Comparing Rates Mental Health Diagnosis in Adolescents Evaluated at a Community Clinic versus Detention-based Clinic: Is Traumatic Stress Still Most Salient?

Abstract

Background:

Adolescents involved in the juvenile legal system (JLS) have higher rates of mental health treatment needs compared to their non-justice-involved peers, and they experience disproportionate rates of trauma exposure and trauma-related psychosocial concerns. Most research comparing these adolescent groups draws data from separate studies, making it more challenging to understand meaningful differences between the two groups. Research documenting making such comparisons can guide prevention and intervention strategies for communities and their juvenile detention centers. The current study involves a secondary data analysis of billing claims from an outpatient community clinic with a satellite location embedded in the local juvenile detention center.

Method:

Descriptive statistics determined the top five primary diagnoses in the sample apart from Post Traumatic Stress Disorder (PTSD): Major Depressive Disorder, ADHD; Adjustment Disorder, Unspecified Trauma and Stressor Related Disorder (UTD), and Generalized Anxiety Disorder (GAD). Hierarchical logistic regression models were used to understand if being evaluated in a detention or community setting, sex, and age predicted the likelihood of having one of these top primary diagnoses.

Results:

Participants (N = 981) were adolescents ages 12-20 (M_age = 15.93 years, SD_age = 1.924, 60.6% Male) seen for psychiatric evaluation in a community mental health agency and a juvenile detention facility. Among detained adolescents (n=635), 141 were diagnosed with PTSD (22.2%), compared to 36 in the community clinic (10.4%). The odds of an adolescent in detention receiving a PTSD diagnosis were 2.5 times higher compared to adolescents evaluated in the community (p<0.001, ExpB 2.54). There was a decreased likelihood for adolescents to receive a GAD or UTD diagnosis while in detention.

Discussion:

In this sample, adolescents evaluated in detention had significantly higher odds of being diagnosed with PTSD and a lower likelihood of receiving GAD or UTD that was not PTSD. This finding supports previous literature that adolescents in detention have higher rates of PTSD than their peers in the community. It is crucial to implement evidence-based trauma treatment in detention settings, and research should continue to examine the feasibility, acceptability, and effectiveness of these interventions.

Adolescents involved with the juvenile legal system (JLS) are known to have higher rates of mental health treatment needs than adolescents in the general population, with a meta-analysis showing that 70% of JLS-involved adolescents have a diagnosable mental health problem (Vincent et al., 2008). The most common issues documented in these reviewed studies were substance use disorders, disruptive behavior disorders, anxiety, Attention Deficit Hyperactivity Disorder (ADHD), and Post-Traumatic Stress Disorder (PTSD) (Schubert et al., 2011; Schubert & Mulvey, 2014). In the past decade, trauma-related mental health concerns among adolescents in the JLS have received more attention in the literature, as these young people experience disproportionately high rates of trauma exposure and trauma-related psychosocial problems compared to their non-JLS-involved peers. Adolescents in detention have experienced exceptionally high rates of trauma exposure, with upwards of 80% of these young people having one or more traumas in their lifetime (Abram et al., 2013). They also experience a higher prevalence of PTSD than in community samples of adolescents without JLS involvement (Wasserman & McReynolds, 2011; Wood et al., 2002). Rates of PTSD in detained adolescents are up to eight times higher than in community samples of similar-age adolescents (Saigh et al., 1999). However, estimates of PTSD prevalence in the JLS vary widely (between 3% and 50%) depending on sample demographics, the assessment instrument used, and the time frame assessed (e.g., past month, past year, or at the time of diagnostic interview; Wasserman et al., 2004; Wolpaw & Ford, 2004; Ford et al., 2010; McNair et al., 2019),.

Mental health concerns can serve as precipitating factors to JLS involvement, especially those following trauma exposure. There is a significant association between early life trauma, PTSD, and delinquent behavior in adolescents (Ford et al., 2006; Lansford et al., 2007; Riggs Romaine et al., 2011). Repeated trauma exposure and reactions related to these experiences can precipitate disruptive behavior and other externalizing disorders in addition to PTSD, thus often serving as a catalyst for adolescents’ involvement in the JLS (Greenwald, 2002; Kerig & Becker, 2010; Gardner et al., 2019). For instance, the school-to-prison nexus literature conveys a pattern of disruptive behavior leading to disciplinary action like suspensions and expulsions. Such disciplinary action can create a vulnerability to decreased academic achievement, social isolation, and school drop-out. These vulnerabilities are known risk factors for delinquent behaviors, and these vulnerabilities are exacerbated for Black and Latinx adolescents who experience disproportionate policing and surveillance in school (Goldstein et al., 2019). A vast majority of adolescents involved with the JLS report having adverse childhood experiences, which put them at risk for continued involvement in the system (Wasserman & McReynolds, 2011; Baetz et al., 2021; Folk et al., 2021).

PTSD is a significant facet of the mental health profile for adolescents with JLS involvement. In general, JLS-involved adolescents commonly experience high rates of other diagnosable mental health conditions as well. In a systematic review of 47 studies conducted in 19 countries, comprising 32,787 adolescents (28,033 male [85%]), prevalence rates of PTSD, conduct disorder, major depression, ADHD, and psychotic disorders suggested these disorders are substantially more common among detained adolescents compared with the general US adolescent population (Beaudry et al., 2021). Detained adolescents had significantly higher rates of PTSD than the general population: 18% of female and 9% of male detained adolescents, compared to 2% in males and 8% in females in the general population (Merikangas et al., 2010). The prevalence of conduct disorders in detained adolescents was 61.7% for males and 59.0% for females, whereas the estimated lifetime prevalence of conduct disorders in the general U.S. adolescent population is approximately 10% (Nock et al., 2006). The prevalence of psychotic disorder in detained adolescents was between 2.7 and 2.9%, compared to a prevalence of approximately 1% of age-equivalent adolescents in the general population (Kirkbride et al., 2006; Costello et al., 2005). Additionally, rates were disproportionately higher among detained adolescents than in the general population for significant depression (10.1% versus 7.5%; Avenevoli et al., 2015;) and ADHD (17.3% versus 11%; Visser et al., 2014). The vast differences in mental health disorder prevalence highlighted by this review underscore the severity of psychiatric morbidity in detained adolescents. This review also suggests that trauma and mental health disorders may be insufficiently assessed and diagnosed in this population, which could have significant ramifications for prevention and intervention efforts in correctional settings.

The studies above draw comparisons of JLS-involved adolescents to community populations from separate studies. While the existing literature provides essential evidence of the disproportionate rates of mental health disorder diagnoses among JLS-involved adolescents compared to the general adolescent population and offers a valuable framework for understanding the significance of this issue, many of these studies used different assessment methods and samples. The studies included in Beaudry and colleagues’ systematic review, for example, had samples of varying age ranges (e.g., studies could meet inclusion criteria provided their sample fell anywhere between ages 10 and 19), had samples from detention and carceral environments that differed widely in size and structure, and employed meaningfully different assessment methods (e.g., assessments conducted by non-professionals using standardized measures versus evaluations conducted by clinical psychologists or psychiatrists). In addition, only 21 of the 47 studies included in the systematic review assessed for PTSD. Given the established high rates of exposure to community violence and potentially traumatic events among this population, we saw this as an area necessitating further exploration as far as rates of the disorder in detained youth. The variance in methodology across studies represents a gap in the literature on this vital topic, and there is an urgency to better understand the needs of JLS-involved adolescents as their service-need profile has increased in complexity. Kim and colleagues (2021) highlighted that youth who have formal contact with the system tend to be considered adolescents with complex rehabilitative needs, with the likelihood of experiencing trauma, mental health problems, and committing a re-offense. Further understanding of the mental health treatment needs of JLS-involved adolescents, especially those who are detained and are most likely in this “complex needs” subgroup, is critical to improve rehabilitation goals and decrease the potential harm of unresponsive care to these adolescents.

Previous research has been limited in comparing rates of mental health disorders in detained youth to community samples within a single study, and single-study investigations are needed to give the variance in a methodology inherent to comparing across studies outlined above. The most recent study to directly compare mental health diagnoses among samples of JLS-involved youth was the study by Cauffman and colleagues (2007). This study did examine associations between factors like substance dependence and the likelihood of confinement to a detention center or waiver to adult court; however, it did not obtain diagnostic information on PTSD or other psychiatric disorders that are highly prevalent among this population, a weakness that the current study aims to address. This study also did not compare JLS-involved adolescents to a community sample of adolescents, which is necessary for providing further evidence of higher rates of PTSD among the detention population.

Wood and colleagues (2002) directly compared PTSD symptomatology among a sample of incarcerated adolescents to a sample of high school students from similar inner-city communities. This study had many strengths, and its implications were meaningful, as it found significantly higher rates of PTSD among incarcerated adolescents compared to the non-incarcerated sample. This finding was associated with higher levels of sexual and community violence exposure among the incarcerated youth compared to their high school counterparts. One weakness in this study that we aim to improve upon was the use of a self-report measure (the Los Angeles Symptom Checklist) to assess for PTSD among both samples. While this is an internally consistent and reliable measure for assessing PTSD (Foy et al., 1997), self-report instruments have some limitations, mainly that they can introduce variance that otherwise could be reduced through using clinician-administered assessments. Additionally, we are interested in advancing the literature by comparing a sample of detained adolescents to a sample of adolescents evaluated at a community mental health outpatient clinic rather than a general sample of adolescents, as we expect to see higher rates of PTSD among JLS-involved adolescents even compared to adolescents who are presenting for clinical assessment and treatment, not just those in a presumably healthy sample.

Current investigation

This study involves a secondary data analysis of billing claims from an outpatient community clinic with a satellite location embedded in the local juvenile detention center. Consistent with existing literature, we hypothesize that rates of PTSD will be higher among adolescents in detention than adolescents evaluated in the community clinic. We present this against the background of other salient diagnostic patterns found in the sample and address a gap in the literature by eliminating common confounding factors (e.g., clinician training and instruments used in evaluation). Clinicians from the same organization conducted evaluations at both sites using the same procedures and instruments. Thus, the current investigation provides an update to the existing literature.

Methods

Participants

Participants (N = 981) were adolescents ages 12-20 (M_age = 15.93 years, SD_age = 1.924, 60.6% male) seen for psychiatric evaluation in a community mental health agency or a juvenile detention facility. Out of these cases, 635 (64.7%) were evaluated in the detention setting (M_age = 15.97, SD_age = 1.65, 70.1% Male), and 346 were evaluated at the community mental health agency (M_age = 15.83, SD_age = 2.35, 56.6% Female). Sex was more evenly split among the community clinic sample, with only a slight majority of female adolescents (n = 196, 56.6%) compared to male adolescents (n = 150, 43.4%). The detention setting sample had a majority of male adolescents (n = 445, 70.1%) compared to female adolescents (n = 189, 28.9%).

Data collection

Claims data was collected for adolescents seen for psychiatric evaluation by a community mental health agency with two programs: outpatient services and consultation services within a secure detention facility. Both settings were in an urban location in the northeast region of the United States. The community mental health agency serves children ages three to 20 with Medicaid insurance. The detention facility holds adolescents ages 12-20 years who had felony-related charges, were deemed as a flight risk, or posed serious community safety concerns.

During the study period (2017-2021), five board-certified child and adolescent psychiatrists conducted diagnostic evaluations at a detention center and outpatient clinic. Two of these psychiatrists conducted evaluations at both the detention and outpatient clinics. One conducted evaluations only at the detention center, and this psychiatrist had forensic fellowship training. Two psychiatrists conducted evaluations at the outpatient clinic only. Six masters-level clinicians provided intakes and engaged in collaborative documentation with the psychiatrist during the period. All of these clinicians were licensed and had masters-level training. The use of structured assessments described below was a part of this training.

Adolescents seen in the community clinic presented as routine patients seeking outpatient services. Adolescents in detention were referred for psychiatric evaluation if they had already been prescribed medication in the community, were court-ordered for a medication evaluation, or had a change in mental status that warranted evaluation as identified by mental health staff. Across both settings, psychiatric evaluations involved a clinical interview that incorporated standardized instruments such as the Patient History Questionnaire 9 (PHQ-9; Kroenke & Spitzer, 2002), Child PTSD Symptom Scale 5 (CPSS-5; Foa et al., 2018), and Columbia Suicide Severity Rating Scale (C-SSRS; Posner et al., 2011).

Claims data was extracted from the electronic medical record (EPIC platform) from 2017 to 2021 for both adolescents in detention and the community. Claims data for all intake comprehensive biopsychosocial evaluations (CBE) completed at the outpatient clinic and the detention program for adolescents ages 12- 20 years were included. Variables in the billing report include patient age at the time of evaluation, primary diagnosis at the visit, and demographic information such as sex assigned at birth. Billing data only captured the primary diagnosis. Billing data did not capture whether adolescents seeking community care were JLS involved. The only exclusion criteria included patients ages 11 years and younger and 21 years and older. Any repeat visits within a year of each other were excluded to limit carryover diagnosis. Any adolescents who presented at the detention center and community clinic were excluded, as some evaluations could relate to continuity of care.

Procedure

This study was determined not to require IRB review because the dataset did not include identifiable information, and there was no chance of re-identifying subjects. The study was, therefore, not human subjects research.

Data analysis

Data was analyzed using SPSS version 28 (IBM, 2021). Descriptive statistics were generated to understand the top five primary diagnoses apart from PTSD in the sample. From these results, chi-square tests were run to understand significant relationships between detained and non-detained youth. For diagnoses that had a significant relationship, hierarchical logistic regressions were run. These models were used to understand likelihood ratios for being diagnosed with one of the top six diagnoses in the sample, based on whether an adolescent was seen in the community versus the secure detention setting (Harris, 2021). Independent variables included in the logistic regression tests were whether the youth was evaluated in detention, the reported natal sex at the time of evaluation, and age in years. Binary dependent variables were the diagnosis of PTSD, Major Depressive Disorder (MDD), ADHD, Adjustment Disorder, Unspecified Trauma and Stressor Related Disorder (UTD), and Generalized Anxiety Disorder (GAD), which were the six most frequent diagnoses found in the entire sample. Hierarchical regression models were guided by a theory of cascading risk for a PTSD diagnosis, that detention would be most impactful (e.g., detention status; McNair et al., 2019; Ford et al., 2010), followed by natal sex (Hiscox et al., 2023), then age (Kongshoj et al., 2023).

Results

Descriptive Statistics

Among detained adolescents (n=635), 141 (22.2%) had a primary diagnosis of PTSD, compared to 36 (10.4%) in the community clinic. Aside from PTSD, the five most frequently occurring mental health disorder diagnoses among both the detention and community clinic sample were MDD, ADHD, Adjustment disorder, UT, and GAD. The prevalence of major depressive disorder was similar among both samples (22.2% vs 23.4%). Prevalence of ADHD (23.4% vs. 20.9%), adjustment disorder (13.2% vs. 9.6%), unspecified trauma disorder (10.4% vs. 5.8%), and generalized anxiety disorder (10.4% vs. 2.2%) were significantly higher among adolescents in the community clinic compared with detained adolescents. Table 1 shows primary diagnosis rates across the two samples.

Table 1.

Primary Diagnosis Rates

Diagnosis	Community clinic		Detention		Total Sample
	N	%	N	%	N	%
PTSD *	36	10.40%	141	22.20%	177	18%
Depressive Disorder	81	23.40%	141	22.20%	222	22.60%
ADHD	81	23.40%	133	20.90%	214	21.80%
Adjustment disorder	46	13.20%	61	9.60%	107	10.90%
Unspecified trauma/stressor-related disorder*	36	10.40%	37	5.80%	73	7.40%
Anxiety disorder *	36	10.40%	14	2.20%	50	5.10%
Total	346	35.30%	635	64.70%	981	100%

Note. PTSD=Posttraumatic stress disorder; ADHD=attention deficit hyperactivity disorder;

^*indicates p<0.05

Chi-square analyses revealed significant associations between being evaluated in detention and receiving a primary diagnosis of PTSD (χ2=22.04, df=1, p<0.001); UTD and being evaluated in detention (χ2=29.32, df=1, p<0.001), and GAD (χ2=31.13, df=1, p<0.001). No significant associations were found for ADHD, Adjustment Disorder, and MDD. Therefore, hierarchical logistic regression would be appropriate for PTSD, GAD, and UTD.

Logistic Regression

A hierarchical logistic regression was performed to determine whether age, sex, and being evaluated in the juvenile detention center were related to the probability of receiving a primary PTSD diagnosis. Table 2 shows the results of the logistic regression. The overall model was statistically significant (n=981; χ2 _(df=2)= 36.58; p<0.001), correctly classifying 100% of the cases. Being evaluated in detention (B=0.92, p<.001) and age (B=0.16, p<0.001) significantly predicted the likelihood of a primary PTSD diagnosis. The odds of an adolescent in detention receiving a primary PTSD diagnosis were 2.5 times higher for adolescents evaluated in detention compared to those in the community (p<0.001, OR= 2.53). The results of the Cox and Snell, and Nagelkerke R2 indicate that being evaluated in detention accounted for 3.6-5.9% of the variance in PTSD diagnoses. Each additional unit increase in age (i.e., year) was associated with an increase of 1.176 in the odds of an adolescent receiving a primary PTSD diagnosis. The likelihood of primary PTSD diagnosis did not differ by sex (p=0.921).

Table 2.

Results of the Logistic Regression: Likelihood of Primary PTSD Diagnosis

Factor	B	Wald	df	p	OR	CI
Detention	0.92	19.69	1	<0.001	2.53	1.78 - 3.80
Sex	−0.02	0.01	1	0.921	0.98	0.69 - 1.40
Age at Visit	0.16	12.02	1	<0.001	1.18	1.07 - 1.29
Constant	−4.77	35.13	1	<0.001	0.01

Note. Overall Model: χ2 _(df=3)=36.58; p<0.001; Goodness-of-fit: −2LL= 896.12, χ2=11.57_(df=11), p=0.171.

A second hierarchical logistic regression was performed to determine whether age, sex, and being evaluated in the juvenile detention center were related to the probability of receiving a primary GAD diagnosis. Table 3 shows the results of this logistic regression. The overall model was statistically significant (n=981; χ2 _(df=2)=38.33; p<0.001), correctly classifying 100% of the cases and being evaluated in detention (B=−1.50, p<.001) significantly predicted the likelihood of a primary GAD diagnosis, and being evaluated in detention led to a 78.7% decrease in the odds of a GAD diagnosis (OR=0.22). The results of the Cox and Snell, and Nagelkerke R2 indicate that being evaluated in detention accounted for 3.8-11.6% of the variance in anxiety diagnoses. Each additional unit increase in age (i.e., year) was associated with an increase of 1.190 in the odds of an adolescent receiving a primary GAD diagnosis (OR=1.19, p=0.015). The likelihood of primary GAD diagnosis did not differ by sex (p=0.137).

Table 3.

Logistic Regression Results: Likelihood of Primary GAD Diagnosis

Factor	B	Wald	df	p	OR	CI
Detention	−1.50	19.93	1	<0.001	0.22	0.11 - 0.43
Sex	0.47	2.21	1	0.137	1.59	0.86 - 2.94
Age at Visit	0.17	5.86	1	0.015	1.19	1.03 - 1.36
Constant	−5.26	18.74	1	<0.001	0.01

Note. Overall Model: χ2 _(df=3)=38.33; p<0.001; Goodness-of-fit: −2LL= 356.73, χ2=11.48_(df=8), p=0.176

A third hierarchical logistic regression was performed to determine whether age and being evaluated in the juvenile detention center were related to the probability of receiving a primary UTD diagnosis. Table 4 shows the results of the logistic regression. The overall model was statistically significant (n=981; χ2 _(df=3)=31.98; p<0.001), correctly classifying 100% of the cases. Being evaluated in detention (B=−1.45, p<.001) at evaluation significantly predicted the likelihood of a primary UTD diagnosis, and being evaluated in detention led to a 75.5% decrease in the odds of a primary UTD diagnosis (OR=0.24). The Cox and Snell and Nagelkerke R2 results indicate that being evaluated in detention accounted for 3.2-9.6% of the variance in a UTD diagnosis. Age and sex were not significant predictors of receiving a primary UTD diagnosis.

Table 4.

Logistic Regression Results: Likelihood of Primary Unspecified Trauma/ Stressor Related Disorder Diagnosis

Factor	B	Wald	df	p	OR	CI
Detention	−1.45	19.65	1	<0.001	0.24	0.12 - 0.44
Sex	0.44	2.08	1	0.149	1.55	0.85 - 2.84
Age at Visit	0.09	1.81	1	0.178	1.09	0.95 - 1.25
Constant	−3.92	1.15	1	<0.001	0.02

Note. Overall Model: χ2(df=3) 31.98; p<0.001; Goodness-of-fit: −2LL= 368.79, χ2=9.71_(df=8), p=0.286.

Discussion

In this sample, adolescents evaluated in detention had dramatically higher odds of having a primary diagnosis of PTSD. This finding supports previous literature that JLS-involved adolescents have higher rates of PTSD than their non-JLS-involved peers (Baetz et al., 2021; Saigh et al., 1999; Wasserman & McReynolds, 2011; Wasserman et al., 2004; Wolpaw & Ford, 2004). Anxiety disorders were more common in the community clinic than in detention, and the trend was the opposite for PTSD, whereby the prevalence rate of PTSD was double that of the outpatient clinic. ADHD rates were slightly higher in the community clinic than in detention; this is in line with literature that reported higher rates of ADHD in community samples (Visser et al., 2014). It is important to note that the diagnosis in the billing data reflects the primary diagnosis that had the most impact on youth functioning. The data did not include comorbidity, so the discussion reflects the likelihood of having a particular primary diagnosis and not the prevalence of the disorder itself.

Rates of UTD were higher in the community compared to detention, and this could be related to the types of trauma exposure generally experienced across groups. JLS-involved adolescents are known to have higher rates of violent victimization and polyvictimization than adolescents without JLS involvement (Ford & Hawk, 2012). Those in the community may be seeking care after a trauma that would not satisfy criteria A for PTSD (e.g., seeking treatment after a death in the family). This potential phenomenon echoes the Criteria A debate about whether Criteria A should be broad or restricted (Kilpatrick et al., 2009) or done away with altogether (Brewin et al., 2009). Lastly, the rates of adjustment disorders were lower in detention than in the clinic. It is also possible that the setting in which youth were evaluated could influence diagnostic prioritization. With increased awareness of high rates of trauma exposure among detained youth and toward providing trauma-responsive care in detention settings, providers may have been more likely to diagnose PTSD as a primary concern among youth in detention. Provider bias matters in interpreting the results, as provider bias could impact diagnostic severity. There could be a tendency for diagnoses or levels of pathology to be worse in the detention setting, considering that detention is considered a high-acuity setting that also serves a public safety function. As noted earlier, youth in detention are presenting with more complex mental health needs as more youth with lower-level offenses are diverted away from detention (Kim et al., 2022).

While we cannot determine the impact of setting on diagnosis, these findings align with what is known about youth who are detained. Detention settings can be traumatizing as adolescents can be exposed to violence there, whereas the natural home setting could be a buffer against the development of PTSD after trauma exposure (e.g., social supports, coping resources; Betancourt et al., 2013; Brooks et al., 2016). As stated earlier, this paper contributes to the existing literature by eliminating common confounding factors in comparing diagnostic prevalence rates between youth in detention and the community due to evaluations at both sites conducted by clinicians from the same organization, using the same procedures and instruments. This study confirms the higher likelihood of receiving a primary PTSD diagnosis and further warrants it as a focal diagnostic concern to address with JLS-involved adolescents. Additionally, our results introduce two diagnostic trends worth further exploration – that youth in detention were less likely to receive a primary GAD or UTD that was not PTSD.

Implications

The results of the current study have implications for clinical practice, forensic practice, and future research. From the clinical perspective, adolescent detention facilities are not traditionally considered nor equipped to be treatment centers. Given the acuity of clinical mental health concerns that adolescents housed there often present with, the need for psychosocial support for adolescents and families interacting with the legal system, and the fact that adolescents in detention have greater physical accessibility to allow for engagement with healthcare, this setting may offer some favorable circumstances to reach adolescents and families for trauma-focused assessment and treatment, which are the standard of care for the treatment of PTSD (Salloum et al., 2022; Dorsey et al., 2017). Scientific evidence is growing to demonstrate the efficacy of evidence-based treatments related to post-traumatic problems for adolescents, such as Dialectical Behavior Therapy, Structured Psychotherapy for Adolescents Responding to Chronic Stress (SPARCS), Trauma Affect Regulation: Guide for Education and Therapy (TARGET, Ford, et al., 2012), Cognitive Processing Therapy (CPT; Resick & Schnicke, 1993), Trauma Grief Components Therapy for Adolescents (TGCTA, Saltzman et al., 2001), Trauma-Focused Cognitive Behavioral Therapy (TF-CBT, Cohen et al., 2016) and Seeking Safety (Lenz et al., 2016). DBT, TARGET, and TGCTA have been studied in the detention setting specifically and have shown promise (Ford et al., 2014; Ford et al., 2016; Olaghere et al., 2021; Kumm et al., 2019). Knowledge of implementation strategies and implementation outcomes of these interventions in detention is not well-established, yet urgently necessary. It could be helpful for adolescents to start trauma-focused treatment in detention and have continuity to continue the work after discharge from detention (Snyder, 2018).

Within facilities, a multi-tiered system of support can meet adolescents and families where they are at the time of treatment engagement and best allocate staff and human resources. This method requires highly skilled staff to determine the level of intervention indicated while balancing the adolescent’s and family’s readiness for treatment. Trauma-informed care is necessary as a universal precaution and standard of care. This standard of care can then support the implementation of trauma-focused treatments. A baseline standard of trauma-informed care is particularly needed in the JLS setting, where both adolescents and families tend to have strong hesitations about treatment engagement, which is often well-founded due to negative previous experiences with other systems of care. Additionally, both adolescents and staff often experience trauma in adolescent detention facilities, suggesting a need to implement trauma-informed practices that not only focus on the adolescents but also on the staff (Baetz et al., 2021).

From a clinical standpoint, mental health care is most effective when the patient/family-provider therapeutic alliance is strong and safe (Fluckiger et al., 2018), which can be challenging to accomplish in a system designed to monitor and control adolescents and families. Highly skilled clinicians may effectively navigate the complexities of this system in a way that fosters a strong therapeutic alliance. However, this ability tends to come with experience, adequate training, and supervision, which often need to be improved in this system. An additional concern when implementing evidence-based models of care in juvenile detention facilities is the cost of this level of care. Previous literature has identified the cost-effectiveness of providing treatment for JLS-impacted adolescents and families in the community (Sheidow et al., 2012; Edmunds et al., 2018). Literature on the implementation of evidence-based models of care notes cost as a significant limitation to these efforts in both detention and the community.

Implementing evidence-based interventions is challenging in the juvenile legal system, including juvenile detention settings, where there is a conflict of missions between mental health and legal systems. Factors impacting implementation include high cost of training, high staff turnover, and difficulty recruiting and retaining mental health providers (Powell et al., 2020). The mental health service capacity for adolescents in JLS settings is limited, with one study reporting that only 15% of adolescents who need treatment in JLS settings receive on-site mental health services (Teplin et al., 2005). This trend compounds for Black and Latino adolescents who have a decreased likelihood of (33%) accessing services generally (White, 2019). Apart from this, even when evidence-based services are implemented, these are often limited by external factors specific to juvenile detention, such as short-term stays, uncertainty about when adolescents may be released, and inability to easily implement inter-professional communication and collaboration (e.g., probation officers, school personnel, outpatient mental health providers, primary care providers). As such, the care received in detention is siloed from what adolescents and families need to promote health and healing and to reduce the likelihood of future legal system contact.

These findings also have implications for forensic evaluators and court service case planners (e.g., probation officers). Court recommendations for rehabilitation, guided by the Risk Needs Responsivity (RNR; Bonta & Andrews, 2007) principle, consider traumatic stress as a vital part of the presentation and trauma-informed practices as an essential system response. The RNR model guides justice system staff to determine an alleged offender’s risk level (e.g., low, moderate, high) to match the appropriate level of service, to consider the factors related to the specific rehabilitation needs, and to understand characteristics of the individual that can increase the likelihood of intervention success, like strengths or learning style (Bonta & Andrews, 2007). Case planning should not neglect behavioral health as a dynamic responsivity factor (Brogan et al., 2015), and rehabilitation recommendations should also consider whether residential settings can address trauma-related concerns that may be a central part of the adolescent’s presentation. A trauma-informed environment, evidence-based interventions, and family engagement are essential when an adolescent is placed outside of their home.

Lastly, there are implications for future research. Research should consider time and ecological contexts more fully. For instance, research should follow a community sample at higher risk for legal involvement to measure clinical diagnoses, how diagnosis may change if a youth is detained, and how symptoms of PTSD may be missed or misdiagnosed in community samples. This data collection could show the impact of incarceration on PTSD symptoms or other mental health concerns, as it is expected that incarceration worsens overall mental health. Those with PTSD may have frequent reminders of their trauma given their court hearings, engagement in therapy, hearing peer stories, seeing restraints/other violence in the facility, and separation from family. For clinical research, our study calls for ongoing attention to the underdiagnosis and treatment of trauma in community settings. As more literature informs our practice, clinicians must recognize what we may be over- or under-diagnosing in the community. For example, a study examined psychiatric diagnoses and treatment before and after placement in correctional facilities, finding that most diagnoses were higher in correctional settings than in the community. However, most youths had received outpatient, inpatient, or emergency services (Whitney et al., 2022).

Limitations

The current study adds to the existing literature by examining primary diagnosis rates across the community and JLS samples in a single city, assessed by clinicians employed by a single clinic and trained in the same methods and assessments. Several noteworthy limitations should be considered when interpreting these findings. Because of the inconsistency of billing data entry, race and ethnicity information could not be included in the analysis. These variables could have impacted the results, knowing that Black and Brown adolescents have increased trauma exposure and decreased service utilization (Metzger et al., 2023), which could be protective against the onset and maintenance of PTSD and other mental health disorders. Because Black and Brown adolescents are overrepresented in juvenile detention, the increased rates of PTSD could be attributed to race/ethnicity factors rather than the settings themselves. Furthermore, the billing data did not capture sexual and gender minority status, which is relevant to rates of PTSD, and the samples were not matched by sex. Billing data also does not capture adolescents with significant trauma exposure who do not meet the full criteria for PTSD. The billing data only captured the primary diagnosis at the evaluation, limiting the ability to understand comorbidities or co-occurring disorders, trauma exposure, differences in prevalence rates of diagnoses in and across groups (e.g., whether youth in detention were more likely to be diagnosed with externalizing disorders), and systemic barriers that can reduce treatment utilization. Lastly, this was a convenience sample drawn from a single geographic area.

Conclusion

Our study is significant in demonstrating that rates of a primary PTSD diagnosis are significantly higher among adolescents in detention compared to a community sample. Future research is needed to explore how clinician training and expertise shape diagnosis rates in detention facilities, as this may impact adolescents’ ability to access indicated and evidence-based treatments. Our sample came from a community clinic associated with an academic center highly committed to enhancing trauma-informed care, practices, and treatments; this may not be the norm across the country. It is crucial to implement evidence-based trauma treatment in detention settings, and research should continue to examine the feasibility, acceptability, and effectiveness of these interventions.