police violence than heterosexual individuals (DeVylder et al., 2018). Moreover, “place” can play an important role in the disparity of police violence. In the perspective of social geography, place is a location or area granted special meanings by individuals and/or groups (Fyfe, 1991; Yarwood, 2007). In this sense, place is not only an objective space, but also a peopled space that has subjective meaning constructed by individuals and/or groups. When there are conflicts between civilians and police officers on the definitions of local places, there can be conditions where police perpetrating violence, such that individuals living in disadvantaged neighborhood in urban areas are at greater risks of police violence resulting in injury and fatalities (Cooper et al., 2004; Schwartz & Jahn, 2020).

Other factors, such as psychiatric disorders and alcohol and substance use should also be considered when examining police violence (Cooper, 2015; Holloway-Beth et al., 2016). One study from Laniyonu and Goff (2021) using a nonrandom sample of nine U.S. police departments from 2011 to 2017 found that police in those departments were 11.6 times more likely to use force against those with a serious mental illness, and those individuals were 10.7 times more likely to experience injury as a result. These relative risk ratios were several times larger than racial and ethnic disparities in police violence in the same cities (Laniyonu & Goff, 2021). Previous research has found that substance users, especially injection drug users, reported higher rates of physical, psychological, and sexual violence and neglect by police (Cooper et al., 2004; Park et al., 2019). Alcohol intoxication is also positively associated with risky behaviors and violent crimes (Kavanagh, 1997; McClelland & Teplin, 2001), which is an important factor in police response. There are other factors contributing to police officers’ choice to use force during their encounters with civilians, including a lack of clear legal and policy guidance on use of force during intervention (Obasogie & Newman, 2017), availability and training on the use of non-lethal weapons such as Tasers or Pepper spray (MacDonald, et al., 2009), and personality factors (Ellrich & Baier, 2016; Waegel, 1984).

These findings all suggest that an intersectional approach could further illuminate risk factors contributing to increasing concern regarding the force used during police interventions. Intersectionality refers to a theoretical framework for examining how different social identities, such as race, gender, sexual orientation, socioeconomic status (SES), and disability, intersect at the micro level of an individual’s experience in order to reflect how privilege and oppression, such as sexism, racism, and classism, are intertwined at the macro level of social structures (Bowleg, 2021 & 2012; Gilbert & Ray, 2016). Intersectionality provides researchers with a framework to conceptualize and analyze disparities and social inequalities in health. Instead of focusing on a single category (i.e., race or gender), the intersectional approach can help examine police violence in complex and multidimensional ways mirroring the experiences of the populations for whom encounters of police violence are most disproportionate. Intersectional approach can also provide a theoretical lens for interpreting novel or unexpected findings.

Despite the available data, which demonstrates a clear and ongoing public health concern, there remains a paucity of research on legal intervention traumatic injury. No reliable official data exist on the number of individuals killed or injured by the police or other details in the legal intervention process in the United States, in part due to long-standing resistance from police departments (Krieger et al., 2015), making it difficult to study this phenomenon. Even though journalists have been monitoring deaths involving police (see Washington Post Police Shooting Database), Epidemiologic data are still incomplete, and trends in nonfatal injuries through police violence, which outweigh deaths by a large margin, are poorly understood (Kaufman et al., 2017). The studies described above are limited to a description of basic and incomplete demographic characteristics of injured civilians via self-reports and medical records. Only one study described risk factors leading to an injury, the types of injuries suffered, and clinical outcomes based on medical record databases for patients treated in all hospitals and trauma units in Illinois between 2000 and 2009 (Holloway-Beth et al., 2016). However, this study was unable to capture race and ethnicity as well as other important contextual factors, such as injury place and time, which are all important variables that could help inform policy and clinical recommendations for best practices.

As shown from the described literature, apart from descriptive studies, it can be exceedingly difficult to take a comparative approach to social problems such as these. As such, methodologies such as propensity score matching may provide novel insights about demographic variables that are most important to consider in the context of legal intervention. Methodologies such as propensity score matching would allow researchers to make comparisons between those injured through legal intervention and those injured via unrelated mechanisms, rather than through comparison to the general population, which may not be representative of those that present with injuries. Although such methodologies have their own limitations, these might provide greater nuance to descriptive studies.

In the current study, we used a medical administrative database in Pennsylvania to capture the nature of legal intervention injury, individual characteristics associated with injuries, contextual factors during legal intervention, and clinical outcomes of injury. The goal of this study was to understand the risk factors associated with traumatic injury due to legal intervention and hospital outcomes associated with this type of injury to inform policies and police practices to reduce police violence and related deaths. In order to capture the unique demographic nature of the legal intervention injury group compared to overall trauma admissions in Pennsylvania, we leveraged the use of a matched sample using propensity score matching, described in greater detail below. By creating a matched sample, we hope to highlight potential demographic differences between those injured from legal intervention in Pennsylvania and the general trauma population. We proposed the following hypotheses, informed by the literature detailed above:

  1. Demographic Variables:
    1. The proportion of men injured by legal intervention will be greater than the proportion of men in a matched sample.
    2. The proportion of patients of color injured by legal intervention will be greater than the proportion of patients of color in a matched sample.
    3. The proportion of patients with lower SES will be greater than the proportion of patients with lower SES in a matched sample.
  2. Pre-existing Conditions and Intoxication:
    1. The proportion of patients with pre-existing psychiatric and/or substance pre-existing conditions (PECs) injured by legal intervention will be greater than those patients in a matched sample.
    2. The proportion of patients with positive alcohol and/or drug tests at admission injured by legal intervention will be greater than those patients in a matched sample.

  3. For patients injured through legal intervention, patients of color will have more firearm injuries, and more severe injury outcomes across all variables assessed, described in greater detail under Methods, compared to their White counterparts, AND

  4. These results will further differ by pre-existing conditions and intoxication at admission. Specifically, patients of color with PECs or who were intoxicated at admission will have greater rates of injury by firearms, and more severe injury outcomes across all variables assessed compared with those with no PECs or not intoxicated.

Supplemental analyses were also conducted examining the impact of race, PECs and intoxication on admission time, injury place as well as discharge destination to inform the observed results.

Method

Data

The current study sample was collected through the Pennsylvania Trauma Outcome Study (PTOS). The PTOS represents a state-wide effort operated by the Pennsylvania Trauma Systems Foundation (PTSF) to ensure the quality of trauma care offered by the registry’s participating hospitals and provide consistent data for system and clinical research for more than three decades (1986 - present). The PTSF included de-identified data with a total of 939,123 cases as of 2019. Obtaining and using the data was approved from PTSF prior to conducting this study. Additional information regarding trauma registry data entry, variable definitions, pre-existing condition categories are presented in the PTSF Operational Manual (2022).

Inclusion Criteria

Patients injured due to legal intervention were identified using the International Classification of Diseases, 9th edition (ICD-9) external cause of injury codes (E-CODES) ranging from E970 to E977, as well as ICD 10th edition (ICD-10) codes containing Y35, which are defined as legal intervention cases. Cases involving legal executions (E978) and terrorism (E979) were excluded. In the United States, diagnoses and procedures related to hospital utilization are officially coded using the ICD clinical modification (ICD-CM) system (Hedegaard & Johnson, 2019). ICD-9-CM is being replaced by ICD-10-CM as of October 1, 2015. ICD-9 legal intervention injury codes (E970–977) are comparable to ICD-10 codes (Y35), which includes injury due to legal intervention by firearms, explosives, gas, blunt object, cutting and piercing instrument, other specified means, unspecified means, and late effects of injuries due to legal intervention. There was no rationale or justification provided for the use of force for this type of injury based on ICD-9, ICD-10 codes, or the PTOS manual.

Variables of Interest

In addition to external cause of injury ICD-9 and ICD-10 codes, basic demographic variables, including age, biological sex, race, and ethnicity were included. Neurotrauma, including head and spinal trauma at admission (Yes/No), the Glasgow Coma Scale (GCS), Injury Severity Score (ISS) and days admitted in Intensive Care Unit (ICU) were used as measures of injury severity. Neurotrauma and injury severity were included due to their associations with more chronic consequences that may persist beyond discharge (Widerstrom-Noga et al., 1999; Wilson et al., 2017).

To capture contextual factors contributing to injury, place of injury, time at admission, blood alcohol levels (Blood Alcohol Content (BAC) on admission), and toxicology results on admission were included. Primary Payor Class was included to indirectly examine patients’ SES. We identified patients that paid through Medicaid or that did not use insurance (self-pay) as those with lower SES, and those that paid through commercial insurance or third parties as a higher SES group. We were particularly interested in PECs and intoxication status at injury that may affect judgment, communication, and motor mobility. Therefore, we included patients’ medical records of PECs. We were specifically interested in psychiatric disorders, alcohol abuse, and substance use disorders. In addition, discharge destination and discharge status (alive/died) were included to study injury and hospital outcomes.

Data Analysis

We used R version 4.1.2 (2021) in RStudio for all data inspections, propensity score matching, analyses, and visualizations. Two sample Z tests of proportions were used to determine whether the two populations of sex (men vs. women), race (i.e., Black vs. White; Asian vs. White), pre-existing conditions (Yes vs. No), alcohol and drug intoxication (Yes vs. No), and payor class (Medicaid/self-pay vs. commercial insurance/third party) differed significantly. Pearson’s Chi-squared test with or without Yates’ Continuity Correction were used to determine if race and other selected factors were related or not (Fisher’s exact test were implemented when the number of patients in one group was less than 5). Additionally, Welch’s two-sample t-tests were implemented to compare means of GCS, ISS scores and ICU days among selected groups.

Propensity Score Matching

We identified 1122 patients injured through legal intervention from 1989 to 2019 in the provided data. 24 were excluded due to their identification as law enforcement officials. This led to 1098 patients injured through legal intervention included in the current study. In an effort to make comparisons in line with our hypotheses, propensity score matching was used to create a comparison sample matched by age and admission year. Age was chosen due to its relationship to trauma outcomes and police involvement (Holloway-Beth et al., 2016), and admission year was chosen to account for potential policy changes, treatment advances, and systematic changes to PTOS manualized data collection. We used 1:1 nearest neighbor propensity score matching without replacement. This matching process was done using the MatchIt package (Ho et al., 2011) and yielded adequate balance. After matching, standardized mean differences were below 0.1, which indicated adequate balance between samples for age and injury year. Because there was missing age data for three subjects in the in the legal intervention injury group, 1095 patients were matched. The final samples contained 1095 unique patients in each sample.

Results

Descriptive Statistics for Legal Intervention Injury Group and Matching Injury Group

Descriptive characteristics and comparison statistics for patients in both groups are provided in Table 1. Notably, there were significantly more men (96.35% vs. 70.59%, p < .001, φ = .35), patients that identified as Black (50.78% vs. 16.99%, p < .001, φ = .36), and patients with lower SES based on the described insurance proxy measure (58.54% vs. 31.33%, p < .001, φ = .30) in the legal intervention injury group than those in the matched injury group. All effect sizes were in the medium range. Thus, our hypotheses regarding demographic variables were supported.

Table 1.

Demographics, injury characteristics, pre-existing conditions, and hospital outcomes

Legal Intervention Injury (Total N = 1095) PTOS Matched Sample (Total N = 1095) χ2 / t p-value Effect Size
(φ, d)
M Age in years 32.32 32.32
Sex
 Male 1055 (96.35%) 773 (70.59%) 263.19 < .001 *** .35
Race Data Missing 46 (4.20%) 44 (4.02%) .05 .832
 White 447 (40.82%) 829 (75.71%) 273.96 < .001*** .35
 Black 556 (50.78%) 186 (16.99) 278.90 < .001*** .36
 Asian 5 (.46%) 6 (.55%) Not calculated; low frequencies -
 Other Race 41 (3.74%) 30 (2.74%) 1.75 .19
Ethnicity Data Missing* 605 (55.1%) 609 (55.62%) .06 .81
 Hispanic/Latino 48 (4.37%) 36 (3.29%) 1.73 .19
Primary Payor Class Data Missing 166 (15.16%) 164 (14.98%) .01 .91
 Medicaid Indemnity 235 (21.46%) 126 (11.51%) 39.35 < .001*** .13
 Medicaid Managed Care 173 (15.80%) 89 (8.13%) 30.56 < .001*** .12
 Self-Pay 233 (21.28%) 128 (11.69%) 36.56 < .001*** .13
 Commercial Insurer Indemnity 68 (6.21%) 228 (20.82%) 99.94 < .001*** .21
 Commercial Insurer Managed Care 124 (11.32%) 117 (10.68%) .23 .63
 Other Third Party 29 (2.65%) 198 (18.08%) 140.24 < .001*** .25
PEC Data Missing 262 (23.93%) 200 (18.26) 10.64 < .001*** .07
 No PEC 446 (40.73%) 635 (57.99%) 65.22 < .001*** .17
 Psychiatric Disorders 139 (12.69%) 109 (9.95%) 4.09 .04* .04
 Alcohol or Substance Use 166 (15.16%) 114 (10.41) 11.07 < .001*** .07
 Comorbid Psychiatric and Substance Use Disorders 82 (7.49%) 37 (3.38%) 17.99 < .001*** .09
BAC Data Missing 375 (34.25%) 313 (28.58) 8.17 .004** .06
 BAC Legally Intoxicated 219 (20%) 361 (32.97) 47.28 < .001*** .15
Drug Screen Data Missing 97 (8.86%) 119 (10.87%) 2.49 .16
 Not Tested for Drug Use 447 (40.82%) 485 (44.29%) 2.70 .10
 Drug Screen Positive 349 (31.87) 263 (24.02%) 16.75 < .001*** .09
Injury Severity
 Neurotrauma 435 (39.73%) 1095 (100%) 944.19 < .001*** .66
 M GCS Score 12.10 12.77 3.15 .002** .15
 M ISS 13.86 14.79 1.74 .08
 M ICU Days 2.23 2.35 .46 .64
 Deceased in hospital 132 (12.05%) 49 (4.47%) 41.49 < .001*** .14
Open in a new tab

Note: φ = phi coefficient. d = Cohen’s d. M = mean. T-tests implemented Welch’s Two Sample t-test. χ2 = Pearson’s chi-squared test.

*

Due to the proportions of missingness for the Ethnicity variable, estimates for patients of Hispanic/Latino ethnicity are likely underestimated. GCS = Glasgow Coma Scale. ISS = Injury Severity Score.

*

p < .05,

**

p < .01,

***

p < .001. t-tests, rather than chi-square, were used to compare injury severity characteristics.

Further, there were significantly more patients with a record of psychiatric disorder (12.69% vs. 9.95%, p = .043, φ = .04), alcohol or substance use disorder (15.16% vs. 10.41%, p < .001, φ = .07), and comorbid psychiatric and substance use disorders (7.49% vs. 3.38%, p < .001, φ = .09) in the legal intervention injury group compared to the matched injury group. Moreover, more patients injured through legal intervention were tested positive for drug use (31.87% vs. 24.02%, p < .001, φ = .09), but more patients in the matched injury group were legally intoxicated based on their BAC (32.97% vs. 20%, p < .001, φ = .15). Thus, our hypotheses regarding PECs and intoxication were mostly supported.

Specific injury mechanisms were presented in the Table 2. Notably, more than half of patients in the legal intervention injury group were injured by firearms (55.43%), followed by manhandling or blow (29.22%), while most patients in the matched injury group were injured by accidents (50.32% transport accidents; 21.19% accidental falls; 3.2% other accidents), followed by homicide and general assault (23.01%). Notably, even though all patients in the matched injury group sustained trauma that led to admission to a trauma center, patients injured through legal intervention injury received worse GCS scores and more of them died while in the hospital (12.05% vs. 4.47%, p < .001, φ = .14).

Table 2.

Specific Injury Mechanisms

Specific Injury Mechanism Legal Intervention Injury
 Firearms 607 (55.43%)
 Manhandled or blow 320 (29.22%)
 Blunt object 99 (9.04%)
 Unspecified means 60 (5.48%)
 Cutting and piercing instrument 7 (.64%)
 Explosives 2 (.18%)
 Other specified means 0
 Late effects of injuries 0
 Transport Accidents -
 Accidental falls -
 Other accidents -
 Suicide And Self-Inflicted Injury -
 Homicide and general assault -
Open in a new tab

The Impact of Race on Legal Intervention Injury Type, Severity and Hospital Outcome

Due to the underrepresentation of Asian (.46%) and other racial groups (3.73%), the examination of our fifth and sixth hypotheses were focused on Black (50.82%) and White (40.80%) patients within the legal intervention group. With respect to firearm injury, significantly more Black patients were injured by firearms than White patients in the current sample (63.08% vs. 46.65%; OR = 1.95; χ2(1, 1005) = 27.17, p < .001, φ = .16).

However, the Black patients were not more likely than the White cohort to experience neurotrauma (37.81% vs. 41.74%; χ2(1, 398) = 1.45, p = .23, φ = .06). The average GCS score of Black patients (N = 433, M = 12.11, SD = 4.57) was not significantly different from the average GCS score of White patients on admission (N = 369, M = 12.21, SD = 4.71), t(711.7) = .30, p = .77. The average ISS score of Black patients (N = 556, M = 14.32, SD = 13.45) was not significantly different from the average ISS score of White patients (N = 443, M = 13.62, SD = 14.19), t(924.22) = −.79, p = .43. Moreover, the average ICU days of admitted Black patients (N = 243, M = 4.89, SD = 9.06) was not significantly more than their White counterparts (N = 209, M = 4.71, SD = 9.15), t(438.62) = −.22, p = .83. Although there were more Black patients (N = 75) that died in the hospital than White patients (N = 50), there was no significant difference in the proportion of Black and White patients that died in the hospital (13.44% vs. 11.16%; χ2(1, 1005) = 1.19, p = .276).

The Interaction of Race and PECs in the Legal Intervention Injury Group

There were 406 Black patients who had data available for PECs, 35.22% of whom had a record of PECs. 358 White patients had data available for PECs, 59.22% of whom had a record of PECs. There was a greater proportion of White patients with PECs than the proportions of Black patients with PECs (59.22% vs. 35.22%; OR = 2.67; χ2(1, 763) = 44, p < .001, φ = .24). There was a significant association between proportions of race and PECs for patients injured by firearms, χ2(1, 384) = 34.04, p < .001, φ = .30. Post hoc analyses showed that that among patients injured by firearms, there was a greater proportion of White patients with PECs (60%) than the proportions of Black patients with PECs (29.91%; OR = 2.01; p < .001), while the proportion of Black patients without PECs (70.09%) injured by firearms was significantly greater than other groups (all p < .05).

There was a significant association between race and PECs for patients admitted with neurotrauma, χ2(1, 311) = 17.12, p < .001, φ = .23. For Black patients, PECs were associated with a lower incidence of neurotrauma (37.58 % neurotrauma vs. 62.42% orthopedic trauma), whereas the opposite association was seen in White patients (61.04% neurotrauma vs. 38.96% orthopedic trauma). Additionally, for patients with PECs, White patients were more likely to be admitted with neurotrauma than Black patients (61.04% vs. 37.58%; OR = 2.60; p < .001), while Black patients without PECs were more likely to be admitted with neurotrauma than White patients without PECs (62.42% vs. 38.96%; OR = 2.60; p < .001). There was significant association between race and PECs for deceased patients in hospitals (fisher-exact p = .015). Post hoc analyses showed that Black patients without PECs (91.67%; N = 22) were more likely to die in the hospital than both White patients with PECs (40.91%; N = 9; OR = 15.89) and without PECs (59.09%, N = 13; OR = 7.62), and that White patients with and without PECs were more likely to die than Black patients with PECs (all p < .05).

There were no significant differences for mean GCS, ISS, or ICU days between Black and White patients with or without PECs (all p > .05).

The Interaction of Race and Intoxication in the Legal Intervention Injury Group

There were 406 Black patients had available data on alcohol or drug screen, 55.17% of whom screened positive for either alcohol or drug intoxication. 347 White patients had available data on alcohol or drug screen, 67.15% of whom screened positive. There was a significant association between race and intoxication for patients injured by firearms, χ2(1, 426) = 13.61, p < .001, φ = .18. Among patients injured by firearms, the proportion of White patients with a positive intoxication screen (66.09%) was greater than proportions of both Black patients with negative (51.98%) and positive intoxication screen (48.02%; all p < .01).

Additionally, there was a significant association between race and intoxication for patients with neurotrauma, χ2(1, 314) = 4.94, p = .026, φ = .13. Among patients admitted with neurotrauma, the proportion of White patients with positive intoxication screen (70%) was greater than proportions of both Black patients with a positive (57.93%) and negative intoxication screen (42.07%; all p < .05). There was a significant association between race and intoxication for deceased patients, χ2(1, 60) = 4.42, p = .035, φ = .27. For patients with a positive intoxication screen, White patients (58.06%, N = 18) were more likely to die in the hospital than Black patients (31.03%, N = 9; OR = 3.08, p = .038), while for patients with negative intoxication screen, Black patients (68.97%, N = 20; OR = 3.08, p = .038) were more likely to die in the hospital than White patients (41.94%, N = 13).

There were no significant differences in the means of GCS, ISS, and ICU days between Black and White patients regardless of intoxication status (all p > .05). However, Black patients without intoxication (M = 14.52) scored higher ISS than intoxicated Black patients (M = 12.07), t(370.64) = 2.13, p = .033, d = 2.14.

The Interaction of Race and SES in the Legal Intervention Injury Group

There were 469 Black patients and 387 White patients had valid data on primary payor class. Based on the payor class defined in the method, 77.61% of Black patients were categorized as lower SES, and 18.12% as higher SES, while 57.36% of White patients were categorized as lower SES, and 31.01% as higher SES. The proportion of Black patients with lower SES than White patient with lower SES (OR = 2.32; χ2(1, 855) = 40.22, p < .001, φ = .22).

There was a significant association between race and SES for patients injured by firearms, χ2(1, 429) = 7.44, p = .006, φ = .13. Post hoc analyses showed that proportions of Black (19.42%) and White (31.13%) patients with higher SES injured by firearms were both significantly lower than proportions of patients with lower SES regardless of being White (68.87%) or Black (80.58%) with Black patients with lower SES endorsing the highest proportion of being injured by firearms (all p < .001).

Similarly, there was significant association between race and SES for patients admitted with neurotrauma, χ2(1, 322) = 9.14, p = .002, φ = .17. Post hoc analyses showed that proportions of Black (21.84%) and White (37.16%) patients with higher SES admitted with neurotrauma were both significantly lower than proportions of patients with lower SES regardless of being Black (78.16%) or White (62.84%) with Black patients with lower SES had the highest proportion of neurotrauma (all p < .01). However, there was no significant association between race and SES for patients deceased in the hospitals (fisher-exact p = 1).

Moreover, there were no significant differences in the means of GCS, ISS, and ICU days between Black and White patients regardless of SES (all p > .05). However, White patients with lower SES (M = 14.12) had greater ISS scores than White patients with higher SES (M = 11.16), t(313.62) = 2.02, p = .044, d = 1.98.

Contextual Factors for Legal Intervention Injury

Overall, the impacts of race, PECs and intoxication on injury time followed a similar trend and are consistent with a larger sample examining a similar phenomenon using PTOS data to examine alcohol and brain injury (see Bernier & Hillary, 2016). Early morning (7AM - 10AM) had the lowest rates of patients with legal intervention injury admitted to the hospitals. After 10AM, the rates started increasing until 11PM – 2AM, achieving the peak of legal intervention injury hospital admission. Notably, intoxicated patients were more likely to be injured in the midnight (11PM – 2AM; 30.92% vs. 23.89%) compared with those who were not intoxicated.

Black patients were more likely to be injured at street or highway (45.67% vs. 32.47%) or unspecified place (25% vs. 19.22%) compared with their White counterparts, while White patients were more likely to be injured at home (27.53% vs. 10.24%; all p < .05). Patients with PECs were more likely to be injured at home (27.93% vs. 8.85%), while those without PECs were more likely to be injured at street or highway (48.70% vs. 30.03%; all p < .05). In terms of discharge destination, Black patients were more likely to be discharged to legal authority (47.62% vs. 40.36%) while their White counterparts were more likely to be discharged to psychiatric facility (11.20% vs. 2.04%; all p < .05). Additionally, patients with PECs were more likely to be discharged to psychiatric facility (12.84% vs. 1.50%), while those without PECs were more likely to be discharged to home (34.16% vs. 22.40%; all p < .01). Moreover, intoxicated patients were less likely to be discharged to psychiatric facility (5.54% vs. 9.65%) or other hospitals/trauma centers (3.55% vs. 6.95%; all p < .05).

Discussion

There has been a paucity of research examining injury during police intervention, including little information regarding risk factors and clinical outcomes. To bridge this gap and provide empirical information for related policy and practice recommendations, the current study aimed to examine potential risk and contextual factors associated with legal intervention injury, and related clinical outcomes based on a Pennsylvania state-wide trauma center registry dataset operated by PTSF.

Demographics and Legal Intervention Injury

Consistent with previous research, there was overrepresentation of Black males in patients admitted for injury following legal intervention within our sample (Krieger et al., 2015; Miller et al., 2017; Mooney et al., 2018). Black or African Americans were around 9.2% to 12% of the population in Pennsylvania and 15.72% of the larger PTOS dataset, yet 50.82% of trauma patients injured through legal intervention were Black in the current sample. More than two thirds of patients in the legal intervention group were aged 18–40 and of lower SES. More than half of patients in this sample were injured by firearms and more than one fourth of patients were injured by physical confrontation during legal intervention. Among admitted patients injured by firearms, 33.59% of them were injured between 1991 – 2000, 30.59% between 2001 – 2010, and 26.15% between 2011 – 2019. Unfortunately, there is no additional information regarding reasons for use of force in the ICD codes or the dataset manual.

Psychiatric and Substance Use Disorders and Intoxication are Important Risk Factors

Aligning with previous research (Heffernan et al., 2003; Holloway-Beth et al., 2016; Laniyonu & Goff, 2021), these findings suggest individuals with psychiatric and/or substance use disorders, or those who were intoxicated, especially intoxicated with psychoactive and illicit drugs, were more likely to be injured through legal intervention and subsequently hospitalized. Intoxicating substances can increase erratic behavior and place individuals at greater risk for legal interaction because they may become aggressive, difficult to understand, and/or have impaired judgment and decision-making abilities (Holloway-Beth et al., 2016; Kesic, et al., 2013). All these interactions carry a risk of escalation, which may lead to use of force from the police. However, existing evidence suggests that preventative strategies, such as providing effective trainings and education to police officers, building up police-assisted referral programs and community treatment programs for mental health and substance use disorder services, can effectively decrease the use of excessive force against these individuals (Davoust et al., 2021; Pollack & Humphreys, 2020). Although 46.35% of patients in our study had pre-existing psychiatric and/or substance use conditions, only 6.10% of these patients were discharged to psychiatric facilities. Moreover, 61.43% of patients tested positive for alcohol and/or drugs on admission, but only one individual was discharged to drug or alcohol rehabilitation.

Legal Intervention Injury in Context

Previous findings suggest that the reasons for injury and fatalities during police intervention are multifactorial and should take social geography and other contextual factors into consideration using an intersectional approach (Cooper et al., 2004; Gilbert & Ray, 2016). In line with the intersectionality framework, our findings suggested that no social category or identity is more salient than another in the context of examining police violence. Black patients who were intoxicated and without PECs had a higher mortality rate in our sample, while PECs, intoxication status and lower SES played greater roles in increasing risk for legal intervention injury and higher injury severity among White patients. Additionally, our study captured the impact of social geography, such that time and place also played an important role in the traumatic legal intervention injury. There was a greater risk of legal intervention injury at night (7PM – 2AM), especially for people under the influence of intoxicating substances. Moreover, Black patients were more likely to be injured at street or highway, and White patients or patients with PECs were more likely to be injured at home. All these findings suggest that police violence is a complex phenomenon, and the inclusion of multiple intersecting identities as well as contextual factors, such as time and place, can help us better understand it, which is one of the intersectionality framework’s greatest assets. Methodological challenges exist and there is much work needed in understanding ideal approaches to include an intersectional framework in the study of police-related violence (Bowleg, 2012). As a first step, collecting data on social identities, including race, ethnicity, age, SES, gender, sexual identity, and disability status, and including them and their interactions in the study may inspire new research questions and findings as well as deepen understanding of social phenomenon.

Our findings highlighted the benefit and importance of using state-level hospital trauma center datasets to understand legal intervention injury. Even though police violence has received increasing public attention, the majority of previous analyses of the epidemiology of injuries resulting from encounters with law enforcement has been focused on fatalities. However, nonfatal injuries through legal intervention, which far outnumber deaths, has been understudied. Moreover, research using a nationally representative data set showed that frequencies of patients injured by law enforcement in the United States has been stable over years, though public attention has increased (Kaufman et al., 2017). This suggests police violence has been a longer-term public health issue. Using state-level hospital trauma center datasets can help researchers understand this ongoing phenomenon and underlying factors. Our findings also emphasized the need for supporting and refining the administrative data collection at trauma centers to facilitate speedy and accurate data collection and preparation process, which will ultimately benefit future research on legal intervention injury and inform injury prevention efforts among clinicians, communities, law enforcement agencies and personnel, and policy makers.

Limitations

There are several limitations of this study. First, patients in the current sample were admitted to accredited hospital’s trauma centers and suffered from traumatic injury. While this helps us to examine the legal intervention injury among trauma patients, the findings may not be generalizable to other circumstances in which the injury does not lead to trauma center admission or in which victims died during legal intervention. Second, there was substantial missing data, and existing data may be biased. For example, data on ethnicity was largely absent from this dataset (55.1% with data), which is crucial to study racial and ethnic disparities in police encounters and associated violence. Additionally, PEC diagnosis may include biased data, as those that are more likely to seek frequent medical care may be more likely to be diagnosed with PECs. Importantly, women were also critically underrepresented in our sample. Previous research has suggested a notable proportion of women experience police violence and related deaths, with significantly higher risks for women of racial and ethnic minorities (Fedina et al., 2018; Ritchie, 2017).

Moreover, the current study enrolled all patients injured through legal intervention from 1986 to 2019 in order to have a larger study sample for subgroup analyses. Even though previous study suggested frequencies of patients injured by legal intervention admitted to trauma centers have been stable over years using nationally representative dataset (Kaufman et al., 2017), we acknowledge potential limitations of including patients across this time frame. However, ​​the assurance of data quality is achieved by utilizing standardized abstraction software that incorporates automatic data checks, a manual defining the data, and both internal and external data audits (Glance et al., 2012).

Of note, ICD codes do not provide enough details on the nature of legal intervention injury. The World Health Organization has defined police officers’ excessive use of force as a form of violence, and identified four domains of violence, including physical, sexual, psychological, and neglectful (Cooper et al., 2004). ICD codes only include physical violence, and they do not provide justification or rationale on the use of force. Moreover, ICD codes of legal intervention injury include injury caused by any law enforcement officials regardless of if they are on duty or not, which makes it difficult to study subgroups of law enforcement officials and other contextual factors.

Finally, payor class was used as an indirect measure of SES, which is an imperfect approximate. Although previous studies have examined payor class similarly (Dell et al., 2021; Nieman et al., 2015), this presents a false binary and does not appropriately account for access to resources or changes that individuals may experience over time. However, given the data available, the authors feel that examining SES, even via limited means, is a necessary consideration. In the future more data on social geography could provide a more comprehensive view of legal intervention injury.

Conclusions

Although police violence has been described as a public health crisis, official, reliable, and valid data on police violence is often inaccessible to researchers and policymakers (Krieger et al., 2015). Our study, which examined a trauma registry dataset in Pennsylvania, provides further evidence that police violence disproportionately impacts racial minority groups, especially Black men, as well as other disadvantaged and marginalized communities, including individuals with psychiatric and substance use disorders and those with lower SES. Researchers have proposed multilevel solutions to tackle police violence and its profound impacts on individuals, such as addressing systemic and structural racism and discrimination within individual police departments and justice systems through changing police practices at the micro-level to reforming laws and policies that contain pervasive racial biases at the macro-level (see DeVylder et al., 2021; Gilbert & Ray, 2016; Hughey, 2015; Obasogie, 2020). We also suggest police departments consider developing, reforming, and evaluating programs designed to reduce the use of force, such as investing in public health prevention and treatment programs and redirecting emergency calls related to mental health to professional service providers. Nevertheless, recognizing police violence as a public health issue is the first step. Future studies should investigate risk and contextual factors of legal intervention injury across a larger sample, including those who are not admitted to trauma centers. More research is needed to develop, implement, and evaluate police practices that have the potential to reduce these disparities in the use of force in policing.

Acknowledgement

We would like to thank and acknowledge the Pennsylvania Trauma Systems Foundation for their contributions to this study. The data presented here were provided by the Pennsylvania Trauma Systems Foundation, Mechanicsburg, Pennsylvania (PTSF). The PTSF specifically disclaims responsibility for any analyses, interpretations, or conclusions. Credit must be given to the Pennsylvania Trauma Outcome Study (PTOS) as the source of the data.

Funding

ZZZ was funded by the National Institutes of Health (T32HD101390). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Footnotes

Disclosure Statement

No known conflict of interest was reported by authors.

Data Availability Statement

Date used in this study can be obtained via application through the Pennsylvania Trauma Systems Foundation: https://www.ptsf.org/research/request-trauma-registry-data/

References

  1. Bernier RA, & Hillary FG (2016). Trends in alcohol use during moderate and severe traumatic brain injury: 18 years of neurotrauma in Pennsylvania. Brain Injury, 30(4), 414–421. 10.3109/02699052.2015.1127998 [DOI] [PubMed] [Google Scholar]
  2. Bowleg L (2012). The problem with the phrase women and minorities: intersectionality—an important theoretical framework for public health. American journal of public health, 102(7), 1267–1273. 10.2105/AJPH.2012.300750 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bowleg L (2021). Evolving intersectionality within public health: From analysis to action. American Journal of Public Health, 111(1), 88–90. 10.2105/AJPH.2020.306031 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Centers of Disease Control and Prevention. National Center for Injury Prevention and Control. Retrieved May 25, 2022, from https://www.cdc.gov/injury/data-statistics-reporting.html
  5. Cooper HL (2015). War on drugs policing and police brutality. Substance Use & Misuse, 50(8–9), 1188–1194. 10.3109/10826084.2015.1007669 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Cooper HL, & Fullilove M (2016). Excessive police violence as a public health issue. Journal of Urban Health, 93(1), 1–7. 10.1007/s11524-016-0040-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Cooper H, Moore L, Gruskin S, & Krieger N (2004). Characterizing perceived police violence: implications for public health. American Journal of Public Health, 94(7), 1109–1118. 10.2105/AJPH.94.7.1109 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Davoust M, Grim V, Hunter A, Jones DK, Rosenbloom D, Stein MD, & Drainoni ML (2021). Examining the implementation of police-assisted referral programs for substance use disorder services in Massachusetts. International Journal of Drug Policy, 92, 103142. 10.1016/j.drugpo.2021.103142 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Dell KC, Staph J, & Hillary FG (2021). Traumatic brain injury in the homeless: health, injury mechanisms, and hospital course. Brain injury, 35(10), 1192–1200. 10.1080/02699052.2021.1958009 [DOI] [PubMed] [Google Scholar]
  10. DeVylder JE, Anglin DM, Bowleg L, Fedina L, & Link BG (2021). Police violence and public health. Annual Review of Clinical Psychology, 18. 10.1146/annurev-clinpsy-072720-020644 [DOI] [PubMed] [Google Scholar]
  11. DeVylder JE, Jun HJ, Fedina L, Coleman D, Anglin D, Cogburn C, Link B and Barth RP (2018). Association of exposure to police violence with prevalence of mental health symptoms among urban residents in the United States. JAMA Network Open, 1(7), e184945–e184945. 10.1001/jamanetworkopen.2018.4945 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Edwards F, Esposito MH, & Lee H (2018). Risk of police-involved death by race/ethnicity and place, United States, 2012–2018. American Journal of Public Health, 108(9), 1241–1248. 10.2105/AJPH.2018.304559 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Ellrich K, & Baier D (2016). The influence of personality on violent victimization–a study on police officers. Psychology, Crime & Law, 22(6), 538–560. 10.1080/1068316X.2016.1168424 [DOI] [Google Scholar]
  14. Fedina L, Backes BL, Jun HJ, Shah R, Nam B, Link BG, & DeVylder JE (2018). Police violence among women in four US cities. Preventive Medicine, 106, 150–156. 10.1016/j.ypmed.2017.10.037 [DOI] [PubMed] [Google Scholar]
  15. Flexon JL, D’Alessio SJ, Stolzenberg L, & Greenleaf RG (2019). Interracial encounters with the police: findings from the NCVS police-public contact survey. Journal of Ethnicity in Criminal Justice, 17(4), 299–320. 10.1080/15377938.2019.1646688 [DOI] [Google Scholar]
  16. Fyfe NR (1991). The police, space and society: the geography of policing. Progress in Human Geography, 15(3), 249–267. [Google Scholar]
  17. Gilbert KL, & Ray R (2016). Why police kill Black males with impunity: Applying public health critical race praxis (PHCRP) to address the determinants of policing behaviors and “justifiable” homicides in the USA. Journal of Urban Health, 93(1), 122–140. 10.1007/s11524-015-0005-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Glance LG, Osler TM, Mukamel DB, & Dick AW (2012). Outcomes of adult trauma patients admitted to trauma centers in Pennsylvania, 2000–2009. Archives of surgery, 147(8), 732–737. 10.1001/archsurg.2012.1138 [DOI] [PubMed] [Google Scholar]
  19. Harrell E, & Davis E (2020). Contacts between police and the public, 2018–statistical tables. Bureau of Justice Statics Report, NCJ, 255730. Available from https://www.ojp.gov/library/publications/contacts-between-police-and-public-2018-statistical-tables [Google Scholar]
  20. Hedegaard H, & Johnson RL (2019). An Updated international dlassification of diseases, 10th revision, clinical Modification (ICD–10–CM) surveillance case definition for injury hospitalizations. [PubMed]
  21. Heffernan EB, Saunders JB, Byrne G, & Finn J (2003). Substance‐use disorders and psychological distress among police arrestees. Medical Journal of Australia, 179(8), 408–411. 10.5694/j.1326-5377.2003.tb05617.x [DOI] [PubMed] [Google Scholar]
  22. Ho DE, Imai K, King G, Stuart EA (2011). “MatchIt: Nonparametric Preprocessing for Parametric Causal Inference.” Journal of Statistical Software, 42(8), 1–28. 10.18637/jss.v042.i08 [DOI] [Google Scholar]
  23. Holloway-Beth A, Forst L, Lippert J, Brandt-Rauf S, Freels S, & Friedman L (2016). Risk factors associated with legal interventions. Injury Epidemiology, 3(1), 1–9. https://doi-org.ezaccess.libraries.psu.edu/10.1186/s40621-016-0067-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Hughey MW (2015). The five I’s of five-O: Racial ideologies, institutions, interests, identities, and interactions of police violence. Critical Sociology, 41(6), 857–871. 10.1177/0896920515589724 [DOI] [Google Scholar]
  25. Hyland S, Langton L, & Davis EJ (2015). Police use of nonfatal force, 2002–11. Washington DC: US Department of Justice, Office of Justice Programs, Bureau of Justice Statistics. Available from https://bjs.ojp.gov/content/pub/pdf/punf0211.pdf [Google Scholar]
  26. Kaufman EJ, Karp DN, & Delgado MK (2017). US emergency department encounters for law enforcement–associated injury, 2006–2012. JAMA surgery, 152(6), 603–605. 10.1001/jamasurg.2017.0574 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Kavanagh J (1997). The occurrence of resisting arrest in arrest encounters: A study of police-citizen violence. Criminal Justice Review, 22(1), 16–33. 10.1177/073401689702200103 [DOI] [Google Scholar]
  28. Kesic D, Thomas SD, & Ogloff JR (2013). Estimated rates of mental disorders in, and situational characteristics of, incidents of nonfatal use of force by police. Social Psychiatry and Psychiatric Epidemiology, 48(2), 225–232. 10.1007/s00127-012-0543-4 [DOI] [PubMed] [Google Scholar]
  29. Krieger N, Chen JT, Waterman PD, Kiang MV, & Feldman J (2015). Police killings and police deaths are public health data and can be counted. PLoS Medicine, 12(12), e1001915. 10.1371/journal.pmed.1001915 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Krieger N, Kiang MV, Chen JT, & Waterman PD (2015). Trends in US deaths due to legal intervention among Black and White men, age 15–34 years, by county income level. Harvard Public Health Review, 3, 1–5. [Google Scholar]
  31. Laniyonu A, & Goff PA (2021). Measuring disparities in police use of force and injury among persons with serious mental illness. BMC Psychiatry, 21(1), 1–8. 10.1186/s12888-021-03510-w [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Loder RT, & Leiser A (2021). Injury patterns and demographics due to legal intervention seen in US emergency departments. Journal of forensic and legal medicine, 79, 102150. 10.1016/j.jflm.2021.102150 [DOI] [PubMed] [Google Scholar]
  33. MacDonald JM, Kaminski RJ, & Smith MR (2009). The effect of less-lethal weapons on injuries in police use-of-force events. American Journal of Public Health, 99(12), 2268–2274. 10.2105/AJPH.2009.159616 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. McClelland GM, & Teplin LA (2001). Alcohol intoxication and violent crime: Implications for public health policy. The American Journal on Addictions, 10, s70–s85. 10.1080/10550490150504155 [DOI] [PubMed] [Google Scholar]
  35. Miller TR, Lawrence BA, Carlson NN, Hendrie D, Randall S, Rockett IR, & Spicer RS (2017). Perils of police action: a cautionary tale from US data sets. Injury Prevention, 23(1), 27–32. 10.1136/injuryprev-2016-042023 [DOI] [PubMed] [Google Scholar]
  36. Mooney AC, McConville S, Rappaport AJ, & Hsia RY (2018). Association of legal intervention injuries with race and ethnicity among patients treated in emergency departments in California. JAMA Network Open, 1(5), e182150–e182150. 10.1001/jamanetworkopen.2018.2150 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Nieman CL, Benke JR, & Boss EF (2015). Does race/ethnicity or socioeconomic status influence patient satisfaction in pediatric surgical care?. Otolaryngology--Head and Neck Surgery, 153(4), 620–628. 10.1177/0194599815590592 [DOI] [PubMed] [Google Scholar]
  38. Obasogie OK (2020). More than bias: How law produces police violence. BUL Rev, 100, 771. [Google Scholar]
  39. Obasogie OK, & Newman Z (2017). Police violence, use of force policies, and public health. American Journal of Law & Medicine, 43(2–3), 279–295. 10.1177/0098858817723665 [DOI] [PubMed] [Google Scholar]
  40. Park JN, Linton SL, Sherman SG, & German D (2019). Police violence among people who inject drugs in Baltimore, Maryland. International Journal of Drug Policy, 64, 54–61. 10.1016/j.drugpo.2018.12.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Pennsylvania Trauma Systems Foundation Operational Manual for the Database Collection System (2022). Available from: https://www.ptsf.org/wp-content/uploads/2021/12/2022-PTOS-Manual-3.14.2022.pdf
  42. Pollack HA, & Humphreys K (2020). Reducing violent incidents between police officers and people with psychiatric or substance use disorders. The ANNALS of the American Academy of Political and Social Science, 687(1), 166–184. 10.1177/0002716219897057 [DOI] [Google Scholar]
  43. R Core Team (2021). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Available from: https://www.r-project.org [Google Scholar]
  44. Ritchie AJ (2017). Invisible no more: Police violence against Black women and women of color. Beacon press. [Google Scholar]
  45. Schwartz GL, & Jahn JL (2020). Mapping fatal police violence across US metropolitan areas: Overall rates and racial/ethnic inequities, 2013–2017. PloS One, 15(6), e0229686. 10.1371/journal.pone.0229686 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Waegel WB (1984). How police justify the use of deadly force. Social Problems, 32(2), 144–15. https://doi-org.ezaccess.libraries.psu.edu/10.2307/800784 [Google Scholar]
  47. Widerström-Noga EG, Felipe-Cuervo E, Broton JG, Duncan RC, & Yezierski RP (1999). Perceived difficulty in dealing with consequences of spinal cord injury. Archives of physical medicine and rehabilitation, 80(5), 580–586. 10.1016/S0003-9993(99)90203-4 [DOI] [PubMed] [Google Scholar]
  48. Wilson L, Stewart W, Dams-O’Connor K, Diaz-Arrastia R, Horton L, Menon DK, & Polinder S (2017). The chronic and evolving neurological consequences of traumatic brain injury. The Lancet Neurology, 16(10), 813–825. 10.1016/S1474-4422(17)30279-X [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Yarwood R (2007). The geographies of policing. Progress in human geography, 31(4), 447–465. 10.1177/0309132507079500 [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

Date used in this study can be obtained via application through the Pennsylvania Trauma Systems Foundation: https://www.ptsf.org/research/request-trauma-registry-data/

RESOURCES