Abstract
Background:
Firearm injuries are the leading cause of death among children aged 0 to 17 years in the United States.
Objective:
To examine the factors associated with recurrent firearm injury among children who presented with acute (index) nonfatal firearm injury in the St. Louis region.
Design:
Multicenter, observational, cohort study.
Setting:
2 adult and 2 pediatric level I trauma hospitals in St. Louis, Missouri.
Participants:
Pediatric patients aged 0 to 17 years presenting with an index firearm injury between 2010 and 2019.
Measurements:
From the St. Louis Region-Wide Hospital-Based Violence Intervention Program Data Repository, we collected data on firearm-injured patient demographics, hospital and diagnostic information, health insurance status, and mortality. The Social Vulnerability Index was used to characterize the social vulnerability of the census tracts of patients’ residences. Analysis included descriptive statistics and time-to-event analyses estimating the cumulative incidence of experiencing a recurrent firearm injury.
Results:
During the 10-year study period, 1340 children presented with an index firearm injury. Most patients were Black (87%), non-Hispanic (99%), male (84%), and between the ages of 15 and 17 years (67%). The estimated risk for firearm reinjury was 6% at 1 year and 14% at 5 years after initial injury. Male children and those seen at an adult hospital were at increased risk for reinjury.
Limitation:
Our data set does not account for injuries occurring outside of the study period and for reinjuries presenting to nonstudy hospitals.
Conclusion:
Children who experience an initial firearm injury are at high risk for experiencing a recurrent firearm injury. Interventions are needed to reduce reinjury and address inequities in the demographic and clinical profiles within this cohort of children.
Primary Funding Source:
National Institutes of Health.
Firearm injury is the leading cause of death among children aged 0 to 17 years in the United States (1). Children in the United States are 11 times more likely to die of a firearm injury than children in other high-income countries (2). The firearm injury epidemic among U.S. children was exacerbated during the COVID-19 pandemic, with injuries and deaths increasing since 2020 (3–5), particularly among young Black males. Recent research underscores the enduring effect of systemic racism and poverty as fundamental drivers of firearm injury disparities among Black children (6).
Although nonfatal firearm injury data in the United States are inconsistently reported, at least 20 000 children present to hospitals each year with nonfatal firearm injuries (7, 8). The current study focuses on pediatric injury in St. Louis, Missouri. Like many urban areas in the United States, St. Louis is characterized by chronic disinvestment, high levels of outmobility, a diminishing tax base, and persistent levels of racial and socioeconomic segregation (9, 10). St. Louis also has one of the highest rates of violent injury per capita in the nation and is ranked second in the country for firearm deaths within pediatric populations (5, 11). A retrospective review of pediatric, trauma registry, firearm injury patients in the St. Louis region (spanning Missouri and lllinois) between 2008 and 2013 found that most of the children younger than age 17 years treated for firearm injuries during the study period were Black males. A disproportionate number of these injuries were clustered in areas with high economic disadvantage, as indicated by an average household income of less than $25000 annually (11, 12). Research suggests that firearm injury risk may vary by patient age, location, and community, indicating a need for coordinated regional and national efforts (13, 14).
People who survive 1 firearm injury are at an increased risk for recurrent firearm injury (15). One study reported that children with a prior firearm injury were 12 times as likely to experience a recurrent firearm injury within 5 years when compared with trauma patients whose index injury was not violence related (16). Across all ages, research indicates that males and Black people are at increased risk for recurrent firearm injury (17–19). However, limited research exists identifying predictors of recurrent injury within the pediatric population. The identification of predictors of pediatric recurrent firearm injury beyond initial incidence is needed to guide violence intervention programming (17).
The objective of this study was to examine the factors associated with recurrent firearm injury among children who presented with acute (index) nonfatal firearm injury in the St. Louis region between 1 January 2010 and 31 December 2019. This study builds on recent work using the St. Louis Region-Wide HospitalBased Violence Intervention Program Data Repository (STL-HVIP-DR) to track fatal and nonfatal firearm injury. This repository houses data on all patients who presentto any 1 of the 4 St. Louis adult or pediatric level I trauma hospitals with a violent injury (for example, firearm injury or blunt assault) from 2010 to the present. These 4 level I trauma hospitals primarily serve the St. Louis Metropolitan Statistical Area, which was home to more than 600 000 youth under the age of 18 years in 2010. The STL-HVIP-DR is able to follow pediatric patient encounters across these 4 partner hospitals and over time (19, 20). These data address a growing need for collaborative, cross-sector data sharing. Such collaboration is essential to support violence reduction through comprehensive analysis of firearm-injured patients (8, 21–23).
Methods
Study Design and Setting
This study is a multicenter, retrospective, observational cohort analysis of all children aged 0 to 17 years who experienced an index firearm injury and presented to 1 of the 4 St. Louis adult or pediatric level I trauma hospitals for care from 1 January 2010 to 31 December 2019. The STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) statement was used to inform the study design (24). Approval for the study was granted by the institutional review board from Washington University in St. Louis, which is the umbrella institutional review board for all participating institutions.
Data Source
Our team developed and maintains the STL-HVIP-DR, which contains electronic health record data from patients presenting to any of the 4 partner adult and pediatric level I trauma hospitals in St. Louis (20). The STL-HVIP-DR was developed in parallel with Life Outside of Violence, the St. Louis region-wide, hospital-based, violence intervention program serving these 4 hospitals, which care for most violently injured patients in St. Louis (19, 20). As patients may receive care from multiple hospitals in the region, this data set is needed for a more comprehensive assessment of firearm reinjury. This data source also offers opportunity for robust reinjury analyses because it includes children treated at both adult and pediatric hospitals, and it is able to track individual patient movement across participant hospitals over time (20). This is of particular importance because the American College of Surgeons guidelines recommend that all children aged 15 years and older can be treated at either adult or pediatric trauma hospitals (25).
All 4 participating hospitals share their discharge data with the Missouri Hospital Association (MHA), a nonprofit member organization that represents all acute care hospitals in the state. The MHA then computes and provides a stable universal identifier that links patients across health systems. Through this identifier, the STL-HVIP-DR longitudinally tracks individual patients at the encounter level over time and across partner hospitals. Repository data include demographic characteristics and International Classification of Diseases (ICD) codes associated with violent injury. Methods used to construct this data set, including converting ICD, Ninth Revision (ICD-9) to ICD, Tenth Revision (ICD-10) codes using General Equivalence Mappings, have previously been published by our team (19, 20, 26). In addition, we queried National Death Index (NDI) data to identify recurrent firearm injury patients who died in the prehospital setting (27).
Participants and Classification of Index and Recurrent Firearm Injury
Inclusion criteria were all firearm-injured children aged 17 years or younger who presented for care at a partner adult or pediatric trauma hospital for an index (acute) firearm injury between 1 January 2010 and 31 December 2019. Index injury was operationally defined as the first firearm injury captured in the repository during the study period. Patients with ICD diagnostic codes in the following categories were included in the analysis: W32, W33, W34, X93, X94, X95,Y22,Y23,and Y24.
The primary outcome of interest was recurrent firearm injury. Patients were considered to have a recurrent firearm injury if they 1) returned to a partner hospital with a new firearm injury or 2) died in the prehospital setting from a new firearm injury as determined through NDI record analysis within the 10-year study period. To more accurately identify acute firearm injuries (both index and recurrent), we applied our study team’s previously validated machine-learning least absolute shrinkage and selection operator classification model to this data set (19, 28). The following NDI codes were used to classify patient death by firearm in the prehospital setting: E922.0 to E922.9, E955.0 to E955.4, E965.0 to E965.4, E970, and E985.0 to E985.4. For patients who experienced 3 or more firearm injuries during the study period, only their initial and second firearm injuries were used in the analysis. Only persons who were discharged from the hospital after surviving their index firearm injury were included in the primary outcome analyses.
Recurrent firearm injury was measured as a time-to-event variable. The time to recurrent firearm injury was calculated as the number of years between the date of the index firearm injury and the date of the recurrent fatal or nonfatal firearm injury. Nonfirearm injury deaths identified in the NDI were treated as competing risks, and people were considered censored on the date of this nonfirearm injury death. Persons with neither recurrent injuries nor nonfirearm injury deaths identified by the NDI were censored on 1 January 2020 at the end of the study period.
Variables
Variables queried from the STL-HVIP-DR included demographic information and clinical factors. Demographic variables included race (Black, White, other, and multiracial), sex (male and female), ethnicity (Hispanic and not Hispanic), age (0 to 4, 5to 9, 10 to 14, and 15 to 17 years), and patient address of residence (used to estimate social vulnerability). Clinical factors included insurance status, treatment at a pediatric or adult hospital, and diagnostic codes (depression, drug use disorder, and alcohol use disorder). The primary method of payment was used to assess insurance status and was classified as either publicly insured, privately insured, or not insured. We used ICD codes and Centers for Medicare & Medicaid Services Chronic Conditions Data Warehouse (CMS CCW) codes to assess diagnostic information. We used the Centers for Disease Control and Prevention (CDC) Social Vulnerability Index (SVI) and U.S. Census data geocoded to patients’ residential Census tracts as proxies for social determinants of health (29). The SVI was categorized as low (SVI <0.25), average (SVI between 0.25 and 0.75), and high (SVI >0.75) and corresponded to the percentile of social vulnerability represented in that tract compared with all other U.S. Census tracts.
Statistical Analysis
We reported descriptive statistics for patient-level variables and SVI (means and SD for continuous variables, frequency, and percentages for categorical variables, and medians and IQRs for skewed continuous variables). Binomial proportions and Cls were used to report information about hospital of presentation by age category, and χ2 tests were used to compare these proportions. A time-to-event survival analysis generating the cumulative incidence function for recurrent firearm injury in the presence of competing risks was used to estimate the cumulative incidence of recurrent firearm injury over time (30). Estimated firearm reinjury incidence and associated 95% Cls were calculated. Fine and Gray’s proportional subdistribution hazards models and cumulative incidence functions were used to examine the association of covariates with the risk for reinjury over time (30, 31). Mutivariable hazard ratios and 95% Cls were calculated. Children in all racial categories were included in all analyses; however, due to the limited sample sizes in race categories other than Black and White, these are the only 2 racial groups for which effect estimates and survival functions are shown. All analyses were conducted in Stata 17, SAS 9.4, and SAS Enterprise Guide 8.3.
Role of the Funding Source
The funders had no role in the design, data collection, or analysis, or the decision to submit the manuscript for publication.
Results
During the 2010 to 2019 study period, there were 1340 pediatric patients treated for an index firearm injury at 1 of 4 adult and pediatric partner level I trauma hospitals in the St. Louis region. Most patients were Black (87%), non-Hispanic (99%) males (84%) between the ages of 15 and 17 years (67%; Table). Of the 1340 patients treated for an initial firearm injury, 1287 patients (96%) were discharged from the hospital after surviving their index injury. Of this group of 1287 patients surviving their index injury, 169 (13%) experienced a recurrent firearm injury.
Nearly all patients (98%) in the study sample between the ages of 0 and 14 years were treated at a children’s hospital for their initial injury, compared with 69% of children aged 15 to 17 years. Among the children aged 15 to 17 years who were treated at adult hospitals, 86% were Black males, compared with 79% Black males among those aged 15 to 17 years treated at pediatric hospitals. Of the 53 patients who did not survive their index injury, 62% were Black males and 58% were between 15 and 17 years of age. The NDI identified 6 children who survived their index injury and later died in the prehospital setting, with causes of death as follows: 1 general medical issue (seizure), 2 recurrent injuries (nonfirearm), 2 drug overdoses, and 1 firearm suicide.
Time-to-event analyses showed risk for reinjury over time (Figure 1). The median follow-up time for all patients was 3.8 years (IQR, 1.7 to 6.6 years). The median follow-up time for patients with recurrent firearm injuries was 1.4 years (IQR, 0.2 to 3.3 years), compared with 4.3 years (IQR, 3.9 to 71 years) for patients without recurrent injuries. For the small group of patients who died of nonfirearm injuries as identified by the NDI, the median time from index injury to death was 5.7 years (IQR, 3.9 to 7.1 years). Based on cumulative incidence estimates, the risk for firearm reinjury was 6% (95% Cl, 5% to 8%) at 1 year, 14% (Cl, 12% to 17%) at 5 years, and 19% (Cl, 16% to 22%) at 8 years after index injury.
Figure 1.
Cumulative incidence estimates Of recurrent firearm injury for each year after pediatric index firearm injury.
The associations of covariates with risk for firearm reinjury are presented in Figure 2. The estimated risk for firearm reinjury was highest among children who were male (P < 0.001), older (aged 15 to 17 years; P = 0.034), Black (P < 0.001), living in average (P = 0.018) or high (P = 0.013) SVl areas, and seen at adult hospitals (P < 0.001 for adult hospitals 1 and 2; Appendix Table). After adjusting for all covariates of interest, only males, compared with females (hazard ratio [HR], 2.7[Cl, 1.4 to 5.5]), and children presenting at adult hospitals, compared with children’s hospitals (HR for adult hospital 1, 1.8 [CI, 1.2 to 2.9]; HR for adult hospital 2, 1.7 [CI, 1.0 to 2.9]), had increased risk for reinjury.
Figure 2.
Cumulative incidence curve Of recurrent firearm injury stratified by age, gender, race, Social Vulnerability Index, insurance status, and hospital.
P < 0.001 for age, gender, race, hospital type. P = 0.040 for social vulnerability. P = 0.149 for insurance status.
Discussion
This study details the high rate of index and recurrent firearm injuries among youth and young adults in St. Louis, Missouri. This study builds on prior work conducted in St. Louis regional pediatric hospitals by also capturing children who received index firearm injury care at regional adult level I trauma hospitals. In the last decade, St. Louis level I trauma hospitals cared for 1340 children with index firearm injuries. These children were predominately non-Hispanic Black males aged 15 to 17 years. Analyses estimated that the cumulative incidence of experiencing a second firearm injury was 6% at 1 year after index injury and 14% at 5 years after index injury. These estimates are strikingly high compared with rates nationwide, which indicate a pediatric reinjury rate of 1% in the first year after initial injury (32). One comparable study in an urban setting (Seattle, Washington) found a 6% pediatric reinjury rate, with roughly 75% of these injuries occurring within 5 years of initial injury (33). The high cumulative incidence of pediatric firearm reinjury in St. Louis indicates a great need for prevention efforts in this region.
Both the index and recurrent firearm injury cohorts were disproportionately represented by Black male youths, which aligns with both regional and national trends, emphasizing a growing need for interventions tailored to this population (4, 11, 16, 34, 35). Structural racism in the United States has been implicated as a driving force of firearm injury disparities (36). There is a direct association between Black and White racial segregation in U.S. urban areas and firearm mortality (36). Structural racism may be a key contributor to the firearm injury epidemic in St. Louis: St. Louis is 1 of the 10 most racially segregated cities in the nation and is in the top 5 for firearm homicide rates (36, 37).
In addition, although children between the ages of 15 and 17 years comprised most of the study sample, we would be remiss not to highlight the fact that 13% of pediatric firearm injuries (n = 179) occurred in children under the age of 10 years. The representation of youth aged 0 to 9 years in our study sample is higher than national estimates, which indicate that roughly 7% of pediatric firearm injuries occur within this age group (38). This underscores the presence of violence in the home, school, and community contexts in which St. Louis children exist and indicates a need for additional research focused on the safety of our youngest children.
Our findings also indicated that treatment at a children’s hospital was associated with lower risk for recurrent injury in the age groups of 10 to 14 years and 15 to 17 years. However, Black children in these age groups were less likely than White children in these age groups to receive index firearm injury treatment at a children’s hospital. Children’s hospitals often have greater social need supports than adult hospitals, which may be a protective factor for recurrent injury (39, 40). Future research will delve into which children receive treatment at adult hospitals, and if there are causal pathways linking children’s hospital care to lower risk for recurrent injury.
Every child in this study interacted with a hospital system during their initial injury treatment. Hospital-based care provides a window of opportunity to reach youth and families impacted by firearm injury to identify unmet needs and risk for recurrent injury and to intervene accordingly. Public health interventions, such as hospital-based violence intervention programs, provide behavioral health care and connect families with needed economic and social resources (41–43). Connecting children and their families to violence intervention services at the time of hospital-based care is an opportunity to improve the child’s physical and mental health after violent injury and may protect against future firearm injuries.
Our analyses are limited by the retrospective nature of this study design and potential misclassification bias. Identification of patients included in this cohort depended on accurate ICD coding. It is possible that we missed patients who presented with a firearm injury but did not receive a corresponding ICD diagnosis code. In addition, ICD coding misclassification of firearm injury intent is well documented, with a recent study highlighting that assaults are most often misclassified as accidental injury (44). Some patients likely sustained an index injury before 2010, which could underestimate cumulative incidence of recurrent firearm injury. We may also undercount minor firearm injuries treated only in the community setting by a nonpartner hospital, although we expect most pediatric firearm-injured patients to be transferred to regional level I trauma hospitals, and thus be captured in our data repository. It is also possible that the date of death identification in the NDI is incorrect for a small subset of people. When a death is identified in the NDI, it is highly likely that they were injured and died that same day because their injury was too severe to indicate a need to transport them to one of the level I trauma centers in the region. However, it is possible that, for a small percentage of these children, there is a lapse between date of death and date their body was found.
In addition, our region-wide HVIP was implemented during the end of the 10-year study period. A few patients (n = 22) received HVIP services during the final 17 months of our study; removing these patients from the index study sample did not change cumulative incidence of reinjury among patients with recurrent firearm injuries.
We also had limited data on comorbid conditions and patient-level social determinants of health. The CCW depression and substance use disorder codes presented here are likely substantial undercounts in this non-Medicare pediatric population because it is unlikely that pediatric patients consistently received assessment for these conditions. Finally, because the median follow-up time in the sample is 3.6 years, the number of patients with larger follow-up times is limited and any conclusions related to longer follow-up times (for example, reinjury rates at 8 years and beyond) should be interpreted with caution.
In conclusion, this study highlights the high cumulative incidence of children experiencing recurrent firearm injuries in St. Louis and identifies racial and social vulnerability disparities in this group. Findings also indicate the disproportionate risk for recurrent firearm injury for children whose index firearm injury was treated at an adult hospital rather than at a children’s hospital. Since the study by Choi and colleagues describing pediatric firearm injury in St. Louis was published nearly 10 years ago, rates of pediatric firearm injuries in the St. Louis region have continued to increase (11). Studies such as ours underscore the need to better detail nonfatal pediatric firearm injury and reinjury through robust data sources that can longitudinally track patient encounters across hospitals (both pediatric and adult) and health care systems. Accurate, reliable identification of firearm injury and reinjury can inform secondary prevention efforts, such as hospital-based violence intervention programs, and target them to children at greatest risk.
Table.
Descriptive Statistics of Pediatric Firearm Injury Cases in St. Louis (n = 1340), 2010 to 2019
| Characteristics | Index Firearm Injury (n = 1340), n (%) | No Recurrent Firearm Injury (n = 1118), n (%) | Recurrent Firearm Injury (n = 169), n (%) |
|---|---|---|---|
|
| |||
| Age, y* | |||
| 0–4 | 75 (6) | 69 (6) | 4 (2) |
| 5–9 | 104 (8) | 94 (8) | 3 (2) |
| 10–14 | 263 (20) | 228 (20) | 22 (13) |
| 15–17 | 898 (67) | 727 (65) | 140 (83) |
| Sex | |||
| Male | 1130 (84) | 931 (83) | 160 (95) |
| Female | 210 (16) | 187 (17) | 9 (5) |
| Race † | |||
| Black | 1148 (87) | 940 (85) | 164 (98) |
| White | 164 (12) | 154 (14) | 4 (2) |
| Other | 5 (0.4) | 5 (0.4) | - |
| Multiracial (≥2 races) | 4 (0.3) | 3 (0.3) | - |
| Ethnicity † | |||
| Hispanic | 8 (0.6) | 7 (0.6) | 1 (0.6) |
| Not Hispanic | 1330 (99) | 1109 (99) | 168 (99) |
| Insurance status | |||
| Public | 993 (74) | 820 (73) | 137 (81) |
| Private | 185 (14) | 162 (15) | 16 (10) |
| No insurance | 162 (12) | 136 (12) | 16 (10) |
| State † | |||
| Illinois | 221 (17) | 204 (18) | 10 (6) |
| Missouri | 1114 (83) | 909 (81) | 159 (94) |
| Other state | 4 (0.3) | 4 (0.4) | - |
| Index injury intent | |||
| Assault/homicide | 777 (58) | 622 (56) | 122 (72) |
| Unintentional | 471 (35) | 420 (38) | 34 (20) |
| Undetermined | 92 (7) | 76 (7) | 13 (8) |
| Hospital type ‡ | |||
| Pediatric | 1049 (78) | 908 (81) | 106 (63) |
| Adult | 291 (22) | 210 (19) | 63 (37) |
| CMS CCW codes | |||
| Depression | 16 (1) | 14 (1) | 2 (1) |
| Drug use disorder | 73 (5) | 60 (5) | 13 (8) |
| Alcohol use disorder | 7 (0.5) | 7 (0.6) | - |
| SVI category † | |||
| Low | 83 (7) | 80 (8) | 1 (0.6) |
| Average | 425 (33) | 354 (33) | 54 (33) |
| High | 767 (60) | 625 (59) | 111 (67) |
CMS CCW = Centers for Medicare & Medicaid Services Chronic Conditions Data Warehouse; SVI = Social Vulnerability Index.
Age at initial injury.
Missing values for race, ethnicity, state, and SVI category, respectively, were excluded from this table; race, 19 (1%), 16 (1%), 19 (0.6%); ethnicity, 2 (0.1%), 2 (0.2%), not applicable; state, 1 (0.1%), 1 (0.1%), not applicable; SVI category, 65 (5%), 59 (5%), 3 (2%).
Treating hospital at initial injury.
Acknowledgment:
This work would not have been possible without the exceptional collaboration and ongoing support provided by BJC HealthCare Barnes-Jewish Hospital, BJC HealthCare St. Louis Children’s Hospital, SSM Health Cardinal Glennon Children’s Hospital, SSM Health Saint Louis University Hospital, and Washington University in St. Louis. Many individuals spent countless hours discussing the work, preparing data and legal agreements, attending committee meetings, and working with the data.
Financial Support:
This publication in part was supported by the Missouri Foundation for Health and by the Washington University Institute Of Clinical and Translational Sciences, which is, in part, supported by the National Institutes Of Health (NIH) National Center for Advancing Translational Sciences (NCATS), CTSA grant #UL1TR002345. Dr. Mueller is supported by the NIH Eunice Kennedy Shriver National Institute Of Child Health and Human Development (K23HD107520).
Appendix Table.
HRs for Risk for Firearm Reinjury for All Covariates From Univariable and Multivariable Models That Include All Covariates
| Characteristics | Unadjusted HR (95% CI) | Adjusted HR (95% CI) |
|---|---|---|
|
| ||
| Gender | ||
| Female | Reference | Reference |
| Male | 3.28 (1.66–6.48) | 2.73 (1.37–5.47) |
| Age | ||
| 0–4 y | Reference | Reference |
| 5–9 y | 0.57 (0.13–2.61) | 0.61 (0.13–2.83) |
| 10–14 y | 1.56 (0.53–4.61) | 1.28 (0.42–3.83) |
| 15–17 y | 3.01 (1.09–8.31) | 1.53 (0.53–4.41) |
| Race* | ||
| Black | 6.60 (2.45–17.78) | 3.40 (1.22–9.43) |
| White | Reference | Reference |
| Insurance | ||
| No insurance | 1.18 (0.59–2.38) | 0.84 (0.41–1.73) |
| Medicare/Medicaid/other government | 1.58 (0.94–2.67) | 1.37 (0.78–2.40) |
| All other private | Reference | Reference |
| SVI category | ||
| High | 12.16 (1.69–87.63) | 6.37 (0.85–47.63) |
| Average | 10.88 (1.50–79.12) | 7.20 (0.97–53.64) |
| Low | Reference | Reference |
| Hospital | ||
| Adult hospital 1 | 2.32 (1.55–3.47) | 1.85 (1.19–2.88) |
| Adult hospital 2 | 2.35 (1.44–3.83) | 1.74 (1.04–2.93) |
| Pediatric hospital 1 | 1.06 (0.72–1.55) | 1.03 (0.70–1.52) |
| Pediatric hospital 2 | Reference | Reference |
| Index injury intent | ||
| Assault/homicide | 1.96 (1.34–2.87) | 1.31 (0.88–1.95) |
| Undetermined | 1.43 (0.75–2.72) | 1.47 (0.75–2.85) |
| Unintentional | Reference | Reference |
HR=hazard ratio; SVI = Social Vulnerability Index.
All racial categories were included in the model but due to small sample sizes in other categories, meaningful conclusions could only be drawn for the Black versus White categories.
Footnotes
Disclosures: Disclosures can be viewed at www.acponline.org/authors/icmje/ConflictOfinterestForms.do?msNum=M24-0430.
Reproducible Research Statement: Study protocol: Restricted access; contact Dr. Kristen Mueller at Kristen.mueller@wustl.edu for more information. Statistical code: Restricted access; contact Dr. Daphne Lew at daphne.lew@wustl.edu for more information. Data set: Restricted access; contact Benjamin Cooper at ben.cooper@wustl.edu for more information.
Contributor Information
Zoe M. Miller, Brown School of Social Work, Washington University in St. Louis, St. Louis, Missouri.
Benjamin P. Cooper, Institute for Public Health, Washington University in St. Louis School of Medicine, St. Louis, Missouri.
Daphne Lew, Center for Biostatistics and Data Science, Washington University in St. Louis School of Medicine, St. Louis, Missouri.
Rachel M. Ancona, Department of Emergency Medicine, Washington University in St. Louis School of Medicine, St. Louis, Missouri.
Vicki Moran, Trudy Busch Valentine School of Nursing, Saint Louis University, St. Louis, Missouri.
Christopher Behr, SSM Saint Louis University Hospital, Saint Louis University School of Medicine, St. Louis, Missouri.
Marguerite W. Spruce, Section of Acute & Critical Care Surgery, Department of Surgery, Washington University in St. Louis School of Medicine, St. Louis; USA C-STARS ETL, U.S. Air Force School of Aerospace Medicine, St. Louis, Missouri.
Lindsay M. Kranker, Section of Acute & Critical Care Surgery, Department of Surgery, Washington University in St. Louis School of Medicine, St. Louis, Missouri.
Michael A. Mancini, Saint Louis University School of Social Work, St. Louis, Missouri.
Matt Vogel, Department of Emergency Medicine, Washington University in St. Louis School of Medicine, St. Louis, Missouri; School of Criminal Justice, University at Albany, State University of New York, Albany, New York.
Doug J.E. Schuerer, Section of Acute & Critical Care Surgery, Department of Surgery, Washington University in St. Louis School of Medicine, St. Louis, Missouri.
Lindsay Clukies, Department of Pediatrics, Washington University in St. Louis School of Medicine, St. Louis, Missouri.
Megan L. Ranney, Yale School of Public Health, New Haven, Connecticut.
Randi E. Foraker, Department of Biomedical Informatics, Biostatistics & Medical Epidemiology, University of Missouri School of Medicine, Columbia, Missouri.
Kristen L. Mueller, Department of Emergency Medicine, Washington University in St. Louis School of Medicine, St. Louis, Missouri.
References
- 1.Doh KF, Sheline EK, Wetzel M, et al. Pediatric firearm and motor vehicle collision injuries in the United States: diverging trends. Am J Emerg Med. 2022;53:59–62. doi: 10.1016/j2jem.2021.12.029. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Richardson EG, Hemenway D. Homicide, suicide, and unintentional firearm fatality: comparing the United States with other high-income countries, 2003. J Trauma. 2011;70:238–243. doi: 10.1097/TA.0b013e3181dbaddf [DOI] [PubMed] [Google Scholar]
- 3.lantorno SE, Swendiman RA, Bucher BT, et al. Surge in pediatric firearm injuries presenting to US children’s hospitals during the COVID-19 pandemic. JAMA Pediatr. 2023;177:204–206. doi: 10.1001/jamapediatrics.2022.488 1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Marifio-Ramirez L, Jordan IK, Népoles AM, et al. Comparison of US firearm-related deaths among children and adolescents by race and ethnicity, 1999–2020. JAMA. 2022;328:2359–2360. doi: 10.1001/jama.2022.19508 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Bernardin ME, Clukies L, Gu H, et al. COVID-19 pandemic effects on the epidemiology and mortality of pediatric firearm injuries; a single center study. J Pediatr Surg. 2023;58:1500–1505. doi: 10.1016/j.jpedsurg.2022.10.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Voith LA, Atwell MS, Sorensen A, et al. Identifying risk factors and advancing services for violently injured low-income black youth. J Racial Ethn Health Disparities. 2023;10:1809–1822. doi: 10.1007/540615-022-01365-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Parikh K, Silver A, Patel SJ, et al. Pediatric firearm-related injuries in the United States. Hosp Pediatr. 2017;7:303–312. doi: 10.1542/hpeds.2016-0146 [DOI] [PubMed] [Google Scholar]
- 8.Kaufman EJ, Wiebe DJ, Xiong RA, et al. Epidemiologic trends in fatal and nonfatal firearm injuries in the US, 2009–2017. JAMA Intern Med. 2021;181:237–244. doi: 10.1001/jamainternmed.2020.6696 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Cambria N, Fehler P, Purnell JQ, et al. Segregation in St. Louis: Dismantling the Divide. Washington University in St. Louis; 2018. [Google Scholar]
- 10.Fox B, Trolard A, Simmons M, et al. Assessing the differential impact of vacancy on criminal violence in the city of St. Louis, MO. Criminal Justice Review. 2021;46:156–172. doi: 10.1177/0734016821996795 [DOI] [Google Scholar]
- 11.Choi PM, Hong C, Bansal S, et al. Firearm injuries in the pediatric population: a tale of one city. J Trauma Acute Care Surg. 2016;80:64–69. doi: 10.1097/TA.0000000000000893 [DOI] [PubMed] [Google Scholar]
- 12.Harris KM. Mapping inequality: childhood asthma and environmental injustice, a case study of St. Louis, Missouri. Soc Sci Med. 2019;230:91–110. doi: 10.1016/j.socscimed.2019.03.040 [DOI] [PubMed] [Google Scholar]
- 13.Van Dyke ME, Chen MS, Sheppard M, et al. County-level social vulnerability and emergency department visits for firearm injuries - 10 U.S. jurisdictions, January 1, 2018-December 31, 2021. MMWR Morb Mortal Wkly Rep. 2022;71:873–877. doi: 10.15585/mmwr.mm7127a1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Haddad DN, Kaufman EJ. Rising rates of homicide of children and adolescents: preventable and unacceptable. JAMA Pediatr. 2023;177:117–119. doi: 10.1001/jamapediatrics.2022.4946 [DOI] [PubMed] [Google Scholar]
- 15.Rowhani-Rahbar A, Zatzick D, Wang J, et al. Firearm-related hospitalization and risk for subsequent violent injury, death, or crime perpetration: a cohort study. Ann Intern Med. 2015;162:492–500. doi: 10.7326/M14-2362 [DOI] [PubMed] [Google Scholar]
- 16.O’Neill KM, Jean RA, Dodington J, et al. Evaluation of firearm-related reinjury in Connecticut: an opportunity for gun violence prevention. J Surg Res. 2022;274:23–30. doi: 10.1016/j.jss.2021.12.009 [DOI] [PubMed] [Google Scholar]
- 17.Pear VA, McCort CD, Kravitz-Wirtz N, et al. Risk factors for assaultive reinjury and death following a nonfatal firearm assault injury: a population-based retrospective cohort study. Prev Med. 2020;139:106198. doi: 10.1016/j.ypmed.2020.106198 [DOI] [PubMed] [Google Scholar]
- 18.Marshall WA, Egger ME, Pike A, et al. Recidivism rates following firearm injury as determined by a collaborative hospital and law enforcement database. J Trauma Acute Care Surg. 2020;89:371–376. doi: 10.1097/TA.0000000000002746 [DOI] [PubMed] [Google Scholar]
- 19.Mueller KL, Cooper BP, Moran V, et al. Incidence of and factors associated with recurrent firearm injury among patients presenting to St. Louis trauma centers, 2010 to 2019: a cohort study. Ann Intern Med. 2023;176:1163–1171. doi: 10.7326/M23-0069 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Mueller KL, Trolard A, Moran V, et al. Positioning public health surveillance for observational studies and clinical trials: the St. Louis region-wide hospital-based violence intervention program data repository. Contemp Clin Trials Commun. 2021;21:100683. doi: 10.1016/j.conctc.2020.100683 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Fontanarosa PB, Bibbins-Domingo K. The unrelenting epidemic offirearm violence. JAMA. 2022;328:1201–1203. doi: 10.1001/jama.2022.17293 [DOI] [PubMed] [Google Scholar]
- 22.Mayfield CA, Siegal R, Herring M, et al. A replicable, solution-focused approach to cross-sector data sharing for evaluation of community violence prevention programming. J Public Health Manag Pract. 2022;28:S43–S53. doi: 10.1097/PHH.0000000000001426 [DOI] [PubMed] [Google Scholar]
- 23.The Cardiff Violence Prevention Model. Centers for Disease Control and Prevention, National Center for Injury Prevention and Control, Division of Violence Prevention; 2021. [Google Scholar]
- 24.von Elm E, Altman DG, Egger M, et al. ; STROBE Initiative. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: guidelines for reporting observational studies. Int J Surg. 2014;12:1495–1499. doi: 10.1016/j.ijsu.2014.07.013 [DOI] [PubMed] [Google Scholar]
- 25.American College of Surgeons. Resources for the Optimal Care of the Injured Patient: 2022 Standards. American College of Surgeons; March 2022. [Google Scholar]
- 26.Centers for Medicare & Medicaid Services. 2018 ICD-10 CM and GEMs. Centers for Medicare & Medicaid Services; 2018. [Google Scholar]
- 27.Mueller K, Cooper BP, Moran V, et al. Linking out-of-hospital deaths with a regional hospital-based firearm injury database: a clinical researcher’s guide to accessing data from the National Death Index. lnj Prev. 2022;28:374–378. doi: 10.1136/injuryprev-2021-044516 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Ancona RM, Cooper BP, Foraker R, et al. Machine learning classification of new firearm injury encounters in the St Louis region: 2010–2020. J Am Med Inform Assoc. 8 July 2024. [Epub ahead of print]. doi: 10.1093/jamia/ocae173. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Brignone E, LeJeune K, Mihalko AE, et al. Self-reported social determinants of health and area-level social vulnerability. JAMA Netw Open. 2024;7:e2412109. doi: 10.1001/jamanetworkopen.2024.12109 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.SAS Institute. The LIFETEST Procedure: Analysis of Competing-Risks Data. SAS/STAT®. SAS Institute; 2023. [Google Scholar]
- 31.He P, Eriksson F, Scheike TH, et al. A proportional hazards regression model for the sub-distribution with covariates adjusted censoring weight for competing risks data. Scand Stat Theory Appl. 2016;43:103–122. doi: 10.1111/sjos.12167 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Cortolillo N, Moeller E, Parreco J, et al. Readmission and reinjury patterns in pediatric assault victims. Pediatr Surg Int. 2020;36:191–199. doi: 10.1007/s00383-019-04603-0 [DOI] [PubMed] [Google Scholar]
- 33.Kwon EG, Wang BK, Iverson KR, et al. Interpersonal violence affecting the pediatric population: patterns of injury and recidivism. J Pediatr Surg. 2023;58:136–141. doi: 10.1016/j.jpedsurg.2022.09.009 [DOI] [PubMed] [Google Scholar]
- 34.Bottiani JH, Camacho DA, Johnson SL, et al. Annual Research Review: youth firearm violence disparities in the United States and implications for prevention. J Child Psycho! Psychiatry. 2021;62:563–579. doi: 10.1111/jcpp.13392 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Del Pozo B, Knorre A, Mello MJ, et al. Comparing risks of firearm-related death and injury among young adult males in selected US cities with wartime service in Iraq and Afghanistan. JAMA Netw Open. 2022;5:e2248132. doi: 10.1001/jamanetworkopen.2022.48132 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Houghton A, Jackson-Weaver O, Toraih E, et al. Firearm homicide mortality is influenced by structural racism in US metropolitan areas. J Trauma Acute Care Surg. 2021;91:64–71. doi: 10.1097/TA.0000000000003167 [DOI] [PubMed] [Google Scholar]
- 37.Frey WH. Black-white segregation edges downward since 2000, census shows. 17 December 2018. Accessed at www.brookings.edu/blog/the-avenue/2018/12/17/black-white-segregation-edges-downward-since-2000-census-shows on 10 December 2022.
- 38.Hughes BD, Cummins CB, Shan Y, et al. Pediatric firearm injuries: racial disparities and predictors of healthcare outcomes. J Pediatr Surg. 2020;55:1596–1603. doi: 10.1016/j.jpedsurg.2020.02.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Casimir G Why children’s hospitals are unique and so essential. Front Pediatr. 2019;7:305. doi: 10.3389/fped.2019.00305 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Franz B, Cronin CE. Are children’s hospitals unique in the community benefits they provide? Exploring decisions to prioritize community health needs among U.S. children’s and general hospitals. Front Public Health. 2020;8:47. doi: 10.3389/fpubh.2020.00047 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.The Health Alliance for Violence Intervention. Healing systemic and interpersonal violence. 2020. Accessed at www.thehavi.org on 23 December 2021.
- 42.Mueller KL, Chapman-Kramer K, Cooper BP, et al. A regional approach to hospital-based violence intervention programs through LOV. J Public Health Manag Pract 2023;29:306–316. doi: 10.1097/PHH.0000000000001716 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Nofi CP, Roberts BK, Cornell E, et al. Hospital-based violence intervention programs to reduce firearm injuries in children: a scoping review. J Pediatr Surg. 2023;58:2212–2221. doi: 10.1016/j.jpedsurg.2023.04.020 [DOI] [PubMed] [Google Scholar]
- 44.Miller M, Azrael D, Yenduri R, et al. Assessment of the accuracy of firearm injury intent coding at 3 US hospitals. JAMA Netw Open. 2022;5:e2246429–e. doi: 10.1001/jamanetworkopen.2022.46429 [DOI] [PMC free article] [PubMed] [Google Scholar]