WHAT IS ALREADY KNOWN ON THIS TOPIC
Firearm violence surpassed all other causes of death for children and teens in 2020 in the USA. Haemorrhage is the leading cause of preventable death after trauma, with most deaths occurring in the prehospital period. There is a lack of data regarding prehospital deaths for those who suffer isolated extremity gunshot injury versus other sites of gunshot injury (head, trunk or a combination of sites).
WHAT THIS STUDY ADDS
This study queried the National Violent Death Reporting System for patients aged 0–24 who died from gunshot wounds between 2012 and 2021. We found that extremity gunshot victims were more likely to arrive at the emergency department compared with other site gunshot victims, potentially secondary to haemorrhage control efforts for emergency personnel.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
This is the first national study that compares paediatric and young adult decedents by site of gunshot wound. Extremity gunshot victims can benefit from Stop the Bleed (haemorrhage control) techniques applied immediately by bystanders or themselves. Prevention and harm reduction efforts should be tailored towards the most vulnerable populations, such as implementing a standardised Stop the Bleed curriculum in at-risk local communities.
Background
In 2020, for the first time, firearm violence surpassed all other causes of death among the paediatric population in the USA.1 2 The total number of gun-related fatalities is rising in America, with a concurrent increasing incidence of adolescent firearm injury and mortality.3 Roberts et al recently described a 41.6% increase in firearm death rate among US children from 2018 to 2021.1 Previous epidemiologic studies have found that African American youths are disproportionately affected by firearm fatalities, predominated by males, racial minorities and those suffering from poverty, with a majority of firearm homicides within the age group 15–24 years.4–6
Haemorrhage is the leading cause of preventable death due to trauma. Up to 56% of haemorrhage-related deaths occur in the prehospital period.7 Better data about injury patterns are crucial for identifying interventions to reduce morbidity and mortality from gunshot wounds, including potentially preventable deaths from extremity firearm injuries.7 8 Preventable firearm mortality includes deaths secondary to injuries that may have been treated with, and haemorrhage mitigated by, haemorrhage control techniques such as those described in the ‘Stop the Bleed’ (STB) Initiative. STB was initially launched in 2015 to educate communities on life-saving haemorrhage control interventions, largely as a response to the Sandy Hook Elementary School mass shooting which led to 20 child and eight adult casualties.9–11 STB interventions such as direct pressure and tourniquet use are applicable to extremity haemorrhage and may be lifesaving.12 However, there is a paucity of data regarding the scope of pre-hospital deaths related to isolated extremity firearm injuries. Deaths due to firearms do not affect all communities equally. Previous research has identified higher rates among males and people of colour in urban neighbourhoods.1 13 14
We sought to characterise decedents of isolated extremity gunshot wounds (E-GSW) as compared with decedents of other anatomic site gunshot wounds (O-GSW) using the National Violent Death Reporting System (NVDRS) database. We hypothesised there would be persistent racial and sex disparities in the NVDRS cohort, as well as longer survival times for E-GSW decedents as compared with O-GSW decedents. Identified disparities may help to reveal communities in which STB education could be improved to prevent deaths.
Methods
Data source and cohort
The NVDRS database was queried for all decedents aged 0–24 years who suffered any firearm-related injury from 2012 to 2021. Over the study period, the number of states contributing data to NVDRS increased and concomitantly, overall deaths within NVDRS increased, with 10 322 deaths reported from 2012 to 2016 and 22 087 reported between 2017 and 2021. The NVDRS includes all 50 states within a population-based surveillance system which collects details on deaths from death certificates, medical examiner/coroner files and law enforcement records.15 16 The study cohort was divided into two groups: those who sustained isolated E-GSW and those who sustained O-GSW. Patients with gunshot wounds to both an extremity and another anatomic site were included in the O-GSW cohort. Those with incomplete or missing demographic data and those with unclear injury site classification were excluded. Given NVDRS is a deidentified national database without any individual identifying information, this study was deemed exempt by the Children’s Hospital Orange County Institutional Review Board and granted waiver of informed consent.
Patient and public involvement
The decedents included in this study cohort did not partake in the research process; however, their deidentified data were recorded in NVDRS by medical, community and state partners such as law enforcement, emergency medical services (EMS), hospitals, forensics and coroner services. These patients were subsequently not involved in development of study design, recruitment or conduct and did not suffer any burden of intervention or participation. We hope our study findings are disseminated widely but specifically in patient communities disproportionately impacted by gun violence.
Covariates, exposure and outcomes
Covariates included demographic information such as age, race, sex and ethnicity. Incident-related variables included the type of incident as classified in NVDRS (single homicide, legal intervention, accident, homicide-suicide, undetermined intent) and time period of the observed incident (2012–2016 vs 2017–2021). Single homicide is defined as a mechanism in which the decedent was the only individual who was injured by GSW intentionally by another individual(s). Multiple homicides were included. Deaths secondary to legal intervention are defined as when a decedent is killed by a police or other peace officer (person of legal authority to use deadly force) in the line of duty. Toxicology data were also collected for amphetamine, alcohol, cocaine, opioids and marijuana; however, only half of patients were tested for drug use and therefore these percentages only represent the proportion of those who were tested. The primary exposure of interest was the anatomic site of the gunshot wound, classified as either E-GSW or O-GSW. Other outcome measures included EMS arrival, patient transport to the emergency department (ED), admission to the hospital and duration of survival, grouped by minutes (0–59 min), hours (1 hour to 23 hours 59 min), days (1–29 days), months (1–12 months) and years (1 year and longer). A subgroup analysis was performed on a cohort of decedents aged 0–18 years.
Statistical methods
Missingness analysis
Missing data were assessed and documented for key variables. The extent of missingness was evaluated, and appropriate methods such as multiple imputation or complete case analysis were considered based on the pattern and extent of the missing data.
Descriptive analysis
Descriptive statistics were calculated to summarise the demographic and incident characteristics of the study cohort. Categorical variables were summarised as frequencies and percentages, and continuous variables were summarised as means and SD or medians and IQRs, as appropriate.
Bivariate analysis
Bivariate inferential statistics were employed to compare demographic and incident characteristics between the E-GSW and O-GSW cohorts. χ2 and Fisher’s exact tests were used for categorical variables, and the Wilcoxon Rank-sum test was applied for continuous variables.
Results
Cohort description
The NVDRS identified a total of 40 746 decedents of firearm injuries among patients aged 0–24 years from 2012 to 2021. Of these, 39 878 (97.87%) were O-GSW and 868 (2.13%) were E-GSW. The median age in both groups was 20 years old (p=0.99) in the overall cohort of ages 0–24. Within the subgroup analysis of ages 0–18, the median age in both groups was 17 years.
Demographics and incident characteristics
Black individuals were observed to comprise the majority of both groups; however, E-GSW patients had a higher proportion of black individuals (66.6% O-GSW vs 75.3% E-GSW, p<0.0001) and a lower proportion of white individuals (27.1% O-GSW vs 19.4% E-GSW, p<0.0001). E-GSW had a lower proportion of Hispanics (10.3% vs 15.6%, p=0.0002, table 1) as compared with O-GSW.
Table 1
Demographics comparing other-site gunshot wounds (O-GSW) to extremity gunshot wounds (E-GSW)
Single homicides were the most common incidents but more frequently represented in E-GSW than O-GSW (87.2% vs 82.3%, p=0.0001). Further breakdown of incidents resulting in E-GSW and O-GSW are shown in table 2. The specific causes of accidental injury resulting in E-GSW and O-GSW are shown in table 3. Handguns were the most common type of firearm in both cohorts (59.6% O-GSW vs 60.2% E-GSW, p=0.91). In a subgroup analysis of ages 0–18, this trend persisted (table 4).
Table 2
Incident type comparing other-site gunshot wounds (O-GSW) to extremity gunshot wounds (E-GSW)
Table 3
Distribution of accidental shooting injury comparing other-site gunshot wounds (O-GSW) to extremity gunshot wounds (E-GSW)
Table 4
Gun types comparing other-site gunshot wounds (O-GSW) to extremity gunshot wounds (E-GSW)
Toxicology reports indicated that drug use was generally uncommon among decedents aside from marijuana, which was present in 65.7% O-GSW and 68.0% E-GSW (p=0.4) of those tested (table 5).
Table 5
Toxicology comparing other-site gunshot wounds (O-GSW) to extremity gunshot wounds (E-GSW)
Survival and EMS response characteristics
Ages 0–24 cohort
In terms of survival times, a significant proportion of both groups survived up to 59 min: 72.0% in E-GSW and 74.8% in O-GSW. This was followed by those who survived hours: 20.9% E-GSW vs 17.3% O-GSW. A smaller proportion survived days (5.7% vs 5.5%) and months to years (0.3% vs 0.7%) (p=0.222) (table 6). EMS arrival was documented in the majority of both groups 85.1% in E-GSW and 88.6% in O-GSW (p=0.0003). A higher proportion in the E-GSW group were transported to the ED (66.1% vs 53.4%, p<0.0001). A fraction of these patients was then admitted to the hospital (12.1% E-GSW vs 12.8% O-GSW, p=0.34). Most decedents did not survive more than 24 hours, with 94.0% in E-GSW and 93.7% in O-GSW (p=0.88). Among those who survived more than 24 hours, some had had EMS involvement prior to arrival at the hospital (5.98% E-GSW vs 6.14% O-GSW, p=1) and others did not (10.5% vs 5.23%, p=0.28) (table 7).
Table 6
Distribution of survival comparing other-site gunshot wounds (O-GSW) to extremity gunshot wounds (E-GSW)
Table 7
Disposition comparing other-site gunshot wounds (O-GSW) to extremity gunshot wounds (E-GSW)
Ages 0–18 cohort
In a secondary analysis focusing on patients 0–18 years old, similar patterns were observed. Black youths were disproportionately represented in both groups however more with E-GSW (62.9% O-GSW vs 75.5% E-GSW, p=0.0001) (table 1). In terms of survival times, 72.0% O-GSW vs 79.4% E-GSW survived to 59 min, followed by those who survived hours (18.4% vs 15.3%), and a smaller proportion survived days (7.1% vs 2.6%) and months to years (0.5% vs 0.9%) (p=0.957) (table 6). A higher proportion of E-GSW were transported to the ED (55.6% O-GSW vs 67.4% E-GSW, p=0.007) (table 6).
Subgroup analysis by age: 0–5, 6–10, 11–15, 16–20, 21–24 years
In additional stratified analysis, age groups were divided into infants, toddlers and preschool-aged children 0–5 years, elementary-aged children 6–10 years, pre-teenagers and young teenagers 11–15 years, adolescents aged 16–20 years and young adults 21–24 years (online supplemental tables 1-4). Small frequencies were observed in youngest age groups. With increasing age above 11 years old, similar trends persisted of higher representation of black youths in E-GSW compared with O-GSW: 69.0% vs 53.7% in years 11–15, 73.4% vs 65.4% in years 16–20 and 64.6% vs 74.6% in years 20–24, all p<0.0001 (online supplemental table 1). Males predominated in all age groups in both E-GSW and O-GSW compared with females (online supplemental table 1). Incident type by age (online supplemental table 2a) was compared, with higher frequencies of accidents in younger age groups compared with higher homicides in older age groups. Handguns predominated as type of gun used in groups 6–10 years, 11–15 years, 16–20 years and 21–24 ywars (all p<0.05) (online supplemental table 2b). Statistical significance in toxicology reports (online supplemental table 3) was detected in teenage years, with increasing uses of alcohol, opioids and marijuana.
With respect to transport timing (online supplemental table 4), more EMS arrivals were documented for E-GSW versus O-GSW only in age group 21–24 years (81.6% vs 72.3%, p=0.004). No differences were detected in transportation to the ED between groups (all p>0.05). compared with O-GSW, E-GSW were less likely to be admitted to the hospital with EMS present in groups 6–10 years (4.76% vs 13.3%, p=0.01), 16–20 years (10.0% vs 14.8%, p<0.0001) and 21–24 years old (9.64% vs 12.3%, p<0.0001). Similarly, E-GSW were less likely to survive more than 24 hours in groups 6–10 years (0% vs 6.17%, p=0.01), 11–15 years (3.16% vs 5.32%, p=0.05), 16–20 years (1.70% vs 3.52%, p<0.0001) and 21–24 years old (1.90% vs 2.97%, p<0.0001). No differences were detected when assessing the proportion of patients surviving more than 24 hours without EMS present. However, with EMS present, E-GSW were again less likely to survive more than 24 hours in groups 6–10 years (0% vs 6.38%, p=0.03), 11–15 years (3.4% vs 6.08%, p=0.04), 16–20 years (1.73% vs 3.75%, p<0.0001) and 21–24 years old (1.99% vs 3.21%, p<0.0001).
Discussion and conclusions
This is the first national study comparing paediatric and young adult decedents of isolated extremity GSW to other-site GSW. Decedents who had isolated extremity firearm injury were more likely to die before EMS arrival. However, if EMS arrived and transported E-GSW decedents, the subjects were likely to be transported to the ED, potentially secondary to resuscitated or haemorrhage control efforts by EMS. The data reveal that a large proportion of deaths due to isolated extremity GSW occur within the first hour of injury, suggesting that early haemorrhage control intervention, such as compression or tourniquet usage as taught in STB courses, may be lifesaving.9–11
This study provides a contemporary demographic characterisation of the population of paediatric and young adults who died of GSWs—we have demonstrated that E-GSW deaths occurred more frequently among Black males. Our findings are concordant with previous studies, which have demonstrated that the populations that most frequently suffer assaultive gun injuries are urban, male, black children from the poorest median income categories, while those more likely to suffer firearm injury suicide are white and/or female.13 17 However, previous studies did not consider the isolated population of deaths from GSW (as opposed to injury from GSW) or compare anatomic areas of injury. Decedents in this patient population were adolescents with a median age of 20 years, supporting evidence that older children and young adults disproportionately suffer from firearm violence.18–21 Prior research has identified that firearm homicides are focused in urban communities of colour with histories of institutionalised segregation, redlining and economic disinvestment.20–22 Our findings add that Black children disproportionately die from E-GSW: Black youths comprised three-quarters of the decedents, while only 12% of the total US population.23
Extremity GSW are not as lethal as craniocerebral or torso GSWs.24 25 Vascular injuries from extremity penetrating trauma still remain a significant cause of morbidity and mortality in the US paediatric population, with upper extremity injuries being most common and vascular injuries from firearms being the most deadly.26 Our data demonstrate that 72% of E-GSW victims survived up to 59 min to EMS arrival, but only 21% survival up to 24 hours. This highlights the importance of an early window of intervention for E-GSW patients, as deaths from isolated E-GSW are potentially preventable with timely recognition and direct pressure and/or tourniquets to control the haemorrhage.8 Tourniquet application for exsanguinating extremity haemorrhage has been shown to improve mortality in adults, both in military and civilian settings.27–29 Kragh et al examined the efficacy of tourniquet use in children in a combat zone and demonstrated similar survival rates to adults with exsanguinating extremity trauma.30 31 Our previous research has shown that tourniquets are effective in paediatric extremity haemorrhage.32 Courses such as STB provide education on techniques for haemorrhage control such as direct pressure and tourniquet placement to first responders and bystanders.11 33 By increasing the community’s ability to control exsanguinating haemorrhage in the field through STB education, we can potentially prevent deaths from E-GSW in the minutes to hours after injury.8 11 26 33–35 While STB efforts are critical, they occur at the level of secondary prevention and are in the context of the larger epidemic of gun violence which must be addressed through primary prevention efforts including gun safety and control legislation.
A subgroup analysis was performed by age stratification to determine if patterns of survival were affected by age. Low frequencies limited the ability to detect statistical significance within the youngest age group of 1–5 years, and no statistical significance was detected between extremity or other-site GSW in transport rates to the ED. However, new patterns were detected in ages 6–10 years and 16–24 years that O-GSW patients were more frequently admitted to the hospital. O-GSW in 6 years and above also had a small advantage (1–6%) of survival more than 24 hours compared with E-GSW. This trend further emphasises the critical nature of timely intervention and haemorrhage control techniques8 in the narrow window after isolated extremity GSW that can be assisted by STB community education.9–11 27 30–35
Limitations of our study include its retrospective nature and utilisation of the NVDRS database which has missing data and possibly selection biases due to incomplete reporting. Collection of the circumstances around death can vary between states or districts and is not standardised which can also introduce bias.16 Over the study period, the number of states contributing data to the NVDRS increased significantly and reached 50 states by 2021, therefore, the data in the early period may not be representative of the total US population. While the NVDRS database variables start at EMS arrival and become a functional ‘clock’ for survival times, it is not known exactly when the injuries occurred. The NVDRS database also does not collect variables on the interventions EMS performed at the scene such as tourniquet or STB techniques, which could be correlated directly to influence on survival, and also does not provide resuscitation or transfusion data en route to hospitals. The NVDRS database does not explicitly differentiate between head/torso injury including extremity injury or isolated head/torso injury, which might further elucidate survival relationships and nuances in usage of STB techniques to prevent haemorrhage. In addition, there is no specific junctional injury category; thus, junctional injuries may have been attributed to E-GSW or O-GSW, as there is no standardised definition in the database. Junctional haemorrhage is not amenable to STB techniques in many instances, aside from direct compression when possible.
In conclusion, our study demonstrates that in this paediatric and young adult population, Black males disproportionately suffer mortality due to GSW—of all anatomical areas and even more disproportionately among those with isolated extremity GSWs. E-GSW victims can potentially benefit from STB techniques applied immediately in the field by any bystander or even the victim themselves prior to EMS arrival. Prevention and harm reduction techniques should be tailored towards this vulnerable population, such as through initiating a standardised STB curriculum in schools in the most at-risk local communities. More research can be directed at how effective STB curriculum will be at increasing usage of these techniques in the field and then the resultant mitigation of mortality.
Data availability statement
All data relevant to the study are included in the article or uploaded as supplementary information. All data relevant to the study are included in the text and tables of the article.
