Abstract
Objective
To assess the clinical, racial, and social characteristics of victims of Gunshot wounds (GSWs) to the head and assess for associations between these factors and outcomes.
Summary Background Data
Previous literature has not focused on the association of race and socioeconomic factors with these specific injuries.
Methods
We identified patients with GSWs to the head who presented to 2 urban academic medical centers between 1998 and 2020, and extracted patient-level demographic data, information about the clinical and surgical course, and outcomes at discharge and follow-up.
Results
The cohort included 250 patients, 90% (n = 226) of whom were male, with a mean age of 28 years. Forty-five percent were white (n = 112), 19% Black (n = 48), 18% Latinx (n = 45), with 6% “other” (n = 16), and 12% “unknown” (n = 29). The majority of patients presented with assault-related trauma (n = 153, 61%) as compared to self-inflicted injuries (n = 97, 39%). Across the entire cohort, sex, age, race, and median income by ZIP code were not significant predictors of outcome. Victims of assault by GSW to the head were more likely to be age 18 or younger (OR 5.26, P = 0.01), between the ages of 19 and 33 years (OR 4.7, P = 0.001), Black (OR 6.66, P < .001), and Latinx (OR 2.65, P = 0.03). Most patients (n = 155, 63%) had a poor functional outcome (modified Rankin Score 3–6) at discharge.
Conclusion
Age, race, and income status were not independent predictors of mortality or functional outcome at discharge in our population. Assault-related GSWs to the head mostly involved young Black or Latinx men of lower socioeconomic status, while self-inflicted injuries were largely seen in older white men.
Keywords: gunshot wounds, traumatic brain injury, social determinants, health disparities
Introduction
According to the Centers for Disease Control and Prevention (CDC), 109 Americans die each day as a result of gun violence. 1 In contrast, the United Kingdom had 126 deaths due to gun violence in 2015 2 ; Sweden had 129 in 2016 3 ; Canada had 130 in 2017 4 ; and Australia had 226 in 2019. 5 The United States (US) is responsible for 25 times more firearm-related deaths than all other high-income countries worldwide combined. 6
Gunshot wounds (GSWs) to the head are a common cause of a traumatic brain injury (TBI) in the US, accounting for approximately 35% of all TBI-related deaths (motor vehicle accidents, accounting for 31%, are the second most common cause). 6 Gunshot wounds to the head are a highly lethal form of penetrating trauma; nearly 71% of victims die prior to hospital arrival and 50% of those who do arrive to a hospital die in the emergency department.7-9 Gunshot wounds to the head impose an enormous financial burden on the US healthcare system, with firearm-related injuries costing over $30 000 per patient. 10 The disparities related to these injuries are well-described, with young Black men most likely to be affected.10,11 Compared with white males who survive these injuries, Black male survivors of gun violence may receive fewer resources for recovery and are more likely to be criminalized. 12
A meta-analysis of 1774 patients with GSWs to the head demonstrated that increasing age, suicide attempt, lower admission Glasgow Coma Scale (GCS), bilateral dilated pupils, dural penetration, and bi-hemispheric and multi-lobar injury are associated with increased mortality. 13 However, despite the demographic differences observed among survivors of gun violence, no study has examined the influence of racial and social determinants of health on outcomes after cranial GSW. We aim to describe the associations between racial and socioeconomic factors on mortality rates and functional outcomes following these traumatic injuries.
Methods
We conducted a retrospective cohort study of all patients with a GSW to the head injury who presented to Massachusetts General Hospital (MGH) or Brigham and Women’s Hospital (BWH) between January 1998 and December 2020. The Mass General Brigham (MGB) Research Patient Data Registry (RPDR) and the Mass General Trauma Center database were used to ensure complete reporting of the final cohort of patients with GSWs to the head. Cranial GSWs were identified by the reason for visit and the radiology report for a head, face, and neck computed tomography (CT). Patients diagnosed with assault or intentional self-harm by explosive, handgun, other unspecified firearm, rifle, shotgun, and larger firearm discharge were identified (ICD codes X96, X93, X95, X72, E958, and X73). Non-head and neck gunshot wounds were excluded. Polytrauma cases and instances of withdrawal of life sustaining treatment (WLST) were included. Demographic, clinical, surgical, discharge, and follow-up information were manually extracted from the electronic medical record. Consent was not required given data anonymity and no direct harm to patients.
Race was determined by patient self-report, defined as white, Black, Latinx, “other,” or “unknown” when a race was not documented. The white population was white and non-Latinx. We assigned median household income as reported by the Census Bureau American Community Survey in 2010–2014 to each ZIP code recorded in the medical record. The income classification structure (ie, upper-middle, middle, lower-middle, poor or near-poor) was chosen according to the U.S. News & World Report in 2020. 14 Age categories were determined by generational breakdown according to the Pew Research Center. 15 Bullet trajectory was described as perforating (entry and exit wound), penetrating (entry wound but no exit), or tangential (no dural penetration). In addition to parenchymal injuries, tangential injuries were included to capture an accurate range of injury severity when assessing predictors of mortality and functional outcomes. Suicidal intention was determined by detailed chart review; incidents not associated with a suicide attempt were considered assaults.
Dichotomous variables were represented as a percentage of a population. Non-normally distributed variables were reported as a median with interquartile range (IQR). Chi-square tests were performed to compare neurosurgical and socioeconomic data between race, intention, and outcome groups at discharge and final follow-up. Outcome groups were determined by the modified Rankin Score (mRS) assigned at the applicable timepoint. A mRS of 0–2 was considered a good outcome, while a mRS of 3–6 was a poor outcome. Categories found to be significant in the chi-square analyses were entered in a multivariable logistic regression analysis. A P-value less than .05 was considered significant. Statistical analyses were performed using Stata/SE 16 software (StataCorp LLC, College Station, TX). Institutional Review Board approval was obtained from MGB.
Results
Over the study period, a total of 250 patients with GSWs to the head were identified. The median age at presentation was 28 years (IQR: 22-43) and the vast majority were male (n = 226, 90%). A total of 112 (45%) patients were white, 48 (19%) were Black, and 45 (18%) were Latinx. Thirteen (5%) were considered poor or near-poor, 105 (42%) were lower-middle class, 116 (46%) were middle class, and 9 (4%) were upper-middle class. There were 153 (61%) victims of assault, while 97 (39%) patients suffered self-inflicted injuries. Upon arrival to the hospital, 138 (55%) patients presented with GCS 3–8 and 96 (38%) were GCS 13–15. Among the total cohort, nearly half (46%) of patients had penetrating GSWs to the head, 40% tangential, and 12% perforating. Half of the total population (50%) underwent operative neurosurgical intervention. One-hundred eleven (45%) patients were dead at the time of discharge. Of the survivors, 69 (28%) were discharged to home and 70 (28%) were discharged to a rehabilitation facility (Table 1).
Table 1.
Study Population Characteristics 1998–2020; Stratified by Race.
| Total 250 (100%) | White (%) 112 (45%) | Black (%) 48 (19%) | Latinx (%) 45 (18%) | Other (%) 16 (6%) | Unknown (%) 29 (12%) | P | |
|---|---|---|---|---|---|---|---|
| Sex | .78 | ||||||
| Male | 226 (90) | 100 (89) | 44 (92) | 40 (89) | 14 (88) | 28 (97) | |
| Female | 24 (10) | 12 (11) | 4 (8) | 5 (11) | 2 (13) | 1 (3) | |
| Age | |||||||
| ≤18 | 30 (12) | 7 (6) | 6 (13) | 7 (16) | 3 (19) | 7 (24) | <.001 |
| 19–33 | 120 (48) | 37 (33) | 31 (65) | 32 (71) | 11 (69) | 9 (31) | |
| 34–48 | 52 (21) | 32 (29) | 9 (19) | 4 (9) | 2 (13) | 5 (17) | |
| ≥49 | 48 (19) | 36 (32) | 2 (4) | 2 (4) | 0 (0) | 8 (28) | |
| Income class a | <.001 | ||||||
| Poor, near-poor | 13 (5) | 0 (0) | 7 (15) | 5 (12) | 1 (6) | 0 (0) | |
| Lower-middle | 105 (42) | 36 (33) | 22 (46) | 22 (51) | 12 (75) | 13 (46) | |
| Middle | 116 (46) | 66 (61) | 19 (40) | 15 (35) | 3 (19) | 13 (46) | |
| Upper-middle | 9 (4) | 6 (6) | 0 (0) | 1 (2) | 0 (0) | 2 (7) | |
| Intention | <.001 | ||||||
| Suicide | 153 (61) | 66 (59) | 6 (13) | 12 (27) | 1 (6) | 12 (41) | |
| Assault | 97 (39) | 46 (41) | 42 (88) | 33 (73) | 15 (94) | 17 (59) | |
| GCS | .63 | ||||||
| 3–8 | 138 (55) | 56 (50) | 25 (52) | 29 (64) | 10 (63) | 18 (62) | |
| 9–12 | 16 (6) | 7 (6) | 2 (4) | 3 (7) | 1 (6) | 3 (10) | |
| 13–15 | 96 (38) | 49 (44) | 21 (44) | 13 (29) | 5 (31) | 8 (28) | |
| Trajectory | .13 | ||||||
| Perforating | 29 (12) | 13 (12) | 6 (13) | 8 (18) | 0 (0) | 2 (7) | |
| Penetrating | 116 (46) | 48 (43) | 17 (35) | 26 (58) | 9 (56) | 16 (55) | |
| Tangential | 100 (40) | 49 (44) | 25 (52) | 9 (20) | 7 (44) | 10 (34) | |
| Unknown | 5 (2) | 2 (2) | 0 (0) | 2 (4) | 0 (0) | 1 (3) | |
| Operation | .36 | ||||||
| Yes | 125 (50) | 61 (54) | 25 (52) | 19 (42) | 5 (31) | 15 (52) | |
| No | 125 (50) | 51 (46) | 23 (48) | 26 (58) | 11 (69) | 14 (48) | |
| Discharge | .03 | ||||||
| Home | 69 (28) | 24 (21) | 19 (40) | 10 (22) | 7 (44) | 9 (31) | |
| Rehab | 70 (28) | 42 (38) | 10 (21) | 12 (27) | 0 (0) | 6 (21) | |
| Deceased | 111 (45) | 46 (41) | 19 (40) | 23 (51) | 9 (56) | 14 (48) |
Table 1. Frequency of clinical and social characteristics stratified by race. GCS = Glasgow coma scale.
aN = 243, White n = 108, Black n = 48, Latinx n = 43, Other n = 16, Unknown n = 28. Percentage calculations based on patients with documented ZIP code information.
Racial and Socioeconomic Determinants
The Black, Latinx, and “other” populations were significantly younger than the white and “unknown” populations. For instance, for ages 19–33, there were 31 (65%) Black, 32 (71%) Latinx, and 11 (69%) “other” patients, whereas only 37 (33%) patients were white (P < .001). Similarly, while 36 (32%) white patients were older than age 49, Black (n = 2, 4%), Latinx (n = 2, 4%), and “other” (n = 0, 0%) race groups contributed only 4 patients to this age category (P < .001). Most white patients (n = 66, 61%) were classified as middle class, whereas most Black (n = 22, 46%), Latinx (n = 22, 51%), and “other” (n = 12, 75%) populations were more likely to be lower-middle class (P < .001). Of the 13 patients in the poor or near-group, 7 (54%) were Black, 5 (38%) were Latinx, and 1 (8%) was “other.” The majority of Black (n = 42, 88%), Latinx (n = 33, 73%), and “other” (n = 15, 94%) populations were victims of assault, while most white patients (n = 66, 59%) suffered self-inflicted GSWs to the head (P < .001). White patients were discharged to a rehabilitation facility (n = 42, 38%) to a greater degree than Black (n = 10, 21%), Latinx (n = 12, 27%), or “other” (n = 0, 0%) patients (P = .03) (Table 1).
Table 2 lists patients as stratified by GSW to the head intention (ie, assault vs suicide attempt). The majority (n = 92, 60%) of victims of assault were between ages 19 and 33, while more suicide victims were between the ages 34 and 48 (27% vs 17%) or older than 49 (36% vs 9%, P < .001). Seventy-eight percent (n = 66) of suicide victims were white, while assault victims were 34% (n = 46) white, 31% (n = 42) Black, and 24% (n = 33) Latinx (P < .001). Compared to assault victims, more suicide victims were of the middle (55% vs 43%) and upper-middle income classes (4% vs 3%), while more assault victims were of the poor, near-poor (8% vs 1%) and lower-middle income classes (46% vs 39%, P = 0.04). A GCS between 3 and 8 was the most common score for both suicide (n = 57, 59%) and assault (n = 81, 53%) victims, though fewer suicide victims were given a GCS between 13 and 15 (31% vs 43%, P = 0.04). A greater portion of assault victims were discharged home (35% vs 15%), while more suicide victims were discharged to rehab (36% vs 23%, P = 0.002). Bullet trajectory (P = 0.44) and the occurrence of an operation (P = 0.52) did not differ between intention groups.
Table 2.
Study Population Characteristics 1998–2020; Stratified by Intention.
| Total (%) 250 (100%) | Suicide (%) 97 (39%) | Assault (%) 153 (61%) | P | |
|---|---|---|---|---|
| Sex | .31 | |||
| Male | 226 (90) | 90 (93) | 136 (89) | |
| Female | 24 (10) | 7 (7) | 17 (11) | |
| Age | <.001 | |||
| ≤18 | 30 (12) | 8 (8) | 22 (14) | |
| 19–33 | 120 (48) | 28 (29) | 92 (60) | |
| 34–48 | 52 (21) | 26 (27) | 26 (17) | |
| ≥49 | 49 (19) | 35 (36) | 13 (9) | |
| Race b | <.001 | |||
| White | 112 (51) | 66 (78) | 46 (34) | |
| Black | 44 (22) | 6 (7) | 42 (31) | |
| Latinx | 44 (20) | 11 (13) | 33 (24) | |
| Other | 17 (8) | 2 (2) | 15 (11) | |
| Income class c | .04 | |||
| Poor, near-poor | 9 (4) | 1 (1) | 12 (8) | |
| Lower-middle | 116 (48) | 37 (39) | 68 (46) | |
| Middle | 105 (43) | 52 (55) | 64 (43) | |
| Upper-middle | 13 (5) | 4 (4) | 5 (3) | |
| GCS | .04 | |||
| 3–8 | 138 (55) | 57 (59) | 81 (53) | |
| 9–12 | 16 (6) | 10 (10) | 6 (4) | |
| 13–15 | 96 (38) | 30 (31) | 66 (43) | |
| Trajectory | .44 a | |||
| Perforating | 29 (12) | 14 (14) | 15 (10) | |
| Penetrating | 116 (46) | 45 (46) | 71 (46) | |
| Tangential | 100 (40) | 35 (36) | 65 (42) | |
| Unknown | 5 (2) | 3 (3) | 2 (1) | .44 a |
| Operation | ||||
| Yes | 125 (50) | 46 (47) | 79 (52) | .52 |
| No | 125 (50) | 51 (53) | 74 (48) | |
| Discharge | .002 | |||
| Home | 69 (28) | 15 (15) | 54 (35) | |
| Rehab | 70 (28) | 35 (36) | 35 (23) | |
| Deceased | 111 (44) | 47 (48) | 64 (42) |
Table 2. Frequency of clinical and social characteristics stratified by intention. GCS = Glasgow coma scale.
aFisher’s Exact test was applied due to low counts.
bN = 221, Suicide n = 85, Assault n = 136. Percentage calculations based on patients with reported race group.
cN = 243, Suicide n = 94, Assault n = 149. Percentage calculations based on patients with documented ZIP code information.
On multivariable logistic regression analysis, age and race were significant predictors of a patient being a victim of assault, as opposed to suicide attempt (Table 3). By age group, victims of assault were more likely to be between ages 19 and 33 (OR 4.71, CI 95% 1.87–11.82, P = 0.001) and 18 or younger (OR 5.26, CI 95% 1.40–19.76, P = 0.01). Assault victims were more likely to be Black (OR 6.66, CI 95% 2.42–18.29, P < .001), Latinx (OR 2.65, CI 95% 1.07–6.54, P = 0.03), or “other” race groups (OR 6.27, CI 95% 1.26–31.1, P = 0.03). Among those who present with a GSW to head, the predicted probability of the injury being assault-related is 77% if between ages 19 and 33, 88% if Black, and 93% if of the poor or near-poor income class.
Table 3.
Multiple Regression Analysis; Predictors of Assault (N = 250, 100%).
| Risk factor | Odds ratio (95% CI) | P |
|---|---|---|
| Age: Reference: ≥ 49 | ||
| 34–48 | 1.66 (.62, 4.44) | .31 |
| 19–33 | 4.71 (1.87, 11.82) | .001 |
| ≤18 | 5.26 (1.40, 19.79) | .01 |
| Race a : Reference: White | ||
| Black | 6.66 (2.42, 18.29) | <.001 |
| Latinx | 2.65 (1.07, 6.54) | .03 |
| Other | 6.27 (1.26, 31.1) | .03 |
| Income b : Reference: Upper-middle | ||
| Middle | .35 (.04, 2.99) | .33 |
| Lower-middle | .43 (.05, 3.79) | .45 |
| Poor, near-poor | .81 (.05, 12.2) | .88 |
Table 3. Multiple regression analysis of socioeconomic predictors of assault.
aN = 221, Suicide n = 85, Assault n = 136. Odds ratio calculations based on patients with reported race group.
bN = 243, Suicide n = 94, Assault n = 149. Odds ratio calculations based on patients with documented ZIP code information.
The population of the state of Massachusetts is 71% white (non-Latinx), 9% Black, 12% Latinx, and 8% Asian, Alaskan Native, American Indian, or Pacific Islander. 16 The distribution of race in the local population is disproportionate to that of the population of assault-related GSW to the head victims: 34% white (non-Latinx), 31% Black, 24% Latinx, 11% “other.” When the race groups are scaled to each respective racial population in Massachusetts, the count of GSW to the head assault victims per capita (per million) is: 9 White, 68 Black, 39 Latinx, and 28 “other” (Figure 1).
Figure 1.
Count of patients with assault-related GSWs to the head by race group scaled to the respective Massachusetts state populations per 1 million. GSW = gunshot wound.
Mortality and Functional Outcomes
In total, 248 (99%) patients had a documented mRS at discharge, with 93 (38%) patients mRS 0–2 (good outcome) and 155 (63%) mRS 3–6 (poor outcome). When compared to patients with a poor outcome at discharge, the majority of patients with a good outcome suffered tangential bullet trajectories (75% vs 29%, P < .001), had reactive pupils (91% vs 53%, P < .001), underwent operative management (61% vs 43%, P = .004), and arrived with GCS 13–15 (80% vs 14%, P<.001). Most patients with a good outcome at discharge had no midline shift (52% vs 23%, P < .001), no intraventricular hemorrhage (IVH) (57% vs 32%, P < .001), unilateral hemispheric injury (95% vs 35%, P < .001), absence of multi-lobar injury (86% vs 28%, P < .001), and absence of a trans-ventricular bullet trajectory (98% vs 58%, P < .001) (Table 4).
Table 4.
Clinical and Socioeconomic Characteristics 1998–2020; Stratified by mRS at Discharge (N = 248 b , 100%).
| Good Outcome N (%) 93 (38) |
Poor Outcome N (%) 155 (63) | P | Alive N (%) 140 (56) | Dead N (%) 108 (44) | P | |
|---|---|---|---|---|---|---|
| Trajectory | <.001 | <.001 | ||||
| Perforating | 4 (4) | 25 (4) | 11 (8) | 18 (17) | ||
| Penetrating | 19 (20) | 96 (62) | 37 (26) | 78 (72) | ||
| Tangential | 70 (75) | 29 (29) | 92 (66) | 7 (6) | ||
| Unknown | 0 (0) | 5 (3) | 0 (0) | 5 (5) | ||
| Midline shift | <.001 | <.001 | ||||
| None | 48 (52) | 36 (23) | 68 (49) | 16 (15) | ||
| ≤5 mm | 8 (9) | 22 (14) | 10 (7) | 20 (19) | ||
| >5 mm | 3 (3) | 35 (23) | 11 (8) | 27 (25) | ||
| Unknown | 24 (37) | 62 (40) | 51 (36) | 45 (42) | ||
| Pupils | <.001 | <.001 | ||||
| Fixed | 4 (4) | 73 (47) | 5 (4) | 72 (67) | ||
| Reactive | 85 (91) | 82 (53) | 131 (94) | 36 (33) | ||
| Unknown | 4 (4) | 0 (0) | 4 (3) | 0 (0) | ||
| Bi-hemispheric | <.001 | <.001 | ||||
| Yes | 5 (5) | 77 (50) | 24 (17) | 58 (54) | ||
| No | 88 (95) | 54 (35) | 116 (83) | 26 (24) | ||
| Unknown | 0 (0) | 24 (15) | 0 (0) | 24 (22) | ||
| Multi-lobar | <.001 | <.001 | ||||
| Yes | 13 (14) | 87 (56) | 34 (24) | 66 (61) | ||
| No | 80 (86) | 44 (28) | 106 (76) | 18 (17) | ||
| Unknown | 0 (0) | 24 (15) | 0 (0) | 24 (22) | ||
| Trans-ventricular | <.001 | <.001 | ||||
| Yes | 1 (1) | 39 (25) | 8 (6) | 32 (30) | ||
| No | 91 (98) | 90 (58) | 131 (94) | 26 (24) | ||
| Unknown | 1 (1) | 26 (17) | 1 (1) | 26 (24) | ||
| IVH | <.001 | <.001 | ||||
| Yes | 4 (4) | 43 (28) | 11 (8) | 36 (33) | ||
| No | 53 (57) | 50 (32) | 72 (52) | 31 (29) | ||
| Unknown | 36 (39) | 62 (40) | 57 (41) | 41 (38) | ||
| Operation, yes | 57 (61) | 66 (43) | .004 | 101 (72) | 22 (20) | <.001 |
| GCS | <.001 | <.001 | ||||
| 3–8 | 13 (14) | 124 (80) | 31 (22) | 106 (98) | ||
| 9–12 | 6 (6) | 10 (6) | 15 (11) | 1 (1) | ||
| 13–15 | 74 (80) | 21 (14) | 94 (67) | 1 (1) | ||
| Intention | .02 | .21 | ||||
| Suicide | 28 (30) | 69 (45) | 50 (36) | 47 (44) | ||
| Assault | 65 (70) | 86 (55) | 90 (64) | 61 (56) | ||
| Sex | .66 | .81 | ||||
| Male | 83 (89) | 141 (91) | 127 (91) | 97 (90) | ||
| Female | 10 (11) | 14 (9) | 13 (54) | 11 (10) | ||
| Age | .13 | .24 | ||||
| ≤18 | 12 (13) | 17 (11) | 16 (11) | 13 (12) | ||
| 19–33 | 50 (54) | 69 (45) | 71 (51) | 48 (44) | ||
| 34–48 | 20 (22) | 32 (21) | 32 (23) | 20 (19) | ||
| ≥49 | 11 (12) | 37 (24) | 21 (15) | 27 (25) | ||
| Race c | .28 | .53 | ||||
| White | 36 (44) | 75 (54) | 67 (54) | 44 (47) | ||
| Black | 22 (27) | 25 (18) | 28 (22) | 19 (20) | ||
| Latinx | 15 (19) | 29 (21) | 22 (18) | 22 (23) | ||
| Other | 8 (10) | 9 (7) | 8 (6) | 9 (10) | ||
| Income class d | .11 | .06 a | ||||
| Poor, near-poor | 4 (4) | 9 (6) | 4 (3) | 9 (9) | ||
| Lower-middle | 45 (49) | 60 (40) | 62 (45) | 43 (42) | ||
| Middle | 37 (40) | 77 (52) | 64 (46) | 50 (49) | ||
| Upper-middle | 6 (7) | 3 (2) | 8 (6) | 1 (1) |
Table 4. Frequency of clinical and neurosurgical characteristics at discharge stratified by outcome. IVH = intraventricular hemorrhage, GCS = Glasgow Coma Scale.
aFisher’s Exact test was applied due to low counts.
bN = 248. Percentage calculations excluded 2 patients who did not have a reported outcome in medical record.
cN = 219. Good Outcome n = 81, Poor Outcome n = 138; Alive n = 125, Dead n = 94. Percentage calculations based on patients with reported race group and outcome.
dN = 241. Good Outcome n = 92, Poor Outcome n = 149; Alive n = 138, Dead n = 103. Percentage calculations based on patients with documented zip code information and reported outcome.
Of the 248 patients, 140 (56%) were alive and 108 (44%) were dead at the time of discharge. Most non-surviving patients, as compared to those who lived, suffered penetrating (72% vs 26%, P < .001), bi-hemispheric (54% vs 17%, P < .001), multi-lobar (61% vs 24%, P < .001), and trans-ventricular (30% vs 6%, P < .001) injuries. Though 42% (n = 45) of deceased patients had an unknown midline shift, 25% (n = 27) had a midline shift >5 mm, compared to the 8% (n=11) of surviving patients with a shift of that degree (P < .001). Twenty percent (n = 22) of deceased patients underwent an operation, while 72% (n = 101) of surviving patients did (P < .001). Most of the deceased patients had fixed pupils (67% vs 4%, P < .001), IVH (33% vs 8%, P < .001), and a GCS between 3 and 8 (98% vs 22%, P < .001) (Table 4).
For the entire cohort, functional outcome at discharge did not differ significantly with regard to sex, age, race, and income class. A larger majority (n = 65, 70%) of the patients with a good outcome at discharge were victims of assault, while only about half (n = 86, 55%) of those with a poor outcome suffered an assault-related GSW to the head (P = 0.02). Sex, age, race, income, and intention did not vary significantly between the surviving and deceased patients at discharge (Table 4).
Patients with a poor outcome at discharge were significantly more likely to have bi-hemispheric injuries (OR 11.06, CI 95% 2.60–46.95, P = 0.001) and a midline shift >5 mm (OR 5.91, CI 95% 1.23–28.39, P = 0.03). These patients with a poor outcome were also less likely to have reactive pupils (OR .11, CI 95% .03–.47, P = 0.003). Deceased patients were similarly more likely to have bi-hemispheric injuries (OR 4.33, CI 95% 1.19–15.80, P = 0.03) and were less likely to have reactive pupils (OR .07, CI 95% .02–.23, P < .001). Additionally, deceased patients were significantly less likely to undergo an operation (OR .15, CI 95% .02–.23, P < .001). Intraventricular hemorrhage, bullet trajectory, intention, and multi-lobar and trans-ventricular injuries did not significantly predict outcome (Table 5).
Table 5.
Multiple Regression Analysis of Significant Variables (P < .05); Stratified by mRS at Discharge (N = 248 a , 100%).
| Risk factor | Poor outcome 155 (63) | P | Dead 108 (44) | P |
|---|---|---|---|---|
| Odds ratio (95% CI) | Odds ratio (95% CI) | |||
| Bi-hemispheric: Reference: No | ||||
| Yes | 11.06 (2.60, 46.95) | .001 | 4.33 (1.19, 15.80) | .03 |
| Multi-lobar: Reference: No | ||||
| Yes | .65 (.16, 2.57) | .54 | .98 (.24, 4.04) | .98 |
| Trans-ventricular: Reference: No | ||||
| Yes | 4.28 (.40, 46.03) | .23 | 1.94 (.43, 8.71) | .39 |
| Unknown | .57 (.03, 12.12) | .72 | 4.00 (.01, 1232.8) | .64 |
| IVH: Reference: No | ||||
| Yes | 2.74 (.59, 12.72) | .20 | 1.12 (.25, 5.00) | .88 |
| Unknown | 2.79 (.77, 10.06) | .12 | .50 (.07, 3.54) | .49 |
| Trajectory: Reference | ||||
| Perforating | 1.46 (.31, 6.87) | .63 | 1.56 (.39, 6.31) | .53 |
| Tangential | .82 (.15, 4.37) | .82 | .41 (.07, 2.30) | .31 |
| Midline shift: Reference: None | ||||
| ≤5 mm | 1.31 (.32, 5.40) | .71 | 3.39 (.69, 16.62) | .13 |
| >5 mm | 5.91 (1.23, 28.39) | .03 | 2.33 (.57, 9.52) | .24 |
| Unknown | .78 (.21, 2.93) | .72 | 2.19 (.34, 14.23) | .41 |
| Pupils: Reference: Fixed | ||||
| Reactive | .11 (.03, .47) | .003 | .07 (.02, .23) | <.001 |
| Operation: Reference: No | ||||
| Yes | 2.02 (.84, 4.83) | .12 | .15 (.06, .43) | <.001 |
| Intention: Reference: Suicide | ||||
| Assault | .55 (.24, 1.26) | .16 | 1.11 (.38, 3.26) | .85 |
Table 5. Multiple logistic regression analysis of significant socioeconomic and clinical characteristics at discharge.
N = 248. Odds ratio calculations excluded 2 patients who did not have a reported outcome in medical record.
Discussion
In the present report, across the entire cohort of patients presenting with GSWs to the head, age, race, and income status were not independent predictors of mortality or functional outcome at discharge, possibly indicating standardization of neurosurgical and neurocritical care. Specific to patients presenting as a result of an assault-related injury, however, we demonstrate that age and race are critical determinants of being a victim of an assault-related GSW to the head. While previous studies have revealed clear racial and social disparities in prevalence and outcome after gun-related violence generally, 17 to our knowledge this is the first report in a neurosurgical cohort to highlight age, race group, and income status with GSW to the head assault specifically. While assault-related GSWs to the head mostly involve young Black or Latinx men of lower socioeconomic status, self-inflicted injuries were largely seen in older white men.
Chiu et al explored racial disparities in the inpatient management and outcome of 333 GSW to the head victims using the Nationwide Inpatient Sample. 18 They found no significant variation among racial/ethnic groups with respect to mortality, length of hospital stay, or rate of surgical intervention after GSW to the head and conclude that racial disparities may not be as prevalent as previously thought, although age and income status were not studied. Similar to their findings, we did not find that age, race, or socioeconomic status significantly predicted mortality or mRS at discharge in our population.
Asemota et al. investigated racial and insurance-based disparities in access to post-acute inpatient rehabilitation services after traumatic brain injury (TBI). 19 Among 307 675 TBI survivors from the Nationwide Inpatient Sample, they found that insured Black, Latinx, and Asian patients had reduced odds of discharge to rehabilitation services compared to the insured white patient group. We also found that significantly more white patients were discharged to a rehabilitation center compared to any other group, though GCS did not differ at presentation and race groups did not predict outcome at discharge. In future studies, we may consider adding insurance type to the socioeconomic variables when evaluating the discharge and outcome patterns of patients with GSW to the head.
Maragkos et al. conducted a meta-analysis of 1774 GSW to the head patients and demonstrated that increasing age, suicide attempt, lower admission GCS, bilateral dilated pupils, dural penetration, and bi-hemispheric and multi-lobar injury were associated with increased mortality. 13 In our analysis, bullet trajectory, midline shift, fixed and dilated pupils, IVH, operative management, low admission GCS, and bi-hemispheric, multi-lobar, and trans-ventricular trauma were all predictors of a poor outcome and death at the time of discharge. As an individual’s outcome is largely determined by the extent of cerebral damage at the time of injury and neurological status on arrival to the hospital, the greatest and most impactful intervention it seems is prevention of the GSW to the head.
Gun violence victimization is thought to be due to individual, situational, and community risk factors. 20 In California, Latinx individuals were more likely to be victims of both gang and non-gang related gun violence, as compared to other racial groups. 21 In addition, according to the CDC, Black men comprise 6% of the country but account for 51% of gun-related homicides. 22 Situational factors, such as the presence of guns within one’s neighborhood, or community factors, such as population density and income equality represent areas of potential intervention and change. Poverty creates a social divide. 23 Gun violence for acquisition and protection of person and/or property may partially explain why assault is higher in poorer neighborhoods. 17 Social epidemiologists studied 13 060 firearm-related deaths in 48 states in 2015 and determined that gun violence is often rooted in social causes, such as income inequality.24,25 It has been demonstrated that only 5% of city blocks and street corners in Boston experience 74% of all gun-related assaults and that 75% of these assaults were from less than 1% of the youth (ages 15–24) population. 26 Even within high-risk populations, gun violence is intensely concentrated, which may represent an opportunity for targeted and integrated gun-violence reduction strategies. In addition to being from poorer neighborhoods, many patients in our GSW to the head assault cohort were young. The Peterborough Adolescent and Young Adult Development Study at the University of Cambridge Institute of Criminology longitudinally tracked more than 700 young people for 5 years. 27 The researchers determined that a combination of environmental factors and personality characteristics, such as impulsivity, were the largest indicators for youth criminal behavior. Some authors believe that the biological and neuropsychological underpinnings of youth make this population more susceptible to criminal activity. 28
This study has limitations. Data were derived from 2 large tertiary care hospitals in a single city and, therefore, may not be generalizable to other settings. It is likely that many victims, regardless of intention, were found dead upon arrival of emergency medical services and, therefore, were not taken to a hospital. In addition, there are several other Level I trauma centers in the Boston metro region, but those data are not included. The rates of WLST were not included despite often occurring in patients with severe traumatic brain injury, which proposes a limitation on the assumptions associated between the clinical presentation and patient outcome. When assessing for predictors of patient outcome, the model is at risk for overfitting given the large number of predictors and limited sample size, which we can validate for other datasets in the future. Finally, racial groups aside from white, Black, and Latinx may be underrepresented. Due to the nature of the statistical analysis and stratification by racial group, Asian patients, for instance, were grouped into the “other” category in order to apply consistent categorical analyses.
Conclusion
Age and race are important determinants of being a victim of assault by GSW to the head, and our data demonstrate that this patient population is often comprised of individuals who are young, male, Black or Latinx, and of lower socioeconomic status. However, these variables do not appear to significantly impact mortality rates or functional outcomes at discharge, perhaps suggesting standardization of neurosurgical care. Instead, mortality and functional outcomes are largely determined by specific clinical and radiographic factors, as previously described. Moving forward, these data can be used to broaden the discussion of gun violence in America within neurosurgery. Future studies can explore potential disparities in post-injury care, recovery, and access to resources and further investigate the intersectionality of gun violence, gun commerce, and political events to assess any temporal trends.
Supplemental Material
Supplemental Material, sj-pdf-1-nho-10.1177_19418744221077552 for Racial and Social Determinants of Civilian Gunshot Wounds to the Head by Myron L. Rolle, Rachel M. McLellan, Pranav Nanda, Aman B. Patel, Chana A. Sacks, Peter T. Masiakos and Christopher J. Stapleton in The Neurohospitalist
Declaration of Conflicting Interests: The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.
Supplemental material: Supplemental material for this article is available online.
ORCID iDs
Myron Rolle https://orcid.org/0000-0002-1746-2522
Christopher Stapleton https://orcid.org/0000-0003-4805-0093
References
- 1.Firearm violence prevention|violence prevention|injury center|CDC. https://www.cdc.gov/violenceprevention/firearms/fastfact.html. Published June 2, 2020. Accessed February 15, 2021.
- 2.Guns in the United Kingdom—Firearms, gun law and gun control. https://www.gunpolicy.org/firearms/region/united-kingdom. Accessed February 15, 2021.
- 3.Guns in Sweden—Firearms, gun law and gun control. https://www.gunpolicy.org/firearms/region/sweden. Accessed February 15, 2021.
- 4.Firearm-related violent crime in Canada. https://www150.statcan.gc.ca/n1/pub/85-005-x/2018001/article/54962-eng.htm. Accessed February 15, 2021.
- 5.Guns in Australia—Firearms, gun law and gun control. https://www.gunpolicy.org/firearms/region/australia. Accessed February 15, 2021.
- 6.Grinshteyn E, Hemenway D. Violent death rates: The US compared with other high-income OECD countries, 2010. Am J Med. 2016;129(3):266-273. doi: 10.1016/j.amjmed.2015.10.025. [DOI] [PubMed] [Google Scholar]
- 7.Nance ML, Branas CC, Stafford PW, Richmond T, Schwab CW. Nonintracranial fatal firearm injuries in children: Implications for treatment. J Trauma. 2003;55(4):631-635. doi: 10.1097/01.TA.0000035090.99483.0A. [DOI] [PubMed] [Google Scholar]
- 8.Gressot LV, Chamoun RB, Patel AJ, et al. Predictors of outcome in civilians with gunshot wounds to the head upon presentation. J Neurosurg. 2014;121(3):645-652. doi: 10.3171/2014.5.JNS131872. [DOI] [PubMed] [Google Scholar]
- 9.Neurosurgical treatment for gunshot wound head trauma. https://www.aans.org/Accessed February 15, 2021.
- 10.Spitzer SA, Staudenmayer KL, Tennakoon L, Spain DA, Weiser TG. Costs and financial burden of initial hospitalizations for firearm injuries in the United States, 2006-2014. Am J Publ Health. 2017;107(5):770-774. doi: 10.2105/AJPH.2017.303684. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Who Pays for Gun Violence? John Jay college research and evaluation center. https://johnjayrec.nyc/2020/05/11/whopays/. https://johnjayrec.nyc/2020/05/11/whopays/Published May 11, 2020. Accessed February 15, 2021.
- 12.Armstrong M, Carlson J. Speaking of trauma: The race talk, the gun violence talk, and the racialization of gun trauma. Palgrave Commun. 2019;5(1):1-11. doi: 10.1057/s41599-019-0320-z. [DOI] [Google Scholar]
- 13.Maragkos G, Papavassiliou E, Stippler M, Filippidis A. Meta-analysis of functional outcomes in 5508 patients sustaining a gunshot wound to the head. Neurosurg. 2019;66(nyz310_832). doi: 10.1093/neuros/nyz310_832. [DOI] [Google Scholar]
- 14.Where do I fall in the American economic class system? US news & world report. https://money.usnews.com/money/personal-finance/family-finance/articles/where-do-i-fall-in-the-american-economic-class-system. Accessed February 15, 2021.
- 15.NW 1615. Suite 800Washington, inquiries D 20036USA202-419-4300|M-857-8562|F-419-4372|M . Defining generations: Where millennials end and generation Z begins. Pew Research Center. https://www.pewresearch.org/fact-tank/2019/01/17/where-millennials-end-and-generation-z-begins/Accessed April 6, 2021. [Google Scholar]
- 16.U.S. Census bureau quickFacts: Massachusetts. https://www.census.gov/quickfacts/fact/table/MA/AGE295219. Accessed February 15, 2021.
- 17.Zebib L, Stoler J, Zakrison TL. Geo-demographics of gunshot wound injuries in Miami-Dade county, 2002-2012. BMC Publ Health. 2017;17(1):174. doi: 10.1186/s12889-017-4086-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Chiu RG, Fuentes AM, Mehta AI. Gunshot wounds to the head: Racial disparities in inpatient management and outcomes. Neurosurg Focus. 2019;47(5):E11. doi: 10.3171/2019.8.FOCUS19484. [DOI] [PubMed] [Google Scholar]
- 19.Asemota AO, George BP, Cumpsty-Fowler CJ, Haider AH, Schneider EB. Race and insurance disparities in discharge to rehabilitation for patients with traumatic brain injury. J Neurotrauma. 2013;30(24):2057-2065. doi: 10.1089/neu.2013.3091. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Papachristos AV, Wildeman C. Network exposure and homicide victimization in an African American community. Am J Publ Health. 2014;104(1):143-150. doi: 10.2105/AJPH.2013.301441. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Aryan HE, Jandial R, Bennett RL, Masri LS, Lavine SD, Levy ML. Gunshot wounds to the head: Gang-and non-gang-related injuries and outcomes. Brain Inj. 2005;19(7):505-510. doi: 10.1080/02699050400005143. [DOI] [PubMed] [Google Scholar]
- 22.Statistics|Giffords. https://giffords.org/lawcenter/gun-violence-statistics/Accessed February 15, 2021
- 23.Dale JG. Poverty and power: The problem of structural inequality. Contemp Soc J Rev. 2010;39(1):82-83. doi: 10.1177/0094306109356659uu. [DOI] [Google Scholar]
- 24.Noonan D. Gun homicide linked to poor social mobility. Scientific American. https://www.scientificamerican.com/article/gun-homicide-linked-to-poor-social-mobility/Accessed February 15, 2021. [Google Scholar]
- 25.Waldman M. The Second Amendment: A Biography. New York: Simon & Schuster, 2014. [Google Scholar]
- 26.Braga AA, Hureau DM, Papachristos AV. Deterring gang-involved gun violence: Measuring the impact of boston’s operation ceasefire on street gang behavior. J Quant Criminol. 2014;30:113-139. doi: 10.1007/s10940-013-9198-x. [DOI] [Google Scholar]
- 27.The Peterborough Adolescent and Young Adult Developmental Study (PADS+) . Design overview—centre for analytic criminology. https://www.cac.crim.cam.ac.uk/research/padspres. https://www.cac.crim.cam.ac.uk/research/padspres February 15, 2021. Accessed
- 28.Steinberg L, Cauffman E, Woolard J, Graham S, Banich M. Are adolescents less mature than adults?: Minors’ access to abortion, the juvenile death penalty, and the alleged APA “flip-flop”. Am Psychol. 2009;64(7):583-594. doi: 10.1037/a0014763. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplemental Material, sj-pdf-1-nho-10.1177_19418744221077552 for Racial and Social Determinants of Civilian Gunshot Wounds to the Head by Myron L. Rolle, Rachel M. McLellan, Pranav Nanda, Aman B. Patel, Chana A. Sacks, Peter T. Masiakos and Christopher J. Stapleton in The Neurohospitalist