Trauma Focused Cognitive Behavioral Therapy for Adolescents with Problem Sexual Behaviors: Feasibility and Acceptability for Alleviating Traumatic Stress, Regulation Problems, and Sexual Behavior Problems

Abstract

While there are multiple risk factors associated with adolescent sexual violence perpetration, childhood trauma and adversity may be one of the most prominent factors. Trauma sequela including regulation deficits, trauma symptoms, and attachment-based disruptions may be key modifiable targets in interventions that aim to impact functional outcomes, yet treatment targets and philosophies of change are not always well defined or not robustly integrated into problem sexual behavior treatment programs for adolescents. Trauma-Focused Cognitive Behavioral Therapy for problem sexual behaviors among adolescents (TF-CBT for PSB-A) has emerged as a new treatment option for adolescents with histories of victimization or adversity who may concurrently be at risk for or have detected PSB. This study explored the initial feasibility, acceptability, and preliminary promise in providing the enhanced TF-CBT for PSB-A in community-based and juvenile justice settings to youth and their caregivers. Results show feasibility and acceptability of this intervention but found notable readiness factors that warrant consideration. Results also demonstrated statistical significance and positive trends from pre-to post-test on treatment targets. This study offers promise for trauma-specific treatment services among youth and opportunities for strengthening training and implementation.

Introduction

Sexual violence (SV) is a significant social problem that leads to detrimental health outcomes, social instability, and economic disparities (Basile et al., 2022; Loya, 2014; Miller et al., 1996; Strathearn et al., 2020; Waechter & Ma, 2015). The study of SV perpetration can reveal opportunities and effective strategies to disrupt patterns of abuse. Specifically, SV perpetration is most likely to peak during adolescence (Finkelhor et al., 2009). In fact, adolescents with problem sexual behaviors (PSB-A)– irrespective of their formal status with the juvenile justice system– are a specialized group who contribute to the overall rate of SV, particularly against children and youth under the age of 18. A recent study found that while both adults and adolescents perpetrate SV against children, adults are more likely to abuse children under 13, and adolescents are more likely to abuse other young people ages 14 to 17 (Gewirtz-Meydan and Finkelhor, 2020). Overwhelmingly, adolescents with PSB discontinue sexual offending in adulthood; a meta-analysis of decade over decade trends in recidivism found that while general recidivism was approximately 44%, the rates of sexual recidivism were reliably low across studies at 8% (Lussier et al., 2023). Thus, it is essential to evaluate the utility of therapeutic programs that are customized to the specialized risks and unmet needs of adolescents with problematic sexual behaviors and their families.

Defining and identifying common characteristics of problematic sexual behavior (PSB) in adolescence is complex, with variations in types of offending behaviors and ways youth exhibit different forms of sexually problematic behaviors (Hunt et al., 2024; Malvaso et al., 2020). Authors have distinguished between what constitutes sexually harmful and what is considered problem sexual where some have argued that problem sexual is considered developmentally inappropriate behaviors that can include actions such as excessive self-stimulation and may exclude behaviors that constitute sexual abuse of others; whereas sexually abusive behavior or harmful sexually behavior may encompass non-consensual acts and may include the use of force or coercion (Malvaso et al., 2020). Other papers (see: Gray et al., 1997, pg. 271) included in scoping reviews have argued that PSB may still include “quite serious and harmful sexual acts” (Hunt et al., 2024 pg. 2572). Throughout our work, we will use the term"problematic sexual behaviors"(PSB) to designate adolescents who exhibit a range of these behaviors, from sexual exposure to sexual offending. We will advocate for the consistent term of"PSB"to be used as it is more comprehensive, non-pejorative, and suitable to encompass the range of problem and/or abusive behaviors that warrant therapeutic intervention.

Trauma as a Critical Risk for Onset and Course of PSB-A

While there are multiple risk factors associated with SV perpetration, research demonstrates that childhood adversity may be one of the most relevant factors contributing to SV perpetration in young people (Burton, 2003; Burton et al., 2011; Creeden, 2013; Leroux et al., 2020; Levenson et al., 2017; Van Wijk et al., 2006). Developmental adversities can contribute to multiple poor health outcomes and have specifically been noted as indicated risks for SV perpetration (Baglivio & Wolff, 2021; Creeden, 2013). There are many forms of childhood adversity that co-occur and may include direct forms of physical, psychological, emotional, or sexual abuse or indirect forms like economic strain, family stress, witnessing violence, or community violence exposure (Alexander et al., 2020; Baglivio & Wolff, 2021; Levenson et al., 2017; Seto & Lalumiere, 2010). Comparative and longitudinal research consistently reveals links between childhood adversity and SV perpetration among youth. For instance, relative to general youth and delinquent youth samples, PSB-A samples have elevated rates of childhood emotional, physical, sexual, intrafamilial, and cumulative (co-occurring) adversity (Levenson et al., 2017; Seto & Lalumiere, 2010; Yoder et al., 2017a, 2020). Furthermore, in a prospective longitudinal study, childhood physical abuse and neglect were identified as early risks for SV perpetration with sexual abuse exposures approaching significance (Widom & Massey, 2015).

Intermediary Factors

The relationship between childhood adversity and youth SV perpetration is complex with no definitive cause-and-effect; while they are correlated, children exposed to adversity do not always commit acts of SV during adolescence (Widom & Massey, 2015). It is critical therefore to understand the intervening variables explaining this relationship. A nuanced look at the way adverse experiences are relatively internalized and coded can better contextualize this connection. For instance, researchers have found prefrontal neurological zones responsible for planning, organizing, understanding consequences of decisions, and regulation are implicated as a result of adversity (Steinberg, 2008; Zelazo et al., 2013) that may partially explain SV perpetration. Behavioral, emotional, and cognitive regulation are key mediators in the relationship between trauma and perpetration of SV; trauma contributes to deficits in regulation, which in turn can lead to SV perpetration (Brown et al., 2021; Yoder, et al., 2019). Interestingly, young people who commit SV often experience more internalization problems like loneliness, anxiety, and low self-esteem compared to those who engage in delinquent behavior (Seto & Lalumière, 2010; Van Wijk et al., 2006, 2007), subsequently contributing to feelings of helplessness and hopelessness that precipitate SV (Brown & Grady, 2019). In fact, studies show that the connection between childhood abuse and internal mental health struggles are stronger among young people who perpetrate SV relative to non-sexual crimes (Boonmann et al., 2016).

The negative internalization of traumatic events during critical developmental stages may manifest as trauma symptoms or even a formal Post-Traumatic Stress diagnosis (Perry, 2009; Turner et al., 2012). Research has found inverse relationships between trauma symptoms and executive functioning capacities (Nooner & Leaberry, 2013; Polak et al., 2012). More notably, trauma symptoms have been indicated as a mediator whereby adversity exposures, specifically sexual abuse has led to a greater likelihood for trauma symptoms that can create deficits in executive functioning among youth that perpetrate SV (Brown et al., 2021). Trauma symptoms may be an area of specific concern given they can be unambiguous outward markers of internalized distress (Nooner & Leaberry, 2013) that manifest externally.

Researchers have also found attachment-based disruptions resulting from trauma that result in compensatory behaviors like SV. For example, researchers found a connection between physical abuse and the development of insecure attachment styles (Grady et al., 2018). There may also be a bi-directional link between relational experiences in the home and trauma whereby they are mutually reinforced and co-occur to impact SV perpetration; research has found that harsh and inconsistent parenting can lead to both traumatic experiences and insecure attachment (Yoder et al., 2018) suggesting that adversity exposure and caregiver relational experiences have a reciprocal relationship that contributes to attachment-based disruptions. Further, young people who commit sexual harm often experienced harsher parenting and greater traumatic events than those who engage in delinquent behavior (Yoder et al., 2018). Yet, protective mechanisms like family and caregiver support have been shown to mitigate the impacts of trauma on SV behaviors (Yoder et al., 2020) suggesting there may be therapeutic benefits to building relational connections and experiences in the context of family or support systems (Yoder et al., 2017b).

Trauma Targets Not Consistently Integrated in PSB-A Treatments

As such, trauma sequela including regulation deficits, trauma symptoms, and attachment-based disruptions may be key modifiable targets in interventions that aim to impact functional outcomes (Cohen et al., 2017). However, our treatment targets and philosophies of change are not always well defined or not robustly integrated consistently into PSB-A treatment programs (Drisko et al., 2019; Wampold, 2018). A common challenge in many treatment programs, particularly those addressing complex issues like youth SV, is the lack of clear, consistent, and integrated treatment targets and philosophies (Wampold, 2018). This is especially true given that adolescent treatments were largely borrowed from philosophies widely applied to adults who commit SV (Dwyer & Letourneau, 2011; Letourneau & Borduin, 2008). This has created treatment programs that are grounded in conflicting theoretical approaches (like the adult-based application of cognitive behavior and relapse prevention or the lack of family integration) with confusing directions and unclear targets for tailored adolescent treatment (Letourneau & Miner, 2005). Such disjointed philosophies lead to fragmented care, ineffective interventions, and a failure to address the root causes of problematic behaviors (Crits-Christoph & Gibbons, 2021). Consequently, some treatments may focus primarily on the sexually harmful behavior, without addressing the underlying causes of such behavior like trauma symptoms, regulation, and relational connections.

For instance, a recent meta-analytic review of specialized treatment programs for PSB-A modeled after adults found great variability across and within these approaches (Kettrey & Lipsey, 2018). Some programs fostered empathy and accountability in young people who have committed sexual offenses, while others focused on topics such as personal goal setting, relapse prevention, and developing healthy relationships. While the programs did not overtly identify a cognitive behavioral therapy (CBT) approach, many employed CBT components (i.e., thought change, automatic thoughts, behavioral plans). Importantly, most programs included some form of counseling, but only a few included family therapies (Kettrey & Lipsey, 2018) despite studies suggesting therapeutic benefits of family-focused approaches (St. Amand et al., 2008). Cognitive behavioral therapies targeting individual-level thought-based behavior change (Dopp et al., 2015; Dopp et al., 2015; McGrath et al., 2010) may not always be supplemented with caregiver and family-inclusive approaches. Findings from a 2008 meta-analysis – with study samples of adolescents up to age 16 revealed that parenting/behavior management components yielded the largest effects on PSB outcomes and null effects when adult model elements were used (i.e., relapse prevention, assault cycle, or arousal reconditioning) (St. Amand et al., 2008). Thus, refining targets that aptly address evidence-driven risks for SV perpetration (i.e., trauma, regulation, and relational connections) for PSB-A that include a parenting/caregiver component is critical in reducing recidivism among PSB-A. Clear and well-defined treatment targets addressing the underlying risk for onset and course of SV trajectories necessitates the use of trauma work; yet, little is known about the about the utility and promise of such approaches.

Trauma Focused Cognitive Behavioral Therapy Enhanced for PSB-A

TF-CBT for PSB-A has emerged as a new treatment option for juveniles with histories of victimization or adversity who may concurrently be at risk for or have detected PSB. Trauma-Focused Cognitive Behavioral Therapy (TF-CBT) is a therapy designed to help children and their caregivers cope with the effects of trauma. It combines cognitive behavioral techniques with trauma-focused approaches to address symptoms of trauma, improve emotional and behavioral regulation, and strengthen relationships. The therapy involves gradual exposure through a trauma narrative and follows an approach that follows a PRACTICE acronym: Psychoeducation and Parenting Skills, Relaxation skills, Affective modulation skills, Cognitive coping skills, Trauma Narrative and cognitive processing of the traumatic event, In vivo Mastery of Trauma Reminders, Conjoint family sessions, and Enhancing safety. A large part of the approach includes a trauma narrative which involves gradually exposing individuals to traumatic memories in a safe environment, helping them develop coping mechanisms and reduce the impact of the trauma. TF-CBT has recently been enhanced to include targets that address SV risk and behaviors (Grady et al., 2023, 2024). This integrated approach does not deviate from the original model and uses the PRACTICE acronym and includes components like development of relaxation skills during trauma reminders, psychoeducation about healthy sexual behavior, cognitive processing around sexual behavior problems, connecting trauma to PSB, and working with caregivers and youth to address sexual behavior problems (See: Grady et al., 2023). The enhanced model can help youth recognize the connection between trauma and PSB and more effectively cope with the effects of trauma; improve their emotional and behavioral regulation in response to trauma to offset sexual behavior problems; and strengthen their relationships with caregivers using caregivers as support during times of dysregulation. For a full description of the enhanced model, readers are encouraged to review (Blind for Peer Review). While TF-CBT has evidenced reductions in sexual risk behaviors for sexually abused children with no indicated sexual behavior problems (Cohen et al., 2017), the evidence underlying the effectiveness of TF-CBT for SV perpetration among adolescents is only beginning to emerge. In the only known study of TF-CBT for youth with problem sexual behaviors, Everhart Newman et al. (2018) deployed TF-CBT in a residentially based setting and used a pre-and post-test single group design yielding reductions in PTSD, depression, academic problems, suicidal tendencies, and other clinical symptoms. However, the study did not use an enhanced version of TF-CBT for PSB and did not measure PSB risk or behavior directly. In determining feasibility, the authors note stronger organizational support is needed for successful implementation and retention of TF-CBT in residential treatment centers.

Current Study

There is a significant need to better understand how interventions like enhanced TF-CBT for PSB-A work under real world conditions and there is lacking evidence on factors that improve or hinder implementation across all treatment studies of youth SV. Therapeutic service receipt routinely occurs in either community-based practice setting or residential/incarceration-based facilities, depending on the severity of the offense or other dependency factors. In fact, according to a 2009 survey, approximately 74% of male adolescent programs in the United States are in community-based settings while 26% are in residential. Furthermore, 55% of adolescents with PSB are treated in community settings and 44% are treated in residential programs (McGrath et al., 2010). There is a significant gap in assessing the promise and utility of trauma-focused approaches for this group across both service settings. To address these gaps, this study explored the feasibility, acceptability, and preliminary promise in providing the enhanced TF-CBT for PSB-A in community-based and juvenile justice settings to youth and their caregivers. The current study research aims included: 1) What is the feasibility and acceptability of training and implementation of TF-CBT for PSB-A in residential and community-based settings? 2) What are some of the preliminary outcomes among youth who have been exposed to TF-CBT for PSB-A?

Methods

Procedure

This pilot study determined the feasibility, acceptability, and preliminary outcomes of TF-CBT for PSB-A in two different settings who serve adolescents with PSB. This study received Institutional Review Board approvals for all research related processes and required caregiver or guardian consents and youth assents for youth and caregiver involvement. Using existing network relationships, we recruited interested partners in a southern state who work in the Department of Juvenile Justice (DJJ) and a community-based setting who receive referrals for PSB-A. Clinicians from both settings were recruited to participate in a 3-day training on TF-CBT for PSB-A and bi-monthly consultations for 6-months during therapy implementation. All clinicians in the DJJ setting (n = 4) and the community-based setting (n = 3) had at least 2 years of experience working with PSB-A in their respective service settings. The clinician sample size shifted during the study; initially while there were 7 clinicians recruited, only those clinicians with active cases of TF-CBT for PSB-A who were involved in the study were able to continue receiving bi-weekly consultations. Thus, the retained practitioners in DJJ (n = 3) and community-based setting (n = 2) had at least one TF-CBT for PSB-A client on their clinical caseload who engaged in the study from start to end.

Training

Prior to receiving the integrated training, clinicians were required to undergo a 2-day web-based asynchronous standard TF-CBT training, which offered initial exposure and grounding in the model (https://tfcbt2.musc.edu/). The integrated TF-CBT for PSB-A training was offered synchronously online for 3-days for 4-h. This study employed two renown TF-CBT trainers; one who was a nationally certified trainer of TF-CBT and an evidence-informed approach to PSB known as CBT-PSB and a second trainer who has trained on TF-CBT with PSB-A internationally. The training materials were developed by both trainers in consultation and support from the research team and model developers and borrowed heavily from both trainers’ experience implementing TF-CBT in PSB-A service settings. The materials were developed for clinicians with PSB-A clients with concepts integrated to support implementation of skills that would align with their existing PSB approach. In essence, the model was not attempting to introduce a “new way” of implementing PSB work with clients; rather, it was an opportunity to introduce TF-CBT work that mapped onto existing approaches. The DJJ and community-based teams used the Pathways Model Workbook (Kahn, 2023) in their PSB-A work.

Bi-Monthly Consultation

The clinicians with active cases were expected to attend bi-monthly consultation calls that required them to share case examples and walk through specific difficulties, challenges, or “wins” in implementing TF-CBT with their PSB clients. The training team offered tangible feedback, opportunities for new skills acquisition, and help overcoming barriers to implementation. The training team also offered web-based resources to support in more enhanced implementation that included.

Dyad Recruitment

The directors of the programs were asked to help recruit incoming youth and caregivers to the study that would require them to a) complete a pre and post-test measure and b) engage in TF-CBT services delivered by the clinicians. Incoming referrals used a screening process, whereby the caregiver-youth dyads were screened by asking about exposure to initial trauma (yes or no) and whether they would be interested in participating in the project. The clinicians then completed a trauma screening tool (CATS- 2; Sachser et al., 2017) following instructions to determine whether youth would formally qualify for services. Overall, this recruitment approach yielded n = 10 dyads with two dropping out at the beginning of the study due to scheduling and transportation conflicts. The final sample included n = 8 complete caregiver-youth pairs with 100% youth retention from pre-to post during the project period but only n = 6 caregivers completing the full survey at post-test.

Sample Demographics

Youth (n = 8) in the sample were on average 15 years old (M = 15.38, SD = 1.59). All youth in the sample identified as male and 25% (n = 2) identified as Black or African American; 12% (n = 1) identified as Mixed Race; 12% (n = 1) identified as White and Native American; and 50% (n = 4) identified as White. Caregivers in the sample were either maternal caregivers (75%; n = 6) or paternal caregivers (25%; n = 2). The majority of the caregivers identified as White (75%; n = 6) and other racial identities included Mixed Race (12.5%; n = 1) and African American (12.5%; n = 1). All youth in the sample who received TF-CBT for PSB-A were all adjudicated, despite the differing service settings. Data was not collected on the type of charges youth received as IRB approvals were not granted to collect that information.

Data Collection

Given the nature of the study being a feasibility and acceptability pilot study, qualitative and quantitative data were collected. Quantitative data were collected from youth-caregiver dyads using a battery of self-report validated instruments and qualitative data were collected from clinicians and trainers using semi-structured interview guides.

Quantitative

Caregivers or guardians signed consents prior to youth completion of the Qualtrics pre-test and post-test surveys. Then caregivers and youth, separately completed surveys prior to the implementation of TF-CBT for PSB-A and again after the completion of treatment. Caregivers were compensated with a $50 Amazon Gift Card at pre-test and $75 at post-test and clinicians were compensated the same for their support facilitating surveys with youth. Treatment lasted approximately 6 months (M = 5.55 months, SD = 0.60). The surveys included a battery of validated instruments described further below. There were not missing data and no attrition for youth who completed the surveys, but there were varied missing data points across the caregiver surveys with six complete pre- and post-surveys.

Qualitative

Clinicians engaged in focused groups two times during the course of the project period. The clinicians were separated by setting (DJJ and community-based) and answered questions pertaining to the usefulness, applicability, flexibility, and practicality of the training; and the feasibility, acceptability, uptake, and practicality of the TF-CBT for PSB-A model. The data were collected via recorded zoom sessions and lasted approximately 1-h each. The clinicians were compensated with $100 Amazon gift cards for their participation in the focus groups. A total of 4 focus groups with clinicians were completed with n = 3 in the community-based setting that included the supervisor and n = 5 in the DJJ settings that included two supervisors. The Qualitative focus groups were also conducted with trainers and consultants at the completion of the training and consultation services. Questions pertaining to readiness, perceived modifications to training materials for greater internalization and uptake, and strengths and limitations of the model were presented. These data were collected via recorded zoom sessions and lasted approximately 1-h.

Quantitative Measures

Trauma Events, Symptoms, and Functioning

Administered to both caregivers and youth, the Child and Adolescent Trauma Screen (CATS- 2; Sachser et al., 2017) assess traumatic events, symptoms associated with those events, and functional impairment. To assess trauma events, youth and caregivers were asked to self-report or report on behalf of the youth, responding (1 = yes; 0 = no) to 15 total adverse or traumatic experiences (i.e., serious or natural disaster, or seeing someone in the family threatened, hit, or hurt badly). To assess trauma symptoms, youth and caregivers were asked to respond (0 = Never to 3 = Almost Always) to 25 possible symptoms experienced in the last two weeks (i.e., bad dreams, trying not think about it, not being able to remember part of what happened). Finally, the youth and caregivers were asked about how trauma interfered with functioning responding (1 = yes; 0 = no) to five questions (i.e., Getting along with others, School or work, or Family relationships). The tool was used in various ways during analyses. First the youth and caregiver pre-test events subscale was used by summing scores to determine how many traumatic events the youth and caregivers report, respectively, to provide a cumulative trauma score. Although the internal consistency was borderline not acceptable at youth pre-test (α = 0.683) and caregiver pre-test (α = 0.641), it was used only for determining frequency of cumulative trauma. Second, the youth self-report symptoms subscale scores were averaged with strong internal consistency at pre-test (α = 0.934) and post-test (α = 0.873). The caregiver reported trauma symptoms were strong at pre-test (α = 0.911), but there was insufficient data to determine post-test consistencies. Finally, the youth self-report functioning subscale scores yielded unacceptable internal consistencies at pre-test (α = 0.504), but acceptable reliabilities at post-test (α = 0.752) and the caregiver report functioning subscale yielded strong consistencies at pre-test (α = 0.812) with insufficient data at post-test. Thus, individual items were used in data analyses.

Family and Caregiver Relations

To assess dyadic features youth and caregiver relations, the Family Assessment Measure III (Skinner et al., 1995) was reported by caregivers and youth. There are 42 questions that assess aspects of task accomplishment, communication and affective expression, role performance, affective involvement, control, and values and norms. Example questions included: “This person really trusts me” and “This person expects too much of me” with item responses ranging from 0 = Never to 4 = Always. The youth self-report Family Assessment Measure scores were averaged with strong internal consistency at pre-test (α = 0.922) and post-test (α = 0.917). The caregiver report scores were averaged with strong internal consistency at pre-test (α = 0.975) and questionably unacceptable at post-test (α = 0.687). The post-test consistency scores are likely low given the sample size changes from pre-to-post with only 6 post-test assessments. They were used in the analyses in a preliminary way and larger sample sizes are needed for further use of this tool.

Emotional and Behavioral Regulation

Emotional regulation was measured using the youths’ self-report of Difficulties in Emotion Regulation (DERS; Gratz and Roemer, 2004; Hallion et al., 2018). The instrument includes 36 questions pertaining to lack of acceptance of emotional responses (pre-α = 0.816 and post-α = 0.687); difficulty engaging in goal direction behavior (pre-α = 0.734 and post-α = 0.931); impulse control deficits (pre-α = 0.917 and post-α = 0.759); lack of emotional awareness (pre-α = 0.698 and post-α = 0.667); limited access to emotional regulation strategies (pre-α = 0.875 and post-α = 0.778); and lack of emotional clarity (pre-α = 0.735 and post-α = 0.675). Example questions included: “I experience my emotions as overwhelming and out of control” and “I am confused about how I feel” with item responses ranging from 0 = Never to 4 = Always with higher scores equating to greater regulation deficits. Subscale scores were averaged and used separately in the analyses.

Behavioral regulation and overall executive functioning were measured using the youths’ self-report of Behavior Regulation Index of Executive Function (Guy et al., 2004). Executive Functioning was measured using the Behavior Rating Index of Executive Function-Self Report (BRIEF-SR) (Guy et al., 2004). The BRIEF includes 80-items measured on a 3-point scale (0 = Never; 1 = Sometimes; 2 = Often) with higher scores indicating greater behavioral regulation deficits. The instrument has been validated with two overarching dimensions of executive functioning: Behavior Regulation Index (BRI) and Metacognition Index (MI). Taken together, all of the items create the Global Executive Composite score (GEC). The BRI was used for this study and included subscales of Inhibit (e.g., “I blurt things out”) (pre-α = 0.823 and post-α = 0.710) and Behavioral Shift (e.g., “I have trouble changing from one activity to another”) (pre-α = 0.715 and post-α = 0.763). The overall BRI scores were averaged and yielded strong internal consistencies at pre (α = 0.905) and post (α = 0.878), and the overall GEC scores were averaged and yielded strong internal consistencies at pre (α = 0.958) and post (α = 0.958).

Problem Sexual Behaviors

The Adolescent Sexual Behavior Inventory (ASBI; Friedrich et al., 2004) is a youth and caregiver self-report measure of unsafe and risky sexual behaviors among adolescents that has been validated. The tool includes 44 questions that assess sexual knowledge, sexual risk, concerns about appearance, divergent sexual interest, and fear or discomfort over the past three months. Example questions include “The youth has been caught in a sexual act” or “The youth talks about sexual behaviors frequently” with item responses ranging from 1 = Not True to 3 = Very True. The youth self-report ASBI scores were averaged with strong internal consistency at pre-test (α = 0.809) and post-test (α = 0.845). The caregiver report scores were averaged with acceptable internal consistency at pre-test (α = 0.755) and questionably unacceptable at post-test (α = 0.645). Again, the post-test consistency scores are likely low given the sample size changes from pre-to-post with only 5 post-test caregiver assessments. They were used in the analyses in a preliminary way and larger sample sizes are needed for further use of this tool.

Qualitative

Semi-structured interview guides with open-ended questions were used to collect data for focus groups. The post-training focus groups included questions such as a) Was the frequency and duration of training appropriate; b) Did you feel adequately trained in the model to follow it with fidelity; c) In what ways did you use the bi-weekly consultation services? The post-implementation focus groups with clinicians included questions such as a) What challenges arouse during recruitment of dyads; b) how widely was the approach accepted by your agency; and c) how well did the program model work (duration, frequency, intensity, dosage)? The post-training and implementation focus group questions with the trainers included questions such as a) What areas of the training was the most useful for clinicians; b) Did the clinicians consume information well; and c) What training elements would you change or add?

Analyses

Qualitative

The data were transcribed by the built-in- zoom software. Two individuals coded the data, including one member of the research team and one student. The use of at least two coders supports inter-rater reliability and triangulation among observers (Padgett, 2016).

The team developed a plan analyzing the data or a coding schema (Saldana, 2016) that included open-, first-, and second-cycle coding methods that align closely with the purpose of the feasibility study that sought to initially understand the practical nature of this intervention. Open cycle coding used analytic memos to capture initial concepts emerging. In order to establish an agreed-upon codebook, the research team independently coded—using open cycle coding—the same portion of a transcript. Initial inter-rater reliability was found to be lower than 35%. The researchers subsequently convened to improve coding consistency. After coding a different portion of the same transcript, inter-rater reliability increased to approximately 95%. These agreed upon codes informed the code book with the associated definitions (e.g., concept of acceptability with codes: “positive feelings of the intervention”, “burden”, “coherence or understanding of the intervention”, or “self-efficacy in implementing intervention”. After open cycle coding, the team conducted process coding to determine how the intervention is implemented via a pragmatic approach (Saldana, 2016). The second cycle of coding employed pattern coding to identify common concepts and emergent themes within the data. This was accomplished through constant comparison, a method that involves reflexively comparing elements both within and between individual interviews and focus groups. This process allowed for the identification of commonalities and differences within and across transcripts. Finally, the research team met again to identify common themes, discuss emergence of concepts, and debrief to verify findings. Qualitative rigor was upheld in this study by incorporating multiple coders; triangulating data from focus groups, interviews, and written memos; and conducting inter-rater reliability checks during the interview process.

Quantitative Between group Pre- and Post-Test

Quantitative Data analyses were conducted in IBM Statistical Package for the Social Sciences (SPSS; Version 29.0) using two-tailed p-values. Items from each scale and subscale, accordingly, were averaged for each response and then means over time (pre and post) were compared. To assess change in individual youth and caregiver reports of youth over the course of the intervention, paired sample t-tests were performed. For all t-tests examining group differences, an effect size (Cohen’s d) was also calculated to quantify the magnitude of the effect (0.20 = small; 0.50 = medium, and 0.80 = large; Cohen, 1988). To determine between group differences between caregivers and youth on reports of cumulative trauma events, a between groups t-test was performed.

Results

Qualitative: Feasibility and Acceptability of Training

Clinicians identified many strengths of the 3-day training reporting that it was clear and organized. The structure of this 3-day four-hour sequential training created flexibility for clinician’s busy schedules and opportunity to easily internalize materials without feeling overwhelmed. The clinicians indicated that the use of examples and role plays helped them practice the content and offered a good opportunity to translate findings into practice. They indicated that the training reinforced what skills and components were known about trauma-focused work and offered opportunities to integrate it into practice. The clinicians indicated that the support and co-facilitation by two trainers with unique experience in PSB-A and TF-CBT created a greater level of insight. They also indicated the training materials were comprehensive and well-constructed that offered tangible opportunities to engage with their clients. However, there were also challenges noted by the clinicians, with some suggesting the training information offered little opportunities to practice integrating TF-CBT into existing PSB work, making it challenging to know how to concurrently address trauma and PSB. The clinicians also indicated that because the training was virtual, they were challenged with not having “hard copies” of materials that could be useful when organizing and structuring implementation sessions.

Qualitative: Feasibility and Acceptability of TF-CBT for PSB-A

Implementation Enhancing Factors

Overall, the results showcased the feasibility, practicality, and acceptability of TF-CBT for both the community-practice agency and juvenile justice agency in this study. Clinicians and organizations found this to be a needed approach that can offer great promise to their clients. They highlighted the great need for more trauma-related work in the field of PSB-A and noted the lack of evidenced-based approaches using trauma elements. They also reported that there is a great need for the enhanced services that address trauma and co-occurring PSB. Results showed setting level factors that enhanced implementation. For instance, the Department of Juvenile Justice has a built-in reimbursement structure with salaried employees who are reimbursed around the clock and not only when they meet with clients. Thus, the time spent training and during consultations did not create a financial burden and offered opportunities for engagement and investment in professional development.

The community-based setting integrated a preliminary automated screening approach to intake processes whereby caregivers were asked about known and reported adolescent trauma. This created a triage process for the dyad to be recruited for participation. The community-based setting was able to offer some more flexibility in their schedules and meet with clients and caregivers via telehealth. Telehealth was indicated as a better option for engaging caregivers as it allowed them to be more involved in therapy as opposed to when youth were offered services in person during predetermined or structured time points. Clinicians determined that caregivers appreciated a flexible schedule for meeting times outside of traditional work hours and telehealth allowed for greater accommodations for caregivers. Further, while some of the community-based youths were under probation supervision, they were not under the same constraints as facility-based dyads making it easier to sustain involvement with caregivers and create consistency in schedules. After implementation began, clinicians noted the usefulness of the consultation calls and opportunities to share cases, create feedback loops, and receive support and mentorship.

Implementation Barriers. Some factors emerged in the data that showcased barriers to implementation with some variation by setting. Across both settings, clinicians noted the need for more case examples or role play opportunities to practice integration of TF-CBT within the context of existing PSB-A work. It was also noted that implementation for some may be challenging and may require a paradigm shift away from an “offense specific” lens to a more trauma-inclusive, contextual, and a victim and offender narrative. Setting level factors for DJJ emerged including a shift in the programming and placement settings for youth, requirements for clinicians and staff to travel great distances to meet 1:1 with clients, and clients that presented with more complex needs including greater human services involvement, caregiver risks (i.e., incarceration, substance use etc.), lengthily stays in facility-based settings, and overall staff-reported greater risk (although risk assessments were not conducted as part of this study). The barriers to implementation among community-based providers included some logistical difficulties with scheduling appointments during after school hours; the reimbursement structure in the state that only allowed for 20 mental health sessions yearly; and the overall pay structure where clinicians are only reimbursed when meeting with clients and not during training or consultation hours devoted to this project.

Quantitative

Between-Group Analyses: Trauma Events

There was a statistically significant difference in youth-reports of cumulative trauma and caregiver reports of youth’s experiences of cumulative trauma. Youth reported experiencing more trauma than their caregivers reported on their behalf; youth reported an average of 6.62 events (SD = 2.5), while caregivers reported youth experienced an average of 4.86 events (SD = 2.47), t = 2.89, p = 0.034.

Within Group Analyses: Outcomes

Results showed promising trends and some statistically significant findings when youth and caregiver reports were analyzed. Overall, given the small sample size, the vast majority of outcomes of interest did not yield statistical significance. Youth reported reductions from pre-to post-test in the extent to which trauma interferes with general happiness (t = 2.65, p = 0.033, ES = 5.35). Caregivers reported reductions from pre-to post-test in adolescent problem sexual behaviors (t = 2.91, p = 0.044, ES = 0.140). Although the results were not statistically significant, adolescents also report reductions in problem sexual behaviors. Some other interesting trends were noted in the results. For instance, youth and caregiver report reductions in trauma symptoms and improvements in caregiver relationships. Yet, there were discrepancies in youth-report and caregiver report on the extent to which trauma interfered with daily functioning; adolescents indicated reductions in functional impacts from trauma symptoms on almost all functional domains at post-test and caregivers reported only some functional domain reductions (school and work, family relationships, and general happiness with the latter two trending statistically significant, yet they also note increases in functional impairment in other domains.

Some other results trending in the right direction were adolescent self-reports of emotional regulation, with expected trending changes in all domains of emotional regulation from pre-to post-test. Yet, there were some unexpected and inverse effects in behavioral regulation within the subscales of overall behavioral regulation and regulation inhibitions. The composite of executive functioning did not change and reported difficulties with behavioral shift improved. The results also trended in the expected direction for non-sexual delinquency behaviors from pre to post-test. Full results on all outcomes are presented in Table 1.

Table 1.

Within group pre- and post-test results

		Youth Self-Report			Effect Size		Caregiver Report			Effect Size
	Mean Pre	Mean Post	T-Statistic	P-Value		Mean Pre	Mean Post	T-Statistic	P-value
Trauma Symptoms	1.98	1.87	.458	.661	–	1.67	1.51	1.40	.412	–
Trauma Interferes with Getting Along with Others	.25	.125	1.00	.351	–	.33	.67	− 1.00	.363	–
Trauma Interferes with Hobbies and Fun	.38	.25	1.00	.351	–	.17	.83	− 2.00	.102	–
Trauma Interferes with School or Work	.5	.5	.00	1	–	.50	.33	.542	.611	–
Trauma Interferes with Family Relationships	.5	.375	.552	.598	–	.83	.33	2.24	.076	–
Trauma Interferes with General Happiness	.63	.125	2.65	.033*	.535	.83	.33	2.24	.076	–
Positive Caregiver Relationship	1.66	1.71	-.386	.711	–	2.38	2.60	-.136	.631	–
Composite Behavioral Regulation Difficulties	1.68	1.76	-.677	.520	–	–	–	–	–	–
Composite Executive Functioning Difficulties	1.69	1.67	.126	.903	–	–	–	–	–	–
Regulation Inhibitions Difficulties	1.79	1.93	− 1.34	.222	–	–	–	–	–	–
Behavioral Shift Difficulties	1.77	1.62	.636	.505	–	–	–	–	–	–
Lack of Acceptance of Emotional Responses	2.22	2.04	.887	.547	–	–	–	–	–	–
Limited Goal Directed Behavior	2.97	2.65	.829	.434	–	–	–	–	–	–
Difficulty with Impulse Control	2.2	1.89	.966	.366	–	–	–	–	–	–
Emotional Awareness	2.81	3.22	.368	.296	–	–	–	–	–	–
Limited Emotional Regulation Strategies	2.12	2.14	− 1.13	.970	–	–	–	–	–	–
Limited Emotional Clarity	2.2	2.12	.346	.740	–	–	–	–	–	–
Problem Sexual Behavior	1.43	1.38	.460	.659	–	1.37	1.19	2.91	.044*	.140
Non-Sexual Delinquency	1.86	1.79	.799	.451	–	–	–	–	–	–

Effect sizes are only present for statistically significant results

^*p <.05

Discussion

Impact and Implications

This study sought to conduct an initial pilot test of the feasibility, acceptability, and promise of the use of an enhanced version of TF-CBT for PSB-A. Overall, the results offer some great potential for this approach in leading to positive outcomes among youth-caregiver dyads. The quantitative results demonstrated some promise in change from pre-to post-test on the treatment targets including functional responses to trauma symptoms and caregiver-reported PSB behaviors. However, there were differences between youth and caregiver reports, suggesting the need for ongoing study in robust samples and triangulating reports from youth and caregivers to ensure consistency in outcomes. While some outcome indicators were trending in the right direction towards statistical significance, other findings were trending in the opposite and unexpected direction. These included some behavioral regulation subscales suggesting for further evaluation of true change in behavioral regulation. Further, results may be trending on these specific indicators if the therapeutic elements supported the increase in awareness of negative emotional states associated with trauma or SV related behaviors and consequently, youths are still obtaining mastery over regulating or modulating behavioral responses to such emotional distress.

The qualitative results showed that the intervention can be feasibly delivered in a community-based practice and juvenile justice setting and is widely acceptable among the clinical groups. While there were some noted barriers to implementation, this approach was widely touted as beneficial and imperative to the work with this population. Perhaps this sentiment reflets the larger importance of framing trauma related services as interconnected with and integral to PSB-A work and not a separate and distinct approach only reserved for those who have experienced victimization. Fragmenting services that include PSB-A work and only sometimes integrate trauma targets may lead to disjointed or ineffective services. There is a critical need in the field to continue implementing and studying trauma related work that may already be unfolding, whether it is TF-CBT approaches or others. Thus, it is important to think about the results from this feasibility study in context and how a study like this can lead to changes in our conceptualizations of the PSB-A and subsequent theories of change and treatment targets.

This necessitates a similar paradigm shift in justice system responses and treatment offerings to consider rehabilitative approaches on the spectrum that targets both early life victimization risk and potential ongoing PSB-A risk. Addressing trauma can possibly lead to longer term reductions in recidivism given its high association with onset and course (Pettus-Davis et al., 2019; Pettus-Davis et al., 2019). Interrupting these associated consequences of trauma exposures can address both risks for ongoing problem sexual behavior and trauma related symptoms to more comprehensively promote health and wellbeing among these youth. Indeed, given that the field of sexual abuse has largely focused on victimization and offending as binary and mutually exclusive categories, secondary prevention programs are somewhat elusive and not well defined (McCartan et al., 2018). TF-CBT for PSB-A is particularly well-suited for secondary prevention given the focus on early identification of trauma and concurrent initial or early detection of PSB-A risk with concurrent caregiver engagement that offers sustained resilience and wellbeing.

Importantly, the feasibility and acceptability results underscore the need for further refinements in training and implementation across both service settings. Implementation science focuses on how to effectively translate evidence-based interventions into real-world practice. It recognizes that even the most effective interventions can fail to produce desired outcomes if they are not implemented correctly (Bond & Drake, 2020). While this study yielded important facilitators to implementation, there were noted barriers that can advance training, manual components, and implementation modifications. Barriers that included setting level factors, reimbursement systems, state policy, level of clinician knowledge, and prominently, caregiver involvement. Such barriers can inform robust consideration of these factors prior to training and contribute to readiness assessments that measure such factors and collaborative opportunities to build capacity within agencies, organizations, and clinicians. Indeed, these implementation challenges have already informed a readiness process for TF-CBT for PSB-A that triangulates multiple reports from organizations. Based on self-reported readiness, coaching strategies have been identified for offsetting readiness challenges to fully equip them to adopt a new approach. The results from this study can inform nuanced clinician consultations regarding ways to build trust, empathy, and rapport when caregiver engagement is compromised. In fact, a caregiver resource guide has been developed specifically for this integrated TF-CBT for PSB-A model given the disproportionate experiences of shame experienced by caregivers and parents. Such a resource guide can support clinicians in identifying challenges and most effectively supporting sustained involvement from caregivers.

Future studies on TF-CBT for PSB-A ought to advance the testing of setting-level factors that influence implementation. Further consideration of the ways in which the approach meets the diverse needs of incarceration and community-based settings and attends to the diverse youth within their care. There are many opportunities to build in flexible approaches for youth in this model while simultaneously targeting treatment needs with fidelity. However, identifying those differential needs based on diverse backgrounds, race and culture, family circumstance, developmental stages, offending profiles, or trauma histories need to be clearly delineated and understood. Cultural and contextual factors may also influence the way the clinicians implement the intervention. For instance, the juvenile justice system often prioritizes punitive measures over trauma-informed care due to systemic pressures like resource limitations, public safety concerns, and a focus on accountability rather than rehabilitation. Thus, clinicians working with adjudicated youth may not be adequately oriented to balancing trauma-specific services in combination with accountability treatments necessitating coaching around readiness and greater support for addressing these contextual implementation factors.

Future studies on TF-CBT for PSB-A should also integrate and study the implementation checklists enhanced for this model. In this pilot, the clinicians were encouraged to utilize of fidelity checklists during sessions. However, the application of these checklists was inconsistent across clinicians and primarily used as guides for discussion points during the bi-monthly consultations and not given to researchers for study. The implementation checklists should be used for study to determine the extent to which the model was followed with fidelity.

Limitations

Feasibility and acceptability studies with small sample sizes in community-based and juvenile justice settings face several limitations. First, the very small sample size in this study limited statistical power, making it difficult to detect true significant effects, failing to detect a real effect, leading to wider confidence intervals, or unstable effect sizes from any single outlier making it difficult to draw definitive conclusions. Small sample sizes hinder exploratory analyses, such as subgroup analyses, and increase the risk of Type I error, where a difference is falsely concluded. Overall, such small sample sizes significantly constrain analytic results and should be interpreted with caution as very preliminary. The small sample size of dyads and clinicians creates challenges in generalizing findings as they are not representative of the target population. Additionally, identifying trauma, adversity, or even problem sexual behaviors is challenging with small sample sizes, potentially leading to overestimation or underestimation of risks. Assessing feasibility acceptability and outcomes via self-report instruments and focus groups can also be limited by subjective measures and potential bias in participant responses. While triangulation with caregivers and youth was achieved through different measures, it should be acknowledged that bias may have greatly influenced differential results. Self-report data is subject to several limitations including desirability bias, memory limitations, and other response biases such as acquiescence or extreme responding, can skew results. Further, it should be acknowledged that there may be social desirability bias with clinicians and trainers providing answers that they believe are socially acceptable or desirable, rather than their true beliefs or behaviors. Finally, the lack of control or comparison group limits the ability to achieve internal validity in this study; the absence of a control group makes it impossible to isolate the specific effects of the intervention, leaving uncertainty about whether observed changes are due to TF-CBT for PSB-A or other extraneous factors. These results should be interpreted as pilot and preliminary with further and more robust research to be conducted.

Research Implications

This small pilot feasibility and acceptability study provides valuable preliminary data, and has several implications for future research. This study's findings will inform the design of future work that will refine the intervention protocol by further identifying potential barriers to implementation, optimizing recruitment strategies, and informing training elements. In future research, clinician"competence"should be assessed to employ TF-CBT for PSB-A after training is completed. This could serve as some form of provider readiness or inclusion criteria, ultimately creating a stronger study. Additionally, the study's outcome data, though limited, can serve as a baseline for comparison in future research. It is essential to conduct larger-scale, more rigorous studies with a control group to further investigate the preliminary effectiveness of TF-CBT for PSB-A intervention in diverse settings and populations. By addressing the limitations of this pilot study and building upon its findings, researchers can contribute to the development of trauma-specific evidence-based interventions to improve outcomes for youth.

It is important to acknowledge that while it is necessary to extend this work in many new and innovative ways, there are often larger conceptual and system level barriers to conducting high-quality research in the field of youth sexual violence. There is often limited funding or funding sources dedicated towards addressing the needs of youth who may have detected or early detected problem sexual behaviors. While funding has increased in the prevention field, there is limited funding supporting trauma-specific research endeavors for secondary or tertiary prevention for youth sexual violence. Further, randomizing groups is fraught with challenges in ensuring that contamination is minimized when trained clinicians are implementing the approach. Further, the control group should have a standard of care from which to adequately compare to the intervention group. Yet, in the field, there is not one singular or evidence-supported approach that is consistently applied to the treatment of youth who commit SV, making it difficult to standardize a comparison condition. Finally, there is a need to have sustained clinician expertise in PSB, but sufficient grounding in TF-CBT to allow them opportunities to practice integration and caregiver involvement that contributes meaningfully to the therapeutic outcomes.

Authors'Contributions

All authors contributed equally to the development and dissemination of this manuscript.

Funding

There are no external funding mechanisms to report.

Data Availability

The data are not available for use. A data sharing agreement with entities outside the immediate research team was not approved by the IRB and given sensitive nature of the data collected, this would likely not be supported.

Declarations

Ethical Approval

This study received full IRB approval from the WIRB prior to the completion of data collection. Parental consents and youth assents were granted for youth and caregiver participation prior to the pre and post-tests. Prisoner representatives were part of the IRB review and offered input regarding upholding rights and building in safeguards for youth and caregivers.

Competing Interest

There are not conflicts or competing interests to report.