A Prospective US National Trauma Center Study of Firearm Injury Survivors Weapon Carriage and Posttraumatic Stress Disorder Symptoms

Abstract

Objective:

To describe the demographic, injury-related and mental health characteristics of firearm injury patients and trace firearm weapon carriage and posttraumatic stress disorder (PTSD) symptoms over the year following injury.

Summary and Background Data:

Based on the increasing incidence of firearm injury and need for novel injury prevention strategies, hospital-based violence intervention programs are being implemented in US trauma centers. There is limited data on the long-term outcomes and risk behaviors of firearm injury survivors to guide this work.

Methods:

We conducted a secondary analysis of a pragmatic 25-trauma center randomized trial (N=635). Baseline characteristics of firearm-injured patients (N=128) were compared with other trauma patients. Mixed model regression was used to identify risk factors for post-injury firearm weapon carriage and PTSD symptoms.

Results:

Firearm injury patients were younger and more likely to be black, male and of lower socioeconomic status and more likely to carry a firearm in the year prior to injury. Relative to pre-injury, there was a significant drop in firearm weapon carriage at 3- and 6–months post-injury, followed by a return to pre-injury levels at 12-months. Firearm injury was significantly and independently associated with an increased risk of post-injury firearm weapon carriage (Relative Risk=2.08,95% CI[1.34, 3.22],p <0.01) and higher PTSD symptom levels (Beta=3.82,95% CI[1.29, 6.35],p < 0.01).

Conclusions:

Firearm injury survivors are at risk for firearm carriage and high PTSD symptom levels post-injury. The significant decrease in the high-risk behavior of firearm weapon carriage at 3–6 months post-injury suggests that there is an important post-injury “teachable moment” that should be targeted with preventive interventions.

Trial Registration:

ClinicalTrials.gov NCT02655354

MINI-ABSTRACT

Survivors of firearm injury are significantly more likely to engage in the high-risk behavior of firearm weapon carriage and to struggle with PTSD symptoms post-injury. There is, however, a significant drop in firearm weapon carriage at 3- and 6-months post-injury, which, represents an important “teachable moment” for future targeted interventions.

INTRODUCTION

Individual and mass firearm injuries constitute a major public health challenge in the United States.^1–4 Annually, approximately 32,000 individuals die from firearm injuries, and over 67,000 individuals are so severely injured that they seek attention in acute care medical emergency department settings. Over half of these 67,000 individuals require inpatient hospitalization.⁵ In the wake of a firearm injury, survivors are at high risk for the development of posttraumatic stress disorder (PTSD) and related comorbid conditions.^6,7 Firearm injury survivors also have high rates of emergency department and hospital recidivism, including frequent recurrent firearm-related visits. ^1,8–10 High risk behaviors such as firearm weapon carriage and mental health symptoms such as PTSD have been associated with marked functional impairment, high risk of re-injury and increased societal costs after firearm injury.^6–9,11,12

Previously, legislative barriers have made scientific inquiry into primary and secondary preventive interventions for firearm injury challenging to implement, resulting in a substantive research to practice gap.^6,13–17 Recent events, including the COVID-19 pandemic, increasing US firearm sales and social unrest, highlight the need for research that catalyzes the development of primary and secondary preventive interventions.^4,18–21 A series of investigations in adult patients have assessed the impact of hospital-based programs for firearm injury survivors, yet the results of these studies are mixed, and many of these investigations are not optimally designed as preventive intervention trials. ^22–25 A smaller number of well-designed randomized preventive intervention trials with adolescent patients in acute care medical settings suggest that motivational interviewing interventions can effectively target reductions in high risk injury behaviors, including peer violence and weapon carriage.^11,26 Few comprehensive multisite investigations exist, however, that directly inform the development of hospital-based firearm injury prevention programs.

In this prospective study, first we aim to describe the demographic, injury, and clinical characteristics firearm injury survivors admitted to 25 US Level I trauma centers in comparison with non-firearm injury patients admitted during the same time period. We trace the course of post-injury firearm weapon carriage and identify baseline risk factors associated with an increased risk of post-injury firearm weapon carriage to define optimal post-injury timeframes for the delivery of preventive interventions. The investigation also seeks to identify the independent contribution of the firearm injury mechanism to the development of PTSD symptoms over the course of the year after the index injury admission.

METHODS

Design Overview

The investigation was a secondary study and analysis of data embedded within a larger pragmatic randomized trial, designed to assess the impact of a collaborative care on PTSD symptoms.²⁷ The Trauma Survivors Outcomes and Support (TSOS) pragmatic trial was orchestrated by the study team’s data coordinating center, located at University of Washington’s Harborview Medical Center, in close collaboration with the National Institutes of Health (NIH) Health Care Systems Research Collaboratory.^28,29 Sites recruited into the study constituted a representative subsample of all US Level I trauma centers.²⁸ Recruitment for the trial began in January of 2016, and 12-month patient study follow-up ended in November of 2018. Patients were assessed at baseline as inpatients and again at 3-, 6- and 12-months after injury. The Western Institutional Review Board approved the protocol prior to study initiation and informed consent was obtained for participation in the study.

Patient Inclusion/Exclusion Criteria

Survivors of intentional and unintentional injuries ≥ 18 years of age were included in the investigation. Prisoners and non-English-speaking patients were excluded. Patients whose index injury was a suicide attempt or who were psychotic and required immediate psychiatric treatment were also excluded from the trial. In order to ensure adequate follow-up rates, patients were required to provide two pieces of contact information.

Electronic Health Record (EHR) PTSD Screen

The study team had previously developed a 10-domain electronic health record (EHR) screen to detect patients at risk for the development of PTSD (Appendix 1).³⁰ Patients identified by EHR evaluation as high risk for PTSD, with a score of ≥ 3 domains positive, were then formally screened for study entry with the Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV)³¹ PTSD Checklist (PCL-C), anchored to the acute injury event.^32–34 Patients scoring ≥ 35 on the PCL-C, indicating that they had symptoms of PTSD that were at least moderate, were followed longitudinally over the course of the year after the index injury admission.

Patient-Reported Outcome Assessments

The study team attempted to contact all patients at 3-, 6- and 12-months after the index injury admissions. The follow-up interviews contained questions assessing violence risk behaviors, mental health symptoms such as PTSD, physical functioning and health services utilization. All measures had been previously administered in prior study team trials.^11,28,32 Individual assessments used in this investigation are described below.

PTSD Symptoms:

The PCL-C was used to assess the symptoms of posttraumatic stress disorder.^32–34 At baseline during the index hospitalization, patients were asked to rate their symptoms since the injury event and the 3-, 6-and 12-month interviews queried patients about their symptoms over the month prior to the interview. Prior investigations, including studies with injured patients, have established the psychometric equivalence of the DSM-IV and DSM-5 versions of the PCL-C.^31,35–37 Because the study team had validated the 10-domain EHR screen with the PCL-C for DSM-IV and DSM-IV CAPS,³⁰ all primary assessments were performed with the DSM-IV version of the PCL-C.

Firearm Weapon Carriage:

At baseline, inpatients were asked, “During the past year before your injury, did you carry a gun on you?” Patients were then again asked if they had carried a gun since their injury at the 3-, 6-, and 12-month follow-up time points.

Other Questionnaire Items:

At baseline, single items were used to assess patients’ pre-injury gun ownership and carriage of a weapon other than a firearm (e.g., knife, club). The 9-item Patient Health Questionnaire (PHQ-9) brief depression severity measure was used to assess depressive symptoms.³⁸ The Alcohol Use Disorder Identification Test (AUDIT) was used to assess alcohol use problems.^32,39 The Medical Outcomes Study Short Form Physical Components Summary Score (MOS SF PCS) SF-12 was used to assess physical function in the month prior to the injury admission.^32,40 The trauma history screen from the National Comorbidity Survey was used to assess pre-injury lifetime trauma exposures (e.g., combat, physical assaults, motor vehicle crashes), that occurred before the index injury admission.^28,41 Questionnaire items also assessed pre-injury substance use, medical conditions and health service utilization, as well as demographic characteristics (e.g., marital status).²⁸

Trauma Registry Data

Medical record data from the 25 sites’ trauma registries was used to derive injury severity scores and injury mechanism. Other clinical characteristics including insurance status, and length of hospital and intensive care unit stays, were also obtained from trauma registries.

Statistical Analysis

The investigation first described the baseline differences between firearm injury survivors and all other injured patients. Differences between the two groups were examined using the t-test for continuous variables and Chi-square statistic for categorical variables.

Next, the longitudinal course of firearm weapon carriage was examined using mixed effects regression models. Mixed effects regression was used to account for both repeated measures for individuals over time and site-level clustering of individual observations; mixed effects regression was used to identify significant variations in firearm weapon carriage at the 3-, 6- and 12-month post-injury time points, while adjusting for any potential intervention effects. Baseline variables associated with firearm weapon carriage over the course of the 12 months after the index injury hospitalization were also identified using mixed effects regression models. Baseline variables were entered into the regression model using a backwards elimination procedure. Although multiple clinical, sociodemographic, and injury (i.e., intervention and control group status, age, gender, race, insurance, employment status, marital status, level of education, comorbidities, gun ownership, self-reported baseline drug use, depression, ISS, TBI, prior ED visits/hospitalizations/mental health visits, number of serious prior traumas) characteristics were tested, only variables with significant independent associations at the p < 0.05 level (gender, intervention versus control group status, firearm injury, serious prior traumas, self-reported baseline drug use, employment status) were retained in the final model. Using a similar approach, the association between firearm injury and the development of PTSD symptoms longitudinally, was also assessed with mixed model regression. The following variables were included in the final model: intervention versus control group status, firearm injury, age, serious prior traumas, self-reported drug use at baseline, comorbidities. The study team repeated select analyses for the subgroup of patients who were firearm injury survivors (n = 128). Sensitivity analyses were conducted that assessed the impact of removing covariates, including intervention effects, on the mixed model regression results. Statistical analyses were performed using SAS Software Version 9.4 (SAS Institute Inc., Cary, NC, USA) and SPSS version 25 (SPSS Software IBM).

RESULTS

Of the 635 patients followed longitudinally in the trial, 128 (20%) were survivors of firearm injuries. The demographic and clinical characteristics of firearm-injured patients versus all other patients included in the study are presented in Table 1. Firearm injury patients were younger and more likely to be male and black and were more likely to have public insurance and a lower level of education than those with non-firearm related injuries. With regards to weapon characteristics, firearm injury patients were more likely to report carrying a gun in the year prior to their injury despite not being more likely to own a gun. There were no significant differences in rates of follow-up between firearm injury survivors compared to all other injury survivors at any time point.

TABLE 1.

Baseline Patient Characteristics^*

	No. Patients (%)
	Not firearm injury (n = 507)	Firearm injury (n = 128)	P-value
Basic Demographic
Age, mean (SD), y	40.6 (14.7)	32.7 (10.0)	<0.0001
Male gender	225 (44.4)	100 (78.1)	<0.001
Race			<0.001
White	292 (57.6)	23 (18.1)
Black	132 (26.0)	86 (67.7)
Hispanic	40 (7.9)	14 (11.0)
American Indian	13 (2.6)	2 (1.6)
Asian/Pacific Islander	6 (1.2)	2 (1.6)
Other	24 (4.7)	0 (0.0)
Insurance			0.003
Private	171 (33.7)	26 (20.3)
Public/None/VA	336 (66.3)	102 (79.7)
Marital status			0.67
Married/living with a partners	144 (28.5)	34 (26.6)
Other	362 (71.5)	94 (73.4)
Employed prior to injury	307 (60.6)	71 (55.5)	0.30
Level of Education			<0.0001
High school or less	325 (64.4)	112 (88.2)
At least some college	180 (35.6)	15 (11.8)
No. of comorbid conditions			0.003
0  1  2	140 (28.3) 92 (18.6) 77 (15.6)	52 (43.7) 25 (21.0) 14 (11.8)
≥3	185 (37.5)	28 (23.5)
Own/possess a gun	85 (17.0)	19 (15.3)	0.65
Carried a gun in the past year	38 (7.6)	19 (15.3)	0.01
Carried another weapon in the past year	121 (24.0)	33 (26.4)	0.58
Veteran	35 (7.0)	5 (4.0)	0.22
Baseline Symptoms
PTSD (PCL-C4 algorithm) positive	272 (53.7)	63 (49.2)	0.37
Depression (PHQ algorithm) positive	270 (53.4)	55 (43.0)	0.04
SF12 PCS, mean (SD)	49.0 (10.0)	50.9 (8.4)	0.03
SF12 MCS, mean (SD)	44.2 (13.5)	47.1 (12.9)	0.03
AUDIT10 positive	215 (42.4)	44 (34.4)	0.10
Pre-injury self-report drug (opioid/stimulant) use	123 (24.3)	38 (29.7)	0.21
Injury Related
ISS			0.02
0–8	115 (24.9)	22 (20.8)
9–16	156 (33.8)	51 (48.1)
>16	191 (41.3)	33 (31.1)
TBI			<0.0001
None	294 (63.6)	94 (88.7)
Mild	97 (21.0)	4 (3.8)
Moderate/Severe	71 (15.4)	8 (7.6)
Days in hospital, mean (SD)	13.0 (13.0)	12.6 (11.2)	0.82
ICU admission	282 (59.4)	65 (60.7)	0.79
No. of ED visits in past year, mean (SD)	1.7 (3.2)	1.2 (1.6)	0.03
No. of hospitalizations in lifetime, mean (SD)	0.8 (1.7)	0.6 (1.1)	0.03
Any mental health or substance use visit, lifetime	197 (39.1)	31 (24.4)	0.002
Treatment Condition			0.38
Control	291 (57.4)	79 (61.7)
Intervention	216 (42.6)	49 (38.3)
No. of serious prior traumas			0.26
0	33 (7.6)	5 (4.6)
1–3	155 (35.6)	49 (45.0)
4–6	135 (31.0)	35 (32.1)
7–8	58 (13.3)	10 (9.2)
≥9	55 (12.6)	10 (9.2)

^*All data presented as N (%) unless otherwise stated

Abbreviations: VA: Veterans Affairs; PTSD: Post-traumatic stress disorder; PCS: Physical components summary score; MCS: Mental components summary score; ISS: Injury severity score; TBI: Traumatic brain injury; ICU: Intensive care unit

The longitudinal course of firearm weapon carriage is displayed in Figure 1. Visual examination of the data revealed a drop in self-reported weapon carriage at the 3- and 6-month post-injury time points, followed by a return of weapon carriage to approximately pre-injury levels by the 12-month post-injury time point. Mixed model regression analyses quantitatively corroborated that the 3-month (β = −0.85, SE = 0.24, P < 0.01) and 6-month (β = −0.69, SE = 0.21, P < 0.001) time points, but not the 12-month (β = −0.24, SE = 0.18, P = 0.16) time point, were significantly different from baseline pre-injury weapon carriage. Weapon carriage was also examined for the subgroup of firearm injury survivors (n = 128). Firearm injury survivors demonstrated a similar pattern of a drop and recurrence of weapon carriage, albeit with elevated frequencies compared to the full sample (i.e., at baseline, 13% weapon carriage, at 3-months, 6% weapon carriage, at 6-months, 8% weapon carriage, and at 12-months, 14% weapon carriage).

Figure 1.

Percentage of study participants carrying a firearm over the 12 months after an index injury admission.

Mixed model regression identified baseline variables that were significantly associated with firearm weapon carriage over time (Table 2). Firearm injury admission was independently associated with increased post-injury firearm weapon carriage (Relative Risk=2.08, 95% CI [1.34, 3.22], p = 0.01). A history of ≥ 5 serious prior traumas before the injury event was also associated with a greater risk of carrying a firearm over the course of the year after the index injury admission. Female gender, and baseline unemployment were associated with a reduced risk of firearm weapon carriage longitudinally.

TABLE 2.

Demographic & Clinical Characteristics Associated with Weapon Carriage over the Course of the Year after Injury

Variable	Risk Ratio	95% Confidence Interval	P
Firearm injury admission	2.08	1.34 – 3.22	0.001
Female	0.46	0.29 – 0.73	0.001
Not employed at baseline	0.59	0.37 – 0.95	0.03
Randomized to intervention	0.50	0.27 – 0.93	0.03
≥5 serious prior traumas	1.62	1.06 – 2.50	0.03

Baseline firearm injury was independently associated with the development of significantly elevated PTSD symptoms (Beta=3.82, 95% CI=1.29, 6.35 p < 0.003) over the course of the year after the index injury admission (Table 3). Other factors associated with the developed of PTSD symptoms included ≥ 5 serious prior traumas, self-reported baseline substance use, ≥ 3 comorbidities and younger age.

TABLE 3.

Demographic & Clinical Characteristics Associated with PTSD Symptoms over the Course of the Year after Injury

Variable	Beta	95% Confidence Interval	P
Firearm-related index injury	3.82	1.29, 6.35	0.003
≥5 serious prior traumas	5.89	3.95, 7.82	<0.0001
Age (continuous)	−0.10	−0.18, −0.03	0.007
Self-report drug use at baseline	2.46	0.44, 4.48	0.017
≥3 medical comorbidities	5.51	3.30, 7.72	<0.0001
Intervention	2.34	−0.14, 4.83	0.064

Sensitivity analyses that removed covariates did not substantially alter the magnitude, pattern, or directionality of the observed treatment effects.

DISCUSSION

Firearm injury survivors had a significantly increased risk of firearm weapon carriage and higher PTSD symptom levels over the course of the year after injury admission compared to survivors of all other injury mechanisms. This increased risk of post-injury firearm weapon carriage and higher PTSD symptoms were observed even after other relevant demographic and clinical characteristics were accounted for.

Weapon carrying characteristics, even among firearm injury survivors, were not static. Survivors of firearm injury were less likely to carry a firearm from the time of injury to 6-months post-injury with subsequent return to baseline levels by 12 months post-injury; a finding that has important implications. This naturalistic “dip” in firearm carriage post-injury suggests that there may be an important post-injury “teachable moment”. Firearm weapon carriage is known to increase the risk of firearm-related injury⁴². There are several plausible mechanisms via which firearm weapon possession may increase an individuals’ risk of firearm injury; guns may result in a false sense of empowerment that may cause an individual to overreact, may make an individual more likely to enter dangerous environments that they may otherwise have avoided, and the gun possessed by an individuals may be turned on them during an otherwise gun-free conflict⁴². The diminishment and subsequent recurrence of weapon carriage post-injury described in this study represents a window of opportunity for preventive interventions targeting reductions in recurrent violent injury.

There exists an important analogy and precedent for this. Approximately two decades ago, prospective studies informed the development of hospital-based secondary preventive interventions targeting recurrent alcohol use injury risk behaviors.⁴³ These investigations included longitudinal investigations that documented a naturalistic “dip” and subsequent recurrence of alcohol consumption during the days and weeks immediately after an index traumatic injury admission.⁴⁴ Other efficacy, effectiveness and implementation randomized clinical trials documented sustained reductions in alcohol use among patients assigned to brief bedside motivational interviewing interventions targeting reductions in alcohol-related risk behaviors.^45,46 More recently, motivational interviewing interventions that provided “booster” sessions in the days and weeks after an initial injury admission were also shown to effectively reduce alcohol consumption.⁴⁷

At a time when many hospitals and trauma systems are looking to develop effective violence intervention programs to break the cycle of violent injury^25,48, our findings inform these efforts. Firearm-related injury has been previously been shown to be associated with an increased risk of recurrent firearm-related hospitalization, firearm-related death and firearm-related arrest.¹⁰ Re-injury, or recidivism, has become a key outcome metric for injury prevention programs.⁴⁹ To date the approach to and study of violence intervention programs has been variable and few multisite investigations exist to inform programmatic intervention development. In this study, we describe an important naturalistic “dip” in firearm carriage post-injury that lends itself to interventions aimed at maintaining decreased firearm carriage over time. Our data also informs the timing of such interventions suggesting that these efforts should be focused in the early post-injury phase. Interestingly, firearm injury patients in this study were not more likely to own a gun compared to those individuals injured as a result of another mechanism, suggesting that perhaps our focus should be on weapon carriage rather than ownership, as well as the mental health, sociodemographic and related factors that may be contributing to the high-risk behavior of firearm weapon carriage. The optimal nature of these interventions remains to be established, highlighting the need for well-designed future pragmatic intervention trials.

We also found firearm related injury to be independently associated with significantly elevated PTSD symptoms. Although the mental health outcomes of firearm injury survivors could benefit from additional investigation, there are some recent published studies demonstrating the increased mental health burden faced by survivors of firearm violence with which our findings resonate.⁵⁰ Specifically, Joseph et al⁶ reported a high readmission rate for survivors of firearm-related injury and that survivors of a semiautomatic rifle or shotgun-related injury had a higher risks of developing acute stress disorder or PTSD upon readmission compared with survivors of a handgun-related injury. Subsequently, Escobar-Herrera et al⁷ reported that compared with matched MVC survivors, firearm injury survivors were significantly more likely to screen positive for PTSD and had significantly worse mental-health related quality of life. The factors contributing to the differential mental health outcomes after firearm injury are likely multifactorial and complex. A greater perceived threat to life and the community context in which most violence-related firearm injuries occur may be key factors.⁵¹ In addition, firearm survivors often come from more vulnerable socioeconomic neighborhoods with limited social support networks, which can make access to post-discharge care and resources a challenge. The findings of the current study suggest that early trauma center based mental health service delivery should be a central focus of violence intervention programs.

This investigation has limitations. Most notably the study is a secondary analysis of a data derived from a randomized clinical trial of a broader patient population. Although all longitudinal analyses are adjusted for potential intervention effects, a prospective cohort design would have been optimal for describing the naturalistic evolution of post-injury firearm carriage.

The results of this investigation suggest that as with alcohol screening and brief intervention two decades earlier, firearm injury survivors represent a unique and vulnerable injury population, appropriate for the development of hospital-based “teachable moment” preventive interventions. Orchestrated efforts targeting policy and funding could systematically incorporate firearm preventive screening, intervention and referral trial findings into national trauma center requirements and verification criteria.