Utilization of Mental Health Services in Pediatric Patients Surviving Penetrating Trauma Resulting from Interpersonal Violence

Abstract

Background

Violent trauma has lasting psychological impacts. Our institution’s Community Violence Response Team (CVRT) offers mental health services to trauma victims. We characterized implementation and determined factors associated with utilization by pediatric survivors of interpersonal violence-related penetrating trauma.

Methods

Analysis included survivors (0–21 years) of violent penetrating injury at our trauma center (2011–2017). Injury and demographic data were collected. Nonparametric regression models determined factors associated with utilization.

Results

There was rapid uptake of CVRT (2011–2013) after which it plateaued, serving >80% of eligible patients (2017). White race and higher injury severity were associated with receipt and duration of services. In post-hoc analysis, race was found not to be associated with initial consultation, but with continued treatment.

Conclusion

Successful implementation required three years, aiding >80% of patients. CVRT is a blueprint to strengthen existing violence intervention programs. Efforts should be made to ensure that barriers to providing care, including those related to race, are overcome.

Article Summary

This study describes the utilization of mental health services amongst pediatric survivors of penetrating trauma resulting from interpersonal violence through a novel program: The Community Violence Response Team. This study reports the successful implementation of this unique program and highlights the importance of providing mental health services to vulnerable victims of violence.

Introduction

Violent injury among pediatric patients poses an ongoing challenge for trauma systems. Often index injury in this population represents just the beginning of long-term mental and physical health conditions and decreased quality of life.¹ Approximately 50% of pediatric patients with penetrating injury are discharged after their index admission with some form of physical disability.² Furthermore, a majority of violently-injured children also show symptoms of posttraumatic stress disorder (PTSD) following hospitalization.^3,4 Teenage trauma patients are estimated to carry the greatest risk of all age groups for readmission due to mental health issues,⁵ with girls, in particular, demonstrating increased risk for symptoms of PTSD.⁶ Moreover, given high rates of re-victimization^7,8 and cumulative violence exposure,^9,10 the risk for exacerbation of mental health issues is a key concern in the long-term management of violently-injured children and youth.

In this context, hospital-based violence intervention programs (HVIPs) have demonstrated a powerful provider-driven model to stem repeat violent trauma.^7,11 They may be particularly cost effective in settings with established high rates of reinjury due to violence.¹² Our institution’s existing HVIP, the Violence Intervention Advocacy Program (VIAP), provides case management support to those impacted by gunshot and stab wound injuries. Bolstering the work of our HVIP, the Community Violence Response Team (CVRT) was created in 2011 with the primary goal, rather, of providing immediate and long-term mental health services to survivors of life-threatening community violence and their families.¹³ This unique program is embedded within the Division of Trauma Surgery. These services are open to eligible patients of all ages, are non-insurance based, and free to the patients and their families. Approximately 1000 individuals are served yearly.¹³

While CVRT services are offered to eligible patients of all ages, for this investigation, we chose specifically to focus on children and youth survivors of violent penetrating trauma, given the significant post-traumatic psychosocial burden in this population.^3–6 The primary objective of this study was to characterize the implementation of CVRT services at our institution among the pediatric penetrating trauma (PPT) population. Our secondary objective was to determine what factors, if any, are associated with CVRT utilization at index admission and beyond and explore potential reasons for non-utilization.

Materials and Methods

Study Approval and Data Sources

The study was approved by the Boston University Institutional Review Board (H-37975) and chart review was conducted in compliance with HIPAA guidelines. Demographic and injury data were collected from our institution’s trauma registry, which consists of a comprehensive account of every patient who presents to our institution with a traumatic injury. Details regarding each patient’s hospital course, including past medical history, social history, insurance status, and consults received during admission, were obtained through thorough review of the electronic medical record. Finally, service-specific data was obtained from an internal CVRT database containing information regarding the patients for whom services were provided and the nature and duration of those services.

Cohort Description

Our overall cohort consisted of all patients ages 0–21 who presented with a penetrating injury to our urban level one trauma center between 2011 and 2017 (N=618). While definitions of what constitutes pediatric, child, adolescent, or youth vary widely,¹⁴ we extended the definition of “pediatric” beyond the age of majority, as many of our vulnerable patients fall into this older youth category. There is precedent, particularly for social programming, to extend services through the age of 21.¹⁵ Additionally, research has shown that under threat conditions, 18–21 year-old youth respond more similarly to teenagers than to those in their later 20s.¹⁴ Patients determined to have sustained an accidental penetrating trauma or a self-inflicted injury were excluded from analysis, as these patients are not eligible to receive CVRT services. Thus, only patients who were victims of interpersonal penetrating violence and eligible to receive CVRT services were included (n=524). Of these, patients who did not survive their index hospitalization were excluded from final analysis (n=482) (Figure 1).

Figure 1:

Flowchart of final study population selection. Those with injuries due to unintentional causes or self-harm were excluded. Additionally, those who were found to be dead on arrival (DOA) or expired in the Emergency Department were excluded. 482 patients were included for analysis.

Community Violence Response Team

Currently, CVRT consists of a dedicated team of four full-time, Masters-prepared, mental health clinicians and two full-time, Masters-prepared, clinical leaders. Ongoing federal grant funding through the Victims of Crime Act (VOCA) currently makes this possible. The team is directly embedded in regular trauma care and liaises with clinical social workers, case managers, our HVIP, and the trauma surgery team. On a daily basis, all trauma admissions and emergency department (ED) discharges are screened for appropriateness for CVRT services and, if deemed eligible, are assigned to a mental health clinician. Patients eligible include all victims of gunshot and stab injuries as well as life-threatening blunt assaults. The patient and, if applicable, family members, are approached at bedside within 24 hours of admission and provided with information about the program, offered services, and informed that all services are free. For patients discharged from the ED, the team reaches out by telephone within 24 hours of discharge. If the patient decides to make use of the services offered, the mental health clinician follows up daily with services that include therapeutic rapport building, psychoeducation, therapy, and advocacy among others. In both cases, the goal is long-term therapeutic engagement. This may include onsite outpatient therapy, telehealth, and outreach therapy in the client’s home or safe community space. The frequency of services is tailored to individual needs and is flexible given that it is not insurance-based. For those potential clients who initially are not interested in mental health services, the team continues to check in every other week to ensure that their needs are being met. This is continued until the patient asks for these check-ins to stop or they choose to engage in further treatment.

Outcomes

Our main outcome of interest was the receipt of CVRT services, which was analyzed as a dichotomous variable. Receipt of CVRT services was defined as having an initial meeting with patients (and family members), in person or by phone, where services were discussed and offered, and opportunities for long-term provision of mental health services were provided. Whether a patient received services was determined from CVRT’s internal database. This database also contains the number of quarters in which a patient or client received services. This was transformed into duration of services measured in days.

Additional Variables

The demographic variables measured included age, sex, race, and ethnicity. We chose to dichotomize age for analysis (<18 vs ≥18) to account for the potential influence by parents or guardians in decisions to have minors engage in mental health services. Additionally, race was dichotomized for analysis (White vs non-White). Injury characteristics included injury severity score (ISS) and injury type (gunshot wound (GSW) and stab wound (SW)). Variables related to hospital course included admission year, possession and type of insurance, presence of a psychiatric consult, and presence of a social work consult. Medical history variables included current or previous diagnosis of mental illness and current or previous history of substance abuse.

Statistical Analysis

All statistical analysis was performed using STATA 16.0. First, we determined the frequency of CVRT utilization amongst those in the included cohort overall and by admission year. Differences in proportion of patients utilizing CVRT services by admission year were determined using a chi-squared test. Second, we calculated the geometric mean duration of utilization of CVRT services (due to skewed distribution) overall and by admission year. Third, we performed a local-linear kernel nonparametric regression and estimated standard errors and confidence intervals with bootstrapping (100 replications)¹⁶ to determine the association between receiving CVRT services and each covariate. The same was then repeated with the duration of services. Fourth, we performed two multivariable nonparametric regressions, one to determine the independent association between all relevant covariates and CVRT service utilization and another to determine the independent association between all relevant covariates and duration of services. Finally, we performed a post-hoc multivariable nonparametric regression with social work utilization as the outcome. We determined that all patients receiving CVRT would also have been evaluated by social work. Accordingly, social work consultation in this cohort would be indicative of initial engagement with the patient. Thus, using social work as the outcome allowed us to determine if CVRT non-utilization was the result of lack of initial engagement or a failure of longer term patient involvement in the program.

Results

Overall Cohort

The overall cohort consisted of 482 patients (Table 1). The mean age was 18.65 years and the majority were male (89.00%), and Black (69.94%), and uninsured (53.11%). Of those insured, more patients had public insurance (31.12%) than private insurance (15.77%). Most patients violently injured suffered GSWs (66.18%). A medical history of mental illness was present in 19.09% of patients and 67.01% had a history of substance use.

Table 1:

Demographics of analyzed cohort n=482.

GSW=gunshot wound, SW=stab wound, ISS=injury severity score.

	Analyzed Cohort
N	482
Age, mean (SD)	18.65 (2.62)
Male sex, n (%)	429 (89.00)
Race, n (%)
White	32 (6.82)
Black	328 (69.94)
Asian	10 (2.13)
Other	99 (21.11)
Hispanic, n (%)	98 (21.92)
Injury
GSW	319 (66.18)
SW	163 (33.82)
ISS, median (IQR)	4.00 (9.00)
Baseline Mental Illness, n (%)	92 (19.09)
Substance Use, n (%)	323 (67.01)
Insurance Type, n(%)
Public	150 (31.12)
Private	76 (15.77)
None	256 (53.11)

CVRT Utilization

Across all years, CVRT utilization among those eligible to receive services was 72.82%. There was rapid uptake of the program from 2011 to 2013 (38.27% to 82.22%), which then plateaued from 2014–2017 at greater than 75% (Figure 2). The frequency of utilization across years differed significantly (p<0.0001). In bivariate models, later admission year was positively associated with receipt of CVRT services [β=0.0872, 95% CI (0.0478, 0.1166), p<0.0001]. Additionally, there was a negative association with sustaining a SW (vs. GSW) and receiving CVRT services [β=−0.0352, 95% CI (−0.0698, −0.0048), p=0.035]. In our multivariate nonparametric model later admission year [β=0.0517, 95% CI (0.0191, 0.0788), p=0.001]and White race [β=0.1541, 95% CI (0.03325, 0.2665), p=0.009] remained significantly positively associated with utilization of CVRT services (Table 2).

Figure 2:

Implementation of CVRT from 2011–2017. Data points represent proportion of eligible patients who received CVRT services (%). The dotted line represents the line of best fit with associated confidence interval.

Table 2:

Bivariate and multivariate nonparametric regression model for receiving CVRT services.

Reference value for admission year: 2011. Age category refers to patients <18 vs those ≥18 (reference value: <18). Reference value for sex: female. Race is dichotomized into white vs non-white (reference value: white). Reference value for injury type: stab wound. Reference value for baseline mental illness: positive history of mental illness. Reference value for substance use: positive history of substance use. Reference value for insurance type: no insurance.

	Bivariate Model		Multivariate Model
	β - coefficient (95% CI)	p-value	β-coefficient (95% CI)	p-value
Admission Year	0.0872 (0.0478, 0.1166)	<0.0001	0.0517 (0.0191, 0.0788)	0.001
Age category	0.0032 (−0.012, 0.017)	0.667	0.06142 (0.0026, 0.1693)	0.177
Sex	0.0007 (−0.0081, 0.0099)	0.881	−0.0078 (−0.1493, 0.1349)	0.916
White	0.0124 (−0.00754, 0.0306	0.215	0.1541 (0.03325, 0.2665)	0.009
Ethnicity	−0.0110 (−0.0328, 0.0099)	0.300	−0.03395 (−0.1661, 0.0552)	0.547
Injury Type	−0.0352 (−0.0698, −0.0048)	0.035	−0.0753 (−0.2020, 0.0038)	0.159
ISS	0.0076 (−0.0094, 0.0221)	0.339	.0077 (−0.003, 0.0159)	0.063
Baseline Mental Illness	<0.0001 (−0.0010, 0.0162)	0.999	−0.0336 (−0.1512, 0.0793)	0.547
Substance Use	0.0049 (−0.127, 0.0172)	0.555	0.0286 (−0.0520, 0.1181)	0.536
Insurance Type: Public	−0.0099 (−0.0245, 0.0061)	0.245	0.0330 (−0.0327, 0.0897)	0.325
Insurance Type: Private	−0.0052 (−0.0177, 0.01244)	0.476	0.0701 (−0.0607, 0.1776)	0.244

Duration of Services

The mean duration of CVRT services over the study period was 140.70 days. In bivariate models, while admission year was not, higher ISS was associated with longer duration of services received [β=6.7577, 95%CI (1.9832, 11.5685), p=0.005]. Additionally, sustaining a SW was associated with a significantly shorter duration of services [β=−15.0031, 95%CI (−26.3099, 1.2100), p=0.035]. In our multivariable nonparametric model, only a higher ISS was associated with longer duration of services [β=7.5640, 95% CI (4.5245, 13.3807), p=0.001] (Table 3).

Table 3:

Bivariate and multivariate nonparametric regression model of duration of CVRT services (days) after discharge.

Reference value for admission year: 2011. Age category refers to patients <18 vs those ≥18 (reference value: <18). Reference value for sex: female. Race is dichotomized into white vs nonwhite (reference value: white). Reference value for injury type: stab wound. Reference value for baseline mental illness: positive history of mental illness. Reference value for substance use: positive history of substance use. Reference value for insurance type: no insurance.

	Bivariate Model		Multivariate Model
	β - coefficient (95% CI)	p-value	β-coefficient (95% CI)	p-value
Admission Year	6.3844 (−8.3670, 19.7844)	0.361	−3.34 (−20.3879, 4.9102)	0.602
Age category	0.4335 (−5.993, 6.9363)	0.897	12.94 (−17.8471, 61.7636)	0.542
Sex	−1.8861 (−5.0694, 2.0389)	0.282	−10.8761 (−65.4317, 48.8934)	0.707
White	9.5505 (−2.7253, 27.8332)	0.264	75.6919 (−32.6524, 160.6468)	0.108
Ethnicity	−1.1495 (−7.1538, 6.1988)	0.768	25.9343 (−32.2472, 82.2738)	0.361
Injury Type	−15.0031 (−26.3099, 1.2100)	0.035	−41.2218 (−99.7042, 32.3812)	0.149
ISS	6.7577 (1.9832, 11.5685)	0.005	7.5640 (4.5245, 13.3807)	0.001
Baseline Mental Illness	2.8461 (−2.6464, 11.0843)	0.463	25.5797 (−27.3103, 78.5572)	0.329
Substance Use	−0.2012 (−8.0044, 8.5631)	0.961	−24.3884 (−78.8049, 25.8029)	0.326
Insurance Type: Public	−4.4996 (−10.4659, 2.6661)	0.213	−11.0543 (−32.3588, 16.1198)	0.388
Insurance Type: Private	−2.3303 (−6.8483, 4.1505)	0.426	−20.0126 (−61.3087, 41.4207)	0.452

Social Work Utilization

In bivariate models, admission year [β=0.0475, 95% CI (0.0233, 0.0773), p<0.001], ISS [β=0.0085, 95% CI (0.0054,0.0127), p<0.0001], substance use [β=0.0457, 95% CI (0.0175, 0.0752), p=0.004], and age category (<18) [β=0.0281, 95% CI (0.0041, 0.0490), p=0.01] were all positively associated with receipt of a social work consult. In multivariable nonparametric analysis of social work utilization, admission year[β=0.0294, 95% CI (0.0141, 0.0503), p=0.001], ISS[β=0.0110, 95% CI (0.0050, 0.0188), p=0.001], and substance use [β=0.1276, 95% CI (0.0371,0.1918), p=0.002] remained significantly positively associated with receiving a social work consult (Table 4).

Table 4:

Bivariate and multivariate nonparametric regression model for receiving a social work consult during hospital stay.

Reference value for admission year: 2011. Age category refers to patients <18 vs those ≥18 (reference value: <18). Reference value for sex: female. Race is dichotomized into white vs non-white (reference value: white). Reference value for injury type: stab wound. Reference value for baseline mental illness: positive history of mental illness. Reference value for substance use: positive history of substance use. Reference value for insurance type: no insurance.

	Bivariate Model		Multivariate Model
	β - coefficient (95% CI)	p-value	β-coefficient (95% CI)	p-value
Admission Year	0.0475 (0.0233, 0.0773)	<0.001	0.0294 (0.0141, 0.0503)	0.001
Age category	0.0281 (0.0041, 0.0490)	0.010	0.0677 (0.0058, 0.1373)	0.069
Sex	0.0027 (−0.0066, 0.0102)	0.528	0.0586 (−0.0676, 0.1497)	0.300
White	−0.0185 (−0.0363, 0.0004)	0.070	−0.1122 (−0.2775, 0.0426)	0.156
Ethnicity	−0.0063 (−0.0178, 0.0071)	0.289	−0.0217 (−0.1185, 0.0621)	0.666
Injury Type	−0.0181 (−0.0427, 0.0102)	0.171	−0.0488 (−0.1355, 0.0217)	0.232
ISS	0.0085 (0.0054, 0.0127)	<0.0001	0.0110 (0.0050, 0.0188)	0.001
Baseline Mental Illness	<0.001 (−0.0107, 0.01175)	0.995	−0.0085 (−0.1137, 0.0564)	0.846
Substance Use	0.0457 (0.0175, 0.0752)	0.004	0.1276 (0.0371, 0.1918)	0.002
Insurance Type: Public	0.01670 (−0.0057, 0.04122)	0.160	0.0190 (−0.0235, 0.0628)	0.398
Insurance Type: Private	0.0127 (−0.0082, 0.0313)	0.223	0.0646 (0.0007, 0.1553)	0.094

Discussion

The CVRT program is the first and, until very recently, the sole embedded mental-health focused violence intervention program in the United States. We used a retrospective observational study of pediatric patients presenting after violent penetrating trauma to our urban level 1 trauma center, from 2011–2017, to evaluate the implementation of our CVRT program and its uptake over time. We also sought to find demographic and clinical characteristics associated with utilization of CVRT services by these patients and the duration of services rendered. We found that the CVRT program had a very successful implementation period, as evidenced by its rapid uptake during the first three years and continued utilization of services thereafter by >75% of eligible pediatric patients. Additionally, we demonstrated that admission year and White race were associated with utilization of CVRT services in a multivariable model. Only a higher ISS was associated with increased duration of services. The current study shows that a program like CVRT can be successfully implemented within a high-volume trauma division as an adjunct service to provide immediate and ongoing mental health services to victims of violence.

Trauma centers are well equipped to treat violent injuries but, historically, not the social factors that lead to repeat violence.¹⁷ HVIPs aim to reduce both retaliatory injury and repeat victimization by providing intensive case management services to high-risk patients who sustain violent injury.¹¹ These patients often face numerous obstacles after being discharged that can lead to continued engagement in behaviors that increase risk of reinjury;¹¹ those injured violently are nearly twice as likely to have another violent injury requiring hospitalization within 2 years compared to those with nonviolent injuries.¹⁸ HVIPs have the goal of utilizing a unique window of opportunity to effectively engage with victims while they are initially hospitalized after their injury¹¹ and provide support in the community context following the violent event.¹⁹ Studies have demonstrated that HVIPs show promise in decreasing repeat victimization.^11,17 Traditionally, HVIPs have focused on case management and, while CVRT has institutional partnership with VIAP (our HVIP), its focus in on providing ongoing mental health services independent of insurance status to both patients and their families. The ability to provide free services allows for CVRT to forgo the need to apply a diagnosis to patients. This helps to reduce the stigma associated with seeking mental health care. Additionally, it allows CVRT to provide services outside of managed care restrictions, permitting continued engagement with clients until they are ready to utilize CVRT services. The program emphasizes the importance of addressing the lasting psychosocial impacts of trauma in addition to recognizing and working to mitigate the risk of repeat trauma.

Violent injury among this population is prevalent, with 1 in every 100 children seen in the ED presenting with violent injury.²⁰ Violence is a leading cause of death among teenagers and firearm violence in particular is the primary cause among African American children.⁷ Furthermore, the consequences of this trauma are not limited to physical injury. The majority of children victims of violent trauma show symptoms of PTSD following their initial hospitalization.^3,4 Teenagers, in particular, have the greatest risk of all age groups for readmissions for mental health issues following trauma.⁵ Providing mental health services to victims of violence, particularly children and youth, may help to alleviate some of the lasting mental health burden of penetrating injuries in this population. The creation and successful implementation of CVRT supplements the existing HVIP framework, together helping to address the need for prevention of cumulative violence exposure, which can lead to the exacerbation of mental health issues following violent trauma.^9,10

Our analysis suggests that, when race is dichotomized, White race was associated with utilization of CVRT services. Given this disparity, we performed a post hoc analysis by creating a model of social work utilization in the same population. All eligible patients who received CVRT services should also have received a social work consult during their initial hospitalization. This is indicative of initial engagement with the patient. In this analysis we found that, while ISS continued to be associated with receipt of a social work consult, race no longer was associated. This would suggest that, at our institution, while there is no disparity in providers approaching patients to offer services, barriers may exist in the delivery of ongoing mental health services following trauma among non-White patients. Different racial and ethnic groups have differing perspectives regarding when it is appropriate to seek care from mental health professionals.¹⁰ Other studies have postulated that mistrust of providers, fear of coercive treatment, and stigma surrounding mental illness may contribute to disparities in utilization of mental health services among minority patients.²¹ Based on our results, we may be failing to identify and adequately address racial and cultural barriers to the provision of CVRT services to this population.

Additionally, we found in bivariate modeling that age <18 was associated with receiving a social work consult. The association between age <18 and receipt of a social work consult also approached statistical significance in the multivariable model (additionally, 95% CI does not include the null). There is evidence to suggest that many parents are uncertain as to whether their children are in need of mental health services.¹⁰ Additionally, parents may underreport their child’s symptom severity when compared with child self-reports.³ The fact that patients <18 were more likely to receive a social work consult but not CVRT services suggests that parental decision-making in our population may play a role in whether pediatric patients receive ongoing psychosocial support following their exposure to violent penetrating injury.

A higher ISS was associated longer duration of CVRT services. In other populations, increases in ISS have been associated with increased odds of screening positive for PTSD.²² It is intuitive that resources are more likely to be focused on patients with severe injuries. However, symptoms of PTSD can manifest even when a patient has only mild or moderate physical injury.³ It is critical to recognize that, though the physical trauma may be minor, the psychosocial impacts of the event remain significant.

Limitations

This study is limited by its retrospective nature. Absence of active follow up prevents us from determining exact causes for lack of CVRT utilization among eligible patients. Given that clinical data was obtained from chart review, there is a risk of misclassification and inter-user differences in classification of variables. To combat this, definitions of included variables were standardized and all authors involved in data entry underwent the same data collection training. During the initial years of CVRT implementation, there may have been missing data given changes in electronic medical record use and documentation requirements. This may have led to underestimations with regards to social work consults received and history of substance use and pre-existing mental illness. This issue, however, does not affect the determination of frequency of CVRT use, as this data was gathered from an independent database managed by CVRT containing all relevant patient data since the launch of the program in 2011. Finally, there is a statistical limitation in the use of nonparametric models in its limited power to produce consistent estimates.²³ Bootstrapping was used to generate 95% confidence intervals around these beta coefficient estimates. Many of these are wide, indicating this lack of power in the setting of our small sample size.

Recommendations and Future Directions

Despite its limitations, this study illustrates the successful implementation of a mental health program embedded in the trauma surgery service with the primary goal to serve patients and families impacted by community violence at our urban level 1 trauma center with the goal of long-term therapeutic engagement. It provides insight into potential gaps in utilization and can inform the program moving forward as well as provide lessons to other institutions who may be interested in implementing similar programs. Given the finding that White race is associated with increased CVRT utilization, a continued emphasis on racial and cultural sensitivity should inform improvements to our existing programming. Additionally, active interviewing and follow up with patients and parents regarding reasons for declining services should take place to ensure that barriers to care are addressed. CVRT plans to implement a tablet-based anonymous survey prior to discharge to gain a deeper understanding of these barriers. This is particularly critical as communities of color are disproportionately impacted by penetrating trauma.^24–26

In this investigation, we found that higher ISS was associated with increased duration of services, however, mild physical injuries can still result in PTSD.³ To address this, the program has started screening pediatric lower level assaults and is working to connect them with community partners for ongoing mental health and case management needs. Maintaining ongoing relationships with community partners to broaden service provision in the long-term is vital to reducing the overall burden of violent trauma in each community.

Additionally, emphasis is being placed on the assessment of clinical outcomes of patients who have received CVRT services and designing prospective studies of CVRT services to further elucidate reasons for non-utilization. Formal implementation research techniques may prove useful in determining how best to translate the knowledge of the necessity of providing mental health services after violent trauma into effective programming at potential adopting institutions. In ongoing and future studies, we are interested in assessing other vulnerable patient populations who utilize CVRT services to determine the potential unique needs of specific subgroups affected by violence.

Conclusions

Providing mental services to address the hidden burden of violent trauma is critical to reducing repeat victimization, cumulative exposure to violence, and poor health outcomes, particularly in children and youth.^1,7–10 This study illustrates that, in addition to the case management services offered by traditional HVIPs, a team dedicated to the provision of mental health services embedded directly in the care of the surgical trauma patient can be successfully implemented at an urban level 1 trauma center. Additionally, the study identifies factors associated with utilization of services, including ISS and race, that should be targeted to expand utilization and duration of services for this vulnerable population.

Highlights.

• Mental Health Services program successfully implemented in 3 years

• Strengthens existing hospital-based violence intervention programs

• Race associated with continued use of services but not initial engagement

• More work is needed to ensure that barriers to providing care are overcome