Suicide vs Homicide Firearm Injury Patterns, Weapons, and Mortality: A Study of the National Trauma Data Bank (NTDB)

Background: Firearm related mortality in the USA surpassed all other developed countries. This study hypothesizes that injury patterns, weapon type, and mortality differ between suicide groups as opposed to homicide.


Background
Firearm related injury is a complex public health issue worldwide.While rearm death rates declined from 2003-2015 among most high-income countries, the United States (US) increased from 10.2 to 11.2 per 100,000 people [1].In 2015, the overall US rearm death rate was 11.4 times higher than 28 other high income countries combined, thereby giving the US the highest rates of rearm homicide, rearm suicide, and unintentional rearm related deaths per 100,00 people [1].
Firearm homicide and suicide are among the top ten causes of injury deaths in the US.In 2018 there were 38,390 deaths related to rearms suicide and homicide combined which surpassed unintentional falls and motor vehicle collisions.In 2019, rearm homicide and suicide was the leading cause of violence-related injury death among persons 15-34 years of age [2].Suicide and homicide accounted for 8.1% and 5.1% years of potential life lost, respectively, before the age of 65 in 2019 [2].
Firearm injuries may be preventable, however, the issues are multi-faceted.Access to rearms and inadvertent discharge have been readily discussed in the literature.Intentional rearm injury that does not immediately lead to death comes with signi cant morbidity, recovery, and impact on social and interpersonal needs of the patient.In an attempt to hopefully prevent future occurrences, we must try to understand the nature and extent of the problem.This study seeks to compare the differences in population, substance use, injury patterns, weapon type, and in-hospital mortality between suicide and homicide groups that reach trauma centers for care using a large nationally reported trauma database.

Methods
A retrospective study from January 2017 to December 2019 was performed using the American College of Surgeons (ACS) National Trauma Data Bank (NDTB) [3].All methods were performed in accordance with the relevant guidelines and regulations.This study does not require any approval by an Institutional Review Board as the NTDB contains only publically available, de-identi ed data compiled from millions of electronic records out of hundreds of trauma centers across the US.The NTDB collects data from patients who present to all level I and level II trauma centers.Level III and IV trauma centers can participate in the data bank voluntarily.All data are de-identi ed and aggregated, and searchable through the American College of Surgeons.With more than 7.5 million electronic records, the NTDB is the world's largest trauma data repository [4] and the most robust source of clinical information for rearms injury [5].
The demographic information and injury severity of homicide and suicide groups were compared using a Mann-Whitney test for numerical data and Fisher's exact test for categorical data.Logistic regression was used to determine the ways that injury location, injury type/severity, and detected drugs were associated with suicide as opposed to homicide.The association between weapon type and mortality relative to suicide as opposed to homicide was assessed in Fisher's exact tests.Signi cance was de ned as p < 0.05.

Results
During the study period, 100,031 homicides and 11,714 suicides were identi ed.The homicide group was composed of individuals who were categorized as assault (n = 97,639), legal intervention (n = 2,303), military operations (n = 9), terrorism (n = 76), or war operations (n = 4).The suicide group included the Intentional Self-Harm categories as described above (n = 11,714).Accidental Discharge (n = 18,920) was not included as this topic has been heavily reported and analyzed in the literature.Accidental Malfunction (n = 397), and Undetermined Intent (2,748) categories were not included in the analysis as there was no signi cant statistical value.
Demographics between the two groups differed in several ways.Median age for suicide subjects (36, IQR: 19, 54) was 16 years older than for homicide subjects (20, IQR: 14, 30) (Table 1).African-American subjects made up 62% of the homicide group and only 10% of the suicide group.Asian, American Indian, and White subjects had higher percentages in the suicide group than the homicide group (Table 1).Males comprised the vast majority of both the homicide and suicide groups, 88% and 83%, respectively.The percent and number of subjects in each category are provided, and for age and maximum severity, the median and interquartile (IQR) are reported for each group.
A comparison of Abbreviated Injury Scale (AIS) scores between both groups demonstrated more severe injuries within the suicide group than the homicide group (Table 1).Super cial soft tissue injuries were the most common pattern of injury between both groups, however, some signi cant differences emerged between the two groups (Table 2).Higher rates of injuries to the extremities or pelvic girdle were observed among homicide subjects, while higher rates of injuries to the head or neck were observed among suicide subjects.Considering head or neck injuries occurred most frequently among suicide subjects, a greater incidence of a GCS less than 13, respiratory rate less than 10 or more than 29 (breaths per minute), and skull fracture would be expected in this group (Table 3).The homicide subjects had higher odds of having low systolic blood pressure, penetrating injuries, long bone injury, crush injury, amputation, and paralysis.The odds ratio (OR) and its 95% con dence interval (CI) are shown, along with the number and percent of homicides and suicides who had injuries in each location.The odds ratio (OR) and its 95% con dence interval (CI) are shown, along with the number and percent of subjects within homicide and suicide groups for each.GCS (Glasgow Coma Scale), SBP (Systolic Blood Pressure), 10RR29 (abnormal respiratory rate less than 10 or greater than 29), PEN (Penetrating injury).
Homicide and suicide group subjects also differed by substances found in their system (Table 4).Suicide subjects most commonly had no substances in their system.When illicit/recreational drugs were involved, suicide subjects were more likely to have barbiturates, benzodiazepines, tricyclic antidepressants, or 'other' drugs in their system.Illicit/recreational drugs such as amphetamines, cocaine, ecstasy (MDMA), opioids, phencyclidine (PCP), and cannabinoids were more likely observed in the homicide group with the most common being cannabinoids.The odds ratio (OR) and its 95% con dence interval (CI) are shown, along with the number and percent of homicides and suicides who had a drug in their system.(TCA: tricyclic antidepressants) In-hospital mortality rate was signi cantly higher (OR = 6.2, P < 0.001) within the suicide group (44.8%, 95% CI: 43.9%, 45.7%) compared to the homicide group (11.5%, 95% CI: 11.3%, 11.7%).Most homicide and suicide incidents involved a handgun or an unspeci ed rearm (Table 5).In the suicide group, overall inhospital mortality rates were high, between 44-50% for rearms and handguns versus air guns and shotguns at 4-29% (Table 5).This is a stark contrast to the homicide group with an overall in-hospital mortality rate of 11.5% regardless of weapon type.The number of incidents resulting in mortality and the percent within the group and weapon type, are reported.P-values are from Fisher's exact tests to test differences in mortality between groups for each weapon type.

Discussion
This retrospective study adds to the body of literature concerning gunshot wounds by analyzing data from the NTDB from all trauma centers from within the country representing the highest rates of rearm deaths.
In observing the data from patients who arrive to trauma centers alive, we might glean information that may direct us in helping those who can bene t from clinician intervention.Similar to previous publications, we found the suicide group comprised of mostly older white males.A mortality study performed by Branas et al. found most of the suicides were male, older (> 35 years), white, and at least half were unmarried.Their homicide group was also comprised of younger, African-American males [6].A study on gun violence by Manley et al. also discovered the majority of their patients injured with a documented rearm were African American males, with a median age of 28, and aggravated assault as the most common circumstance [7].A retrospective study on interpersonal violence at a Level I trauma center by Livingston et al. also found a signi cant and disproportionate representation by young African American males [8].
Firearm-related injuries are a signi cant public health crisis in the US, contributing to disability, loss of productivity, and tragically, death [9].This retrospective analysis of the severity of injury and in-hospital mortality found the suicide group experienced more severe injuries in addition to increased mortality compared to the homicide group.Suicide attempts have a case fatality rate of nearly 90% [10].An epidemiological trend study on rearm mortality in the US by Goldstick et al., evaluated data from 2006 to 2014, and noted that suicides were consistently the majority of rearm related deaths ranging from 54.6-63.7%,respectively [11].
Firearm related injuries to the head are a leading cause of TBI related morbidity and mortality, with one third of patients arriving to the hospital alive, ultimately succumbing to their injury [9].The remaining two thirds of these will then go on to have TBI with potentially prolonged rehabilitation, motor de cits, cognitive disability, etc that will worsen the quality of their daily lives after attempting to end it.In this study, head and neck injuries were more commonly associated with suicide.Self-in icted gunshot wounds are more likely to involve the head and neck regions and often result in death [12].In this study, extremity injuries were more commonly observed in the homicide group.This group will also likely need additional rehabilitation, possible prosthesis, and will impact their functionality.Fowler et al., described similar results for patients who arrived to the hospital due to unintentional circumstances [9].Handguns account for most of the rearm related injuries in this retrospective study.A study evaluating homicidal gunshot wounds at a single Level 1 trauma center in Kansas found handguns to be the most common weapon of choice in their patients [13].Another study in a rural Midwest level 1 trauma center performed by Guetschow et al. also observed handguns to be the most common weapon type for unintentional rearm related injuries [14].
There are multiple modi able risk factors for rearm suicide.A recent study from the National Violent Death Reporting System (NVDRS) found that 54% of suicide victims did not have a known mental illness [5].This number may be due to lack of access to mental health resources, lack of nancial resources to obtain therapeutic medications, or missed opportunities for treatment of undiagnosed illness.This retrospective study found that detectable antidepressants had much higher odds (OR 2.47, 95% CI: 1.53, 3.98) in the suicide group than homicide.Access to highly lethal means, in the setting of acute life stressors and substance use, even without prior suicidal ideation, may explain the increase risk of impulsivity and suicide by rearm [5].Conversely, risk factors for homicide include involvement in gang activity and illegal rearm possession along with socioeconomic risk factors such as income and social inequality [5].Our study found the homicide group was more likely to have used illicit/recreational drugs.
Initiating substance use counseling on index admission rather than having planned follow-up as outpatient after discharge may reduce subsequent recurrence.Psychiatric services with possible initiation of appropriate antidepressants in suicidal patients in concordance with substance use counseling may bene t those who were missed, self-medicating, undiagnosed, or needing appropriate care while recovering from their injuries might also improve recovery and long-term outcomes Access to rearms has been shown, and well described in the literature, to be associated with increased risk factor for rearm suicide related injuries [15].In fact, after controlling for confounding factors such as unemployment, mental illness, poverty, illicit drug and alcohol dependence, rearm availability increases the risk of suicide [16].A meta-analysis on accessibility of rearms and risk for suicide and homicide

Table 1
Demographics about subjects included in the study.

Table 2
Location of injuries for homicide and suicide groups.

Table 3
Details and severity of injuries for homicide and suicide groups.

Table 4
Drugs observed in homicide and suicide groups

Table 5
Groups differ in the involvement of weapon types and resulting mortality by group.