Show Summary Details
Page of

(p. 19) Theory, Treatment Development, and Research 

(p. 19) Theory, Treatment Development, and Research
Chapter:
(p. 19) Theory, Treatment Development, and Research
Author(s):

Esther Deblinger

, Anthony P. Mannarino

, Judith A. Cohen

, Melissa K. Runyon

, and Anne H. Heflin

DOI:
10.1093/med:psych/9780199358748.003.0002
Page of

Subscriber: null; date: 16 December 2019

As noted earlier, trauma-focused cognitive-behavioral therapy (TF-CBT) was originally developed for use with children, adolescents, and nonoffending parents in the aftermath of child sexual abuse (CSA; Cohen & Mannarino, 1996a, 1998b; Deblinger & Heflin, 1996; Deblinger, Lippmann, & Steer, 1996; Deblinger, McLeer, & Henry, 1990). The model was designed to help caregivers and children cope with the many stressors associated with CSA with a particular focus on addressing posttraumatic stress disorder (PTSD) symptoms, generalized anxiety, depression, feelings of shame, as well as children’s oppositional behaviors, age-inappropriate sexual behaviors, and/or other behavioral difficulties. Since then, the developers have adapted this model for children who have experienced a wide array of other traumas including exposure to domestic violence, traumatic loss, and widespread disasters (Cohen, Mannarino, & Deblinger, 2006). Moreover, while TF-CBT was originally designed to be delivered in individual therapy format in outpatient clinics, its effectiveness in group therapy settings has also been documented (Deblinger, Pollio, & Dorsey, 2015; Deblinger, Stauffer, & Steer, 2001; McMullen, O’Callaghan, Shannon, Black, & Eakin, 2013; O’Callaghan, McMullen, Shannon, Rafferty, & Black, 2013; Stauffer & Deblinger, 1996). More recently, TF-CBT has been used successfully with children and adolescents residing in foster care (Dorsey et al., 2014; Lyons, Weiner, & Scheider, 2006), those placed in residential treatment facilities (Cohen, Mannarino, & Navarro, 2012), as well as boy soldiers and girls who have been sexually exploited in the context of war (McMullen et al., 2013; O’Callaghan et al., 2013).

Theory Underlying the Development of Trauma Symptoms

TF-CBT integrates ideas and information from a variety of theoretical models, including family, empowerment, attachment, developmental neurobiology, (p. 20) and humanistic theories. Cognitive behavioral theory, however, provided the initial and primary foundation and rationale for the development of TF-CBT. Cognitive behavioral therapy is based on the central premise that cognitions, behaviors, emotions, and bodily sensations are highly interdependent. Thus, an intervention that specifically targets one of these areas of human functioning is expected to indirectly affect the other aspects of human functioning and adjustment. TF-CBT, in fact, integrates interventions that target behaviors, cognitions, emotions, and bodily sensations individually and in combination.

To conceptualize the etiology and treatment of psychological symptoms developed in the aftermath of sexually abusive experiences, the CBT theoretical model integrates learning theory particularly the influence of conditioning, contingencies, and models in the environment, with the impact of cognitive factors (Kendall, 1985). This model is used below to explain the development and maintenance of sexual abuse-related symptoms in children as well as the authors’ conceptualization of the theoretical mechanisms underlying the efficacy of TF-CBT.

Observational learning

Modeling is an example of a simple learning process that can explain the development of both positive and negative behaviors in children. Children constantly imitate what they see and hear. Thus, through observational learning, children who have experienced CSA sometimes imitate verbalizations and behaviors exhibited by the perpetrators of their abuse. For example, the child may use foul language or engage in age-inappropriate sexual behaviors as a result of observing the offender. In other words, children’s abuse-related behaviors may be simple reflections of the observational learning processes much like other more common behaviors children learn through observation.

Children’s cognitive views and developing beliefs may also be significantly influenced by role models in their environment. Sex offenders may model dysfunctional attitudes regarding the abuse, sexuality, relationships, the trustworthiness of others, and so on. Even nonoffending, supportive parents may inadvertently contribute to the development of unhealthy views or beliefs. For example, when a well-meaning parent acts as though the sexual abuse is the worst thing that could ever have happened to the child, he/she may unintentionally encourage the child to adopt the same catastrophic view of the experience. In the aftermath of a sexual abuse disclosure, there are also many opportunities for children to learn positive coping behaviors by observing the responses of significant others. For example, if nonoffending parents discuss the sexual abuse in a calm, open, and direct manner, the child is likely to imitate that style in coping with his own thoughts, feelings, and concerns regarding the abuse. Indeed, as noted earlier, nonoffending parents may be children’s most influential role models, though the distress they experience themselves may make it difficult for them to be at their best in the aftermath of CSA.

(p. 21) Respondent and instrumental conditioning

Another learning mechanism that may explain the development of children’s abuse-related symptoms is referred to as two-factor theory. Two-factor learning theory suggests that fears are acquired through respondent conditioning and maintained through instrumental conditioning (Mowrer, 1939). According to respondent conditioning principles, when neutral stimuli are paired with unconditioned fear-evoking stimuli, the neutral stimuli alone begin to elicit fear responses. For example, for children who are sexually abused in the dark, darkness may be the previously neutral stimulus that is paired with the unconditional fear-evoking stimulus: the sexual abuse. As a result of the learned association between darkness and sexual abuse, a child’s fear of being sexually abused in the dark may generalize to a fear of darkness.

When fear responses lead to avoidance of previously neutral stimuli, instrumental conditioning comes into play. Each time avoidance behavior occurs, it is negatively reinforced by a reduction in anxiety, thereby increasing the likelihood that the avoidance behavior will recur. To continue with the example provided above, children who learned to fear darkness as a result of the association between darkness and abuse now begin to avoid darkness to avoid experiencing the fear. Each time they successfully avoid the dark, they experience a reinforcing reduction in anxiety, which increases the likelihood of future attempts to avoid the dark. This association may then lead to problem behaviors including sleep refusal or insistence on sleeping with the lights on.

Recent revisions of the classical two-factor theory suggest other mechanisms for the development and maintenance of avoidance behavior. According to approach-withdrawal theory, avoidance may be maintained as a result of the positively reinforcing qualities of relaxation and/or other safety cues that follow avoidance behavior. Thus the children’s avoidance of the dark is also positively reinforced by the sense of relief they experience upon successful avoidance, as well as by any other positive consequences such as increased parental attention in response to their verbalized fears of the dark. This increased attention sometimes takes the form of sleeping in the parents’ bed, which can become reinforcing not only for its tendency to reduce anxiety but also because of the positive reinforcing feelings of closeness, warmth, safety, and attention.

Through the processes of generalization and higher order conditioning, a wider range of previously neutral stimuli may be paired with fear-evoking stimuli. Subsequently, previously neutral stimuli begin to elicit fear and avoidance responses. For example, the children described above initially may have feared only the darkness in their own bedrooms because that was where the abuse occurred. However, through generalization and higher order conditioning, they learned to fear darkness in general, across different settings. Such responses then become increasingly debilitating as the innocuous stimuli (e.g., darkness) that elicit troubling symptoms generalize from anxiety only when alone in the darkness of their bedroom, to anxiety in any dark room, to being extremely uneasy outside at night but not really understanding what is precipitating these symptoms of anxiety.

(p. 22) These conditioning principles provide a framework for understanding the development of posttrauma symptoms often exhibited by children who have experienced CSA. Many children experience feelings of fear, anxiety, pain, sadness, shame, and/or anger during the episodes of sexual abuse. Although they naturally associate these negative feelings with sexual abuse, some children, particularly those experiencing PTSD, generalize these feelings of distress from the actual experience of abuse to nonthreatening abuse-related cues such as darkness, men, being alone with one person, certain tones of voice, specific smells, bathrooms, enclosed spaces, getting undressed, etc. These cues are not in and of themselves dangerous, but because of their association with the abuse, they may trigger the emotions the children experienced during the abuse. Moreover, any cues that trigger CSA-related memories or thoughts may similarly come to be associated with anxiety, fear, anger, shame, and/or other distressing emotions experienced at the time of the abuse. Thus, in an effort to avoid these disturbing emotions, many children work hard to avoid specific abuse-related stimuli, while also actively avoiding thinking, talking, or being reminded of the abuse.

It should also be emphasized that many children use the same coping mechanisms that were seemingly adaptive in responding to sexually abusive experiences (e.g., dissociation, denial, numbing, and avoidance) to cope with innocuous abuse-related stimuli and/or other stressors in their day-to-day lives (e.g., peer pressure, test anxiety, parent-child conflict). Unfortunately, dissociative and avoidant coping strategies are not particularly successful when children apply them to the everyday stressors of child and adolescent development. In fact, relying on maladaptive strategies such as denial, avoidance, and/or dissociation to cope with abuse-related cues may inadvertently strengthen the inappropriate associations made between innocuous reminders of the abuse and psychological distress. For example, each time an innocuous abuse-related cue is avoided (e.g., darkness), a child experiences a reduction in distress and/or an increased feeling of safety that reinforces the avoidance behavior and strengthens the association between innocuous abuse reminders and emotional distress. Thus, although these coping mechanisms may have helped the child survive the abuse, their continued use once the abuse has ended may be maladaptive. Such continued avoidance may cause children to needlessly avoid innocuous situations, potentially preventing them from enjoying positive and important learning experiences. For example, children who feel anxious around and/or try to avoid dark-haired men or boys because their perpetrator had dark hair may be hindered from participating effectively in situations with other dark-haired males, such as teachers, coaches, neighbors, or peers. Furthermore, continued avoidance of abuse-related thoughts and memories may prevent these children from effectively processing and understanding their abusive experiences, potentially leaving them with misperceptions and inaccurate cognitive schemas related to the abuse. Furthermore, as noted earlier, there is evidence that both adult and child survivors of sexual abuse who use strategies of avoidance and suppression to cope with memories of CSA tend to be more symptomatic than those who rely (p. 23) on more active, constructive coping mechanisms (Leitenberg et al., 1992; Simon et al., 2010).

Theoretical Models Underlying the TF-CBT Approach

Prior empirical investigations have demonstrated the effectiveness of prolonged exposure in treating the symptoms experienced by adult PTSD sufferers (Fairbank & Keane, 1982; Foa, Rothbaum, Riggs, & Murdock, 1991). It seemed natural to build on the groundbreaking work of Dr. Edna Foa and her colleagues, whose work focused on the treatment of adult rape survivors. In fact, prolonged exposure seemed the treatment of choice for children experiencing PTSD symptoms in the aftermath of CSA. However, during early clinical trials with children, it quickly became evident that, unlike adult patients, children were often unwilling to subject themselves to anxiety-provoking stimuli for prolonged periods in exchange for the promise of long-term therapeutic benefits. In addition, many nonoffending parents expressed concern about the high levels of anxiety associated with immediate and prolonged exposure. Systematic desensitization, on the other hand, offered the advantage of using a gradual hierarchy of anxiety-provoking stimuli, which might be easier for children to tolerate. However, all the fears and anxieties children experience may not be apparent at the outset of therapy, and they may not fit neatly into a hierarchy. In addition, it is sometimes difficult to engage children, particularly young and/or developmentally delayed children, in the visualization and/or relaxation exercises required for systematic desensitization procedures.

To meet the needs of children, the first author and her colleagues at the Medical College of Pennsylvania combined elements of systematic desensitization and prolonged exposure in designing the intervention referred to as gradual exposure that essentially begins at the very start of treatment (Deblinger & Heflin, 1996; Deblinger et al., 1990). Like systematic desensitization, gradual exposure encourages children to confront feared stimuli, such as reminders, thoughts, and memories of the abuse, in a graduated fashion. Even though it can be helpful, a highly detailed hierarchy need not be constructed when developing a treatment plan for young children. It is important to assess the degree to which various stimuli or memories provoke anxious or avoidant responses and offer children choices for gradual exposure exercises, thereby ensuring that the process allows them to maintain control in choosing only slightly more anxiety provoking memories to review across sessions. Thus, the therapists may develop a more informal hierarchy as an overall plan that need not be shared with the child if it would provoke anticipatory anxiety. However, with some older children and adolescents, it may give them a greater sense of control if they collaborate in creating a general hierarchy in the middle stage of treatment when they are more confident in their ability to confront abuse-related memories and reminders. Initially, children will be encouraged to endure low-level anxiety-provoking (p. 24) stimuli before moving on to confront more distressing stimuli. For example, in the initial phase of therapy, psychoeducation about CSA may be shared in the abstract because that is less anxiety provoking than encouraging a discussion of the child’s own abusive experience(s). Thus, the therapist may begin by engaging the child in discussions of factual information regarding sexual abuse such as prevalence, impact, reactions and dynamics of CSA. Then over time and during some coping skills training, the child may be asked to describe only her feelings related to the abusive experience and later during the trauma narrative component, the child may be asked to discuss the actual details of the least distressing of his or her own abusive experiences before moving on to discuss more anxiety-provoking memories in detail. As is the case with prolonged exposure, gradual exposure and processing exercises may be repeated until traumatic memories and innocuous stimuli in the environment no longer elicit maladaptive anxiety or avoidance.

With exposure to traumatic memories and/or different anxiety-provoking stimulus, it is expected that the child’s anxiety level may increase somewhat. However, the increase in anxiety seems to reflect the increased anxiety the child experiences naturally upon exposure to abuse-related stimuli outside the therapy setting, rather than being a unique consequence of the therapeutic work. It is the experience of the authors that most children and their parents are able to tolerate this temporary increase in anxiety if it is explained in advance and they understand the long-term purpose of the work. Over time, the repeated exposures to abuse-related stimuli in a safe therapeutic setting will lead to a decrease in anxiety level. By the end of treatment, the child is expected to confront abuse reminders and discuss abuse-related memories without experiencing significant distress. Thus, gradual exposure aims to disrupt the maladaptive associations between innocuous abuse-related cues and the more extreme negative emotions that develop as a result of respondent conditioning. Moreover, when habituation occurs, new associations replace the old ones; that is, adaptive responses such as feelings of control, comfort, bravery, and/or pride become connected to previously anxiety-provoking memories, situations, thoughts, and/or discussion.

In addition to breaking the associations between innocuous abuse-related cues and distressing emotions, the connection needs to be disrupted between avoidance of innocuous abuse-related cues and positive or negative reinforcement. Repeated experiences with reduced anxiety as a result of avoidance behaviors strengthen avoidance and escape habits. Thus, it is very important for children to endure the anxiety-provoking thoughts and/or cues until anxiety decreases naturally without the child engaging in avoidance. In so doing, the child will learn that distress decreases without resorting to the use of escape strategies such as avoidance, dissociation, numbing, acting out, and so on, thereby resulting in decreased dependence on these maladaptive coping mechanisms. In addition, the child will learn that feelings of safety and mastery may be achieved in the face of memories, thoughts, and reminders of the abuse. Many parents also benefit from participating in gradual exposure exercises. At the start of therapy, (p. 25) many parents have less information than the investigator or caseworker about their child’s abuse. Parents often experience high levels of anxiety when confronted with thoughts or reminders of their child’s abusive experience and thus may seek to avoid those reminders as well. In exhibiting such avoidant behavior, parents model ineffective coping responses for their children. Thus, parents are encouraged to participate in gradual exposure exercises in order to become more comfortable with abuse-related stimuli, information, and discussion themselves, so that they can model more successful coping strategies for their children.

For many parents, the development of effective parenting and communication skills is central to their children’s recovery particularly when acting-out behavior problems are exhibited by their children. Acting-out behaviors in children who have experienced sexual abuse may consist of angry outbursts that are sometimes associated with PTSD. They may also reflect behaviors learned over the course of the abuse that are being maintained through inadvertent reinforcement by parents and others.

In summary, the treatment approach described in this book includes educational, skill building, gradual exposure, and processing interventions designed to break the problematic associations children have made between negative feelings and abuse-related cues, such as memories and innocuous reminders. In that way, children become more comfortable confronting and making sense of abuse-related memories. Furthermore, this therapeutic work decreases children’s reliance on maladaptive coping responses to innocuous abuse reminders such as avoidance and dissociative responses. TF-CBT also uses the modeling process described earlier as the therapist models ways to communicate about and cope effectively with troubling abuse-related thoughts, feelings, and reminders. In addition, the therapist coaches the nonoffending parent to model effective coping strategies for the child and to respond more effectively to abuse-related disclosures and behavior problems. More information about the TF-CBT PRACTICE components and the course of treatment from session to session is provided in the chapters that follow.

TF-CBT Development and Research

TF-CBT has its roots in scientific principles derived from general research in psychology as well as treatment outcome research. Many of the component interventions of TF-CBT were included in the model because of their demonstrated effectiveness with populations experiencing symptoms similar to those of children who have experienced sexual abuse. Moreover, as noted, early investigations found cognitive behavioral interventions to be effective in treating posttraumatic stress in adult sexual assault survivors (Foa, Rothbaum, & Ette, 1993; Foa et al., 1991). However, in the course of developing TF-CBT, it became evident that significant modifications were required to apply exposure based CBT interventions to children experiencing PTSD, including incorporating parent participation into the treatment model.

(p. 26) Dr. Deblinger and her colleagues and Drs. Cohen and Mannarino began conducting research studies to enhance the understanding of the therapeutic needs of children who experienced sexual abuse initially through independent studies at separate clinical research sites in the mid-1980s (Cohen & Mannarino, 1988; Deblinger et al., 1989; Mannarino & Cohen, 1986; Mannarino, Cohen, & Gregor, 1989; Mannarino, Cohen, Smith, & Moore-Motily, 1991; McLeer et al., 1988). A preliminary empirical evaluation of the model referred to today as TF-CBT examined the treatment response of children (ages 3 to 16 years) who were experiencing PTSD in the aftermath of CSA (Deblinger et al., 1990). The findings demonstrated no changes in children’s symptomatology during a baseline period, but showed significant symptom improvements pre- to posttreatment. Although children’s PTSD, depression, and behavioral symptoms improved significantly, at post treatment some children continued to remain in the mild depressive range. To better understand the impact of the parent and child interventions on specific symptom domains, Deblinger and colleagues next conducted a study examining the impact of the child and parent interventions separately and in combination. In this study, children and their nonoffending parents were randomly assigned to one of four treatment conditions: the full TF-CBT model including both the parent and child interventions; a child-only TF-CBT intervention; a parent-only TF-CBT intervention (in which parents were taught to serve as their child’s therapeutic agent); or a community treatment comparison condition (Deblinger et al., 1996). The findings of this investigation documented that children assigned to conditions in which they actively participated in TF-CBT (i.e., child only or child/parent conditions) showed significantly greater reductions in PTSD symptoms as compared to children who were not active TF-CBT participants (i.e., parent-only or community comparison). In addition, parents assigned to conditions in which they actively participated in the implementation of TF-CBT (i.e., parent or child/parent conditions) showed significantly greater improvements in parenting practices, and their children exhibited significantly greater reductions in externalizing behavior problems as well as depression compared to conditions in which parents did not participate in TF-CBT (i.e., child-only or community comparison). Most importantly, the significant improvements found at posttreatment were sustained over a two-year follow-up period (Deblinger et al., 1999). This model was further evaluated for its efficacy in a group therapy format. Again, after demonstrating the benefits of TF-CBT in a group format through a preliminary pre-post investigation (Stauffer & Deblinger, 1996), Deblinger and colleagues conducted an additional randomized trial in which young children who had experienced CSA were randomly assigned along with their nonoffending parents to either a TF-CBT group program or education/support group program. The results of this investigation demonstrated that parents assigned to the TF-CBT group program experienced significantly greater improvements with respect to parental abuse-specific distress as compared to those assigned to the support group. Children assigned to the TF-CBT group program showed greater improvements with respect to body safety skills and knowledge as compared to those assigned to the educational/support groups (Deblinger et al., 2001).

(p. 27) During approximately this same time frame, Drs. Cohen and Mannarino began conducting treatment outcome investigations on an early version of TF-CBT as well. In fact, several studies were conducted examining the responses of preschool and school-age children to TF-CBT for CSA as compared to a nondirective supportive therapy condition. The findings in their initial study with preschool children (ages 3 to 6 years) demonstrated that children randomly assigned to TF-CBT showed significantly greater symptom improvements with respect to internalizing symptoms, as well as general behavior problems as compared to those children assigned to the nondirective supportive counseling condition (Cohen & Mannarino, 1996a). Moreover, from a clinical perspective, Cohen and Mannarino (1996a) found that TF-CBT was more effective than the alternative treatment condition in terms of helping children overcome potentially dangerous acting-out behaviors including sexually inappropriate behaviors toward others and was more efficacious in maintaining symptom reductions at the 1-year follow up assessment (Cohen & Mannarino, 1997). The findings of another investigation demonstrated that school-age children (ages 7 to 14 years) who were randomly assigned to TF-CBT in the aftermath of CSA exhibited significantly greater improvements with respect to depression, anxiety, sexual problems, and dissociation at the six-month follow-up as compared to children assigned to the nondirective supportive therapy condition (Cohen & Mannarino, 1998b; Cohen, Mannarino, & Knudsen, 2005). In addition, replicating their findings with preschoolers, the clinical findings of this investigation also suggested that TF-CBT seemed to be more effective than the nondirective therapy approach in addressing children’s sexually inappropriate behaviors. Cohen, Mannarino, Perel, and Staron (2007) also conducted an independent study examining the potential benefits of psychiatric medication in addition to TF-CBT in addressing children’s therapeutic needs in the aftermath of CSA. The findings indicated that with the exception of TF-CBT with sertraline medication being associated with significantly greater improvements in observed child global assessment ratings, TF-CBT with sertraline medication produced no significant benefits above and beyond TF-CBT alone with regard to PTSD, depression, anxiety, behavior problems, and all other clinical outcome measures. Although these findings should be interpreted with caution due to the small and unrepresentative nature of the study’s sample, the authors conclude that the current evidence suggests that for most children with PTSD, including those with co-morbid depression, an initial trial of TF-CBT generally should be offered before combining trauma-focused therapy with medication.

The treatment approach described in this book reflects the integration of the developers’ early independent treatment models (Cohen & Mannarino, 1993; Deblinger & Heflin, 1996), as well as the findings of TF-CBT treatment outcome research and the authors’ clinical experiences to date. In fact, in the mid-1990s the authors conducted the first large-scale multisite collaborative investigation in which participants were randomly assigned to the integrated TF-CBT approach or a child centered therapy condition (CCT) across both clinical research sites (Cohen, Deblinger, Mannarino, & Steer, 2004). The findings of this investigation (p. 28) demonstrated significantly greater benefits of TF-CBT over the child-centered therapy condition in terms of the outcomes of both child and nonoffending parent. Children assigned to TF-CBT demonstrated significantly greater improvements with respect to PTSD, depression, abuse attributions, shame, behavior problems, and personal safety skills as compared to children assigned to the comparison treatment. The nonoffending parents assigned to TF-CBT also reported significantly greater improvements with respect to their own abuse-related distress, personal feelings of depression, parenting skills, and support levels as compared to those parents assigned to the comparison treatment condition (Cohen, Deblinger, et al., 2004). Furthermore, the findings of a follow-up investigation demonstrated that participant improvements across conditions were generally sustained over a one-year period, and children who experienced multiple traumas as well as comorbid depression exhibited poorer outcomes, but only in response to the child centered therapy (Deblinger, Mannarino, Cohen, & Steer, 2006).

Most recently, a dismantling study was conducted examining the impact of treatment length (eight versus 16 TF-CBT sessions) and the completion of a written narrative in the context of TF-CBT delivered to young children (ages 4 to 11 years) with a history of CSA (Deblinger et al., 2011). The results of this study suggested that a longer course of treatment (i.e., 16 sessions) and a greater focus on skill building appeared to be critical in helping parents enhance their parenting skills and children overcome externalizing behavior problems. The findings also documented the benefits of the TF-CBT trauma narrative component in most efficiently and efficaciously helping children overcome abuse-related fear and general anxiety in as little as eight TF-CBT sessions. Recently, Salloum and Overstreet (2012) conducted a similar study in which the results also suggested the value of trauma narration in addition to coping skills training particularly for highly distressed children. A follow-up to the dismantling investigation demonstrated that participants across all TF-CBT treatment conditions generally sustained their symptom improvements over a one-year posttreatment period (Mannarino, Cohen, Deblinger, Runyon, & Steer, 2012). The results also indicated that higher levels of internalizing and depressive symptomatology predicted the small minority of children (i.e., 11%) who continued to meet PTSD criteria at the one-year follow-up (Mannarino et al., 2012). From a clinical perspective, it is interesting to note that replicating earlier findings, many children assigned to the written narrative conditions reported that talking about the sexual abuse was the most helpful part of therapy (Deblinger et al., 2011; Deblinger, Mannarino, et al., 2006).

Over the years, many of the findings of the above-described studies have been replicated by other researchers that have documented the generalizability of the impact of TF-CBT with diverse populations of children and families impacted by CSA. Researchers in Australia, for example, randomly assigned children who had been sexually abused to one of three conditions: a parent and child TF-CBT condition, a child only TF-CBT condition, or a waiting list control condition. Not surprisingly, children assigned to the TF-CBT conditions exhibited greater improvements with respect to child-reported PTSD, fear and anxiety, as well as (p. 29) parent and clinician ratings of global functioning as compared to those children assigned to the waitlist control (King, Tonge, Mullen, et al., 2000). Interestingly, differences with respect to the child only and parent and child conditions emerged at the three-month follow-up, with children assigned to the parent and child TF-CBT condition exhibiting significantly less abuse-related fear than children assigned to the child only TF-CBT condition. The authors suggest that additional hypothesized differences between child only and parent and child TF-CBT may not have been detected due to the small sample size and the lack of focus in the parent and child condition on parenting skills training and the emotional impact of the sexual abuse on the parents.

Another recent randomized controlled trial conducted in the Democratic Republic of the Congo examined the efficacy of TF-CBT delivered in group format to girls who had experienced sexual exploitation, as well as numerous other traumas (O’Callaghan et al., 2013). In fact, the average number of childhood traumas experienced by this sample of girls was approximately 12 (as compared to an average of three to four traumas experienced by children in earlier TF-CBT trials). In addition to sexual abuse and war exposure, girls in this study experienced a severe lack of food or water, severe punishment or revenge, exposure to corpses, looting, fighting, attacks, and/or parental loss through death, divorce, and abandonment. Despite this, the results of this randomized controlled trial replicated the findings of other TF-CBT studies demonstrating that girls randomly assigned to a TF-CBT group exhibited significantly greater improvements with respect to trauma symptoms, depression, anxiety, conduct problems, and prosocial behaviors as compared to girls assigned to the wait list. Moreover, the positive changes observed in these domains of functioning at posttreatment were either maintained or further improved at the three-month follow-up assessments. Interestingly, these Congolese girls, including many who had never spoken about their trauma(s), reported that they valued the opportunity to talk about their traumatic experiences, much like children participating in previous TF-CBT trials in the United States (Deblinger et al., 2011; Deblinger, Mannarino, et al., 2006). In addition, it should be noted that the TF-CBT intervention provided in this context seemed to offer additional benefits due to its group format, including the opportunity to experience peer support from those with similar experiences (Deblinger et al., 2015).

Further clinical development of TF-CBT for children, adolescents, and nonoffending caregivers continues to be closely tied to the empirical literature based on the findings of an increasing number of treatment efficacy, effectiveness, and process studies conducted by the current authors as well as other researchers across the United States and the world (Cohen, Mannarino, & Knudsen, 2004; Cohen, Mannarino, & Staron, 2006; Murray, Familiar, et al., 2013; Murray et al., 2015; O’Donnell et al., 2014; Webb, Hayes, Grasso, Laurenceau, & Deblinger, 2014). These studies have continued to document the efficacy of TF-CBT in addressing CSA as well as a wide array of other childhood traumas including traumatic grief, abandonment, community violence, and other traumas. Several recent studies, in fact, have demonstrated the superior benefits of TF-CBT for (p. 30) children in foster care who often have experienced a wide array of traumas. More specifically, in a quasi-experimental study, Lyons et al. (2006) found that children who received TF-CBT as opposed to usual care not only exhibited significantly greater reductions in traumatic stress symptoms but were also one-tenth as likely to run away from the foster home and about half as likely to experience placement disruptions. This reduction in risk of running away from home and removal from foster homes is critically important given the associations found between placement disruptions and escalating emotional and behavioral problems in foster children (Chamberlain et al., 2006; Leathers, 2006). Moreover, it should be noted that a recent study documented the critical importance of engaging foster parents in the TF-CBT process. The results of this study documented that children and foster parents randomly assigned to TF-CBT with evidence based engagement procedures as opposed to TF-CBT without such procedures, were significantly less likely to drop out prematurely and significantly more likely to complete treatment (Dorsey et al., 2014). In another recent TF-CBT study, the results documented the superior benefits of a brief eight-session TF-CBT model delivered to children and their caregivers in shelters after exposure to domestic violence (Cohen, Mannarino, & Iyengar, 2011). In this context, children who received TF-CBT in comparison to usual care were significantly more likely to exhibit PTSD diagnostic remission and experienced fewer serious adverse events. The benefits of TF-CBT also have been documented with respect to addressing the psychosocial impact of widespread disasters such as 9/11 (CATS Consortium, 2010) and Hurricane Katrina (Jaycox et al., 2010). In addition, the results of another study conducted in the Democratic Republic of Congo demonstrated that TF-CBT provided in group format to boy soldiers (McMullen et al., 2013) produced significantly greater improvements in trauma symptoms, depression, conduct problems, as well as prosocial behaviors than did the waitlist control condition. Another randomized trial conducted in the Netherlands randomly assigned children to TF-CBT vs. Eye Movement Desensitization and Reprocessing (EMDR). The findings of this investigation documented that both interventions produced significant reductions in PTSD symptoms, however, only the children assigned to TF-CBT exhibited significant reductions in depression and hyperactive symptoms as well (Diehl, Opmeer, Boer, Mannarino, & Lindauer, 2015). Finally, a recently published randomized trial conducted in Norway (Jensen et al., 2014) replicated earlier TF-CBT findings and documented PTSD remission rates among the Norwegian participants (ages 10 to 18 years) that were almost identical to those found in the first multisite TF-CBT study conducted in the United States (Cohen, Deblinger, et al., 2004).

Table 2.1 provided here summarizes the samples, methods, and findings of the initial TF-CBT treatment outcome studies designed to address the therapeutic needs of children and their nonoffending caregivers in the aftermath of CSA as well as more recent TF-CBT studies that focus on other populations of youth impacted by diverse childhood traumas. The efficacy of TF-CBT in addressing other traumas, as described above, is important for the CSA population, given that research has demonstrated that children who have been sexually (p. 31) (p. 32) (p. 33) (p. 34) (p. 35) (p. 36) (p. 37) (p. 38) (p. 39) (p. 40) (p. 41) abused frequently have experienced other childhood traumas as well including exposure to domestic violence, traumatic loss, and other diverse traumas (Finkelhor et al., 2007a). Although TF-CBT is not the only treatment approach with empirical documentation of its therapeutic efficacy with children who have experienced trauma, it does appear to have the strongest research track record (Jensen et al., 2014; Saunders, Berliner, & Hanson, 2004; Silverman et al., 2008). To date, there have been at least 25 empirical investigations of this model including 15 completed randomized controlled trials, with the largest proportion of those studies focusing specifically on children with a history of sexual abuse (Cohen, Deblinger, et al., 2004; Cohen & Mannarino, 1996a, 1998b; Deblinger et al., 1996; Deblinger et al., 2011; Deblinger et al., 2001; King, Tonge, Mullen, et al., 2000) (see Table 2.1). In addition, recent TF-CBT outcome research has not only replicated its efficacy with diverse community samples but has also established the cost-effectiveness of this short term model as compared to treatment as usual in the community (Greer, Grasso, Cohen, & Webb, 2014). Moreover, additional research is underway aimed at developing stepped care approaches to the delivery of TF-CBT to further reduce costs and enhance treatment accessibility, by more effectively matching treatment dose and intensity to children’s needs (Salloum et al., 2014).

Table 2.1. TF-CBT Outcome Studies with Children, Adolescents, and Nonoffending Caregivers Impacted by CSA

Study

Target Population

Country of Study

# of Sessions

Design

Treatment/Control

Major Findings

Final Follow-Up

Deblinger, McLeer, & Henry, 1990

Children ages 3–16 years

U.S.

12

Pre-post

19 TF-CBT

Significant improvement in symptoms of anxiety, PTSD and depression as well as significant improvements in behavior problems

N/A

Cohen & Mannarino, 1996a

Children ages 3–6 years & their nonoffending parents

U.S.

12

Random assignment

  • 39 TF-CBT

  • 28 NST

Compared to NST condition, those in TF-CBT condition had significantly greater improvement in PTSD symptoms, internalizing symptoms, and sexual behavior problems

TF-CBT group was more efficacious than NST group in maintaining symptom reduction at 12 month follow up (Cohen & Mannarino, 1997)

Deblinger, Lippmann, & Steer, 1996

Children ages 7–13 years & their nonoffending mothers

U.S.

12

Random assignment

  • 22 TF-CBT Parent Only

  • 24 TF-CBT Child Only

  • 22 TF-CBT Parent + Child

  • 22 Community Control

Compared to conditions without child intervention, conditions with the TF-CBT child intervention had significantly greater reduction in PTSD symptoms. Compared to conditions without parent intervention, conditions with the TF-CBT parent intervention had greater reductions in children’s externalizing behavior problems and depression and significantly greater improvements in parenting practices.

Improvements in PTSD symptoms, depression, and externalizing behavior problems were maintained over the two-year follow up period (Deblinger, Steer, & Lippmann, 1999)

Stauffer & Deblinger, 1996

Children ages 2–6 years & their nonoffending mothers

U.S.

11

Pre-post

19 TF-CBT

Significant decreases in children’s sexualized behaviors and parental distress and significant improvements in parenting practices

Improvements maintained at 3-month follow-up

Cohen & Mannarino, 1998b

Children ages 7–14 years & their nonoffending parents

U.S.

12

Random assignment

  • 30 TF-CBT

  • 19 NST

Significantly greater improvement in depressive symptoms and social competence in TF-CBT condition at post-treatment compared to NST

Significantly greater improvement in PTSD and dissociative symptoms in TF-CBT condition at 12-month follow up (Cohen, Mannarino, & Knudsen, 2005)

King, Tonge, Mullen, et al., 2000

Children ages 5–17 years

Australia

20

Random assignment

  • 12 TF-CBT Child

  • 12 TF-CBT Family

  • 12 WL

Significantly greater improvements across TF-CBT conditions in PTSD symptoms, fear, and anxiety as well as overall functioning compared to WL condition

Improvements were maintained at 12 week follow up. At follow-up, children assigned to the family CBT condition showed significantly greater improvement on abuse-related fear relative to those assigned to child only CBT.

Deblinger, Stauffer, & Steer, 2001

Children ages 2–8 years & their nonoffending mothers

U.S.

11

Random assignment

  • 21 TF-CBT Group

  • 23 Support Group

Compared to support group, TF-CBT group had significantly greater improvement in child-reported knowledge of body safety skills and significantly greater reductions in maternal abuse-related intrusive thoughts and negative parental emotional reactions to the abuse

Improvements maintained at 3-month follow-up. Mothers in support group showed significant reductions in emotional distress though they tended to obtain more additional help during follow-up compared to mothers in TF-CBT group.

Cohen, Deblinger, Mannarino, & Steer, 2004

Children ages 8–14 years & their primary caregivers

U.S.

12

Random assignment

  • 89 TF-CBT

  • 91 CCT

Compared to CCT, TF-CBT condition had significantly greater reductions in children’s PTSD and depressive symptoms, behavior problems, feelings of shame, and abuse-related attributions, as well as greater improvements in caregivers’ depression, abuse-related distress, parenting practices, and parental support of the child

Compared to CCT condition, children in TF-CBT condition reported significantly fewer PTSD symptoms and feelings of shame at 6 and 12 month follow-up. Parents in TF-CBT condition reported less abuse-specific distress at 6 and 12 month follow-up (Deblinger, Mannarino, Cohen, & Steer, 2006)

Cohen, Mannarino, Perel, & Staron, 2007

Girls ages 10–17 years

U.S.

12

Random assignment

  • 12 TF-CBT + sertraline

  • 12 TF-CBT + placebo

Both conditions showed significant pre-post improvement on PTSD and other symptoms. No significant differences between conditions except on ratings of global functioning, which favored the TF-CBT plus sertraline group

N/A

Deblinger, Mannarino, Cohen, Runyon, & Steer, 2011

Children ages 4–11 years & their nonoffending caregivers

U.S.

  • 8 or

  • 16

Random assignment

  • 40 TF-CBT 8 sessions no TN

  • 39 TF-CBT 8 sessions with TN

  • 35 TF-CBT 16 sessions no TN

  • 44 TF-CBT 16 sessions with TN

  • Significant improvements reported across conditions on behavioral, emotional, and skill-based outcomes.

  • 8 sessions with TN condition was significantly more efficient and effective in helping children overcome abuse-related fear and general anxiety as well as in helping parents overcome abuse-specific distress.

  • 16 sessions no TN condition showed significantly greater improvements with respect to children’s externalizing behavior problems and effective parenting practices.

Improvements reported at post-treatment were maintained at 6 and 12 month follow up; additional improvements on parental emotional distress and on child self-reported anxiety were reported at 12 month follow-up (Mannarino, Cohen, Deblinger, Runyon, & Steer, 2012)

O’Callaghan, McMullen, Shannon, Rafferty, & Black, 2013

War-affected girls exposed to sexual exploitation and multiple traumas ages 12–17 years

Democratic Republic of Congo

15

Random assignment

  • 24 group-based, culturally modified TF-CBT

  • 28 WL

Significantly greater reductions in trauma symptoms and significant improvement in depression and anxiety symptoms, conduct problems, and prosocial behavior in TF-CBT group as compared to WL

N/A

Treatment Outcome Studies with Children, Adolescents, and Nonoffending Caregivers Impacted by Other Traumas

Study

Target Population

Country of study

# of sessions

Design

Treatment/Control

Major findings

Final Follow-Up

Cohen, Mannarino, & Knudsen, 2004

  • Children ages 6–17 years & their primary caregivers;

  • Trauma type = CTG

U.S.

16

Pre-post

22 TF-CBT-CTG

Significant reductions in children’s behavioral problems & depressive, anxiety, PTSD, and CTG symptoms. Significant reductions in depressive and PTSD symptoms of parents

N/A

Cohen, Mannarino, & Staron, 2006

  • Children ages 6–17 years & their parents

  • Trauma type = CTG

U.S.

12

Pre-post

39 TF-CBT-CTG

Significant reductions in children’s behavioral problems and PTSD, CTG, anxiety, and depressive symptoms.

N/A

Lyons, Weiner, & Scheider, 2006

Children ages 0–21 years in foster care who experienced a moderate or severe trauma

U.S.

10

Quasi-experimental

  • 69 TF-CBT

  • 82 CPP

  • 65 SPARCS

  • 2218 TAU

Participants in all 3 EBP groups improved significantly in child and adolescent needs and strengths domains. As compared to TAU, youth who received TF-CBT achieved significantly greater gains on traumatic stress symptoms and behavioral/emotional needs, were about one-tenth as likely to run away from placement, and about half as likely to have a disruption in placement.

N/A

CATS Consortium, 2010

Children ages 5 to 21 affected by the World Trade Center disaster on September 11th

U.S.

  • Trauma-specific CBT (TF-CBT for ages 5–12; Trauma and Grief Component Therapy for Adolescents for ages 13–21) 8 to 12 sessions

  • Brief skills CBT 4 sessions

Quasi-experimental

  • 239 trauma specific CBT

  • 67 brief skills CBT

Both groups improved over time (trauma symptoms decreased). Although those who received TF-CBT began treatment with considerable family/environmental adversity and greater trauma-exposure than the brief skills CBT participants, the TF-CBT participants experienced significantly more clinical improvement moving from a probable diagnosis of PTSD to the mild range.

N/A

Jaycox et al., 2010

  • Fourth through eighth graders in New Orleans schools post Hurricane Katrina who reported PTSD symptoms

  • Trauma type = hurricane exposures, vicarious traumas, and personal traumas

U.S.

  • TF-CBT 12 sessions

  • CBITS 10 group sessions, 1–3 individual sessions

Random assignment

  • 60 TF-CBT

  • 58 CBITS

Significant reduction in PTSD symptoms of participants in both treatments

N/A

Cohen, Mannarino, & Iyengar, 2011

  • Children ages 7–14 years

  • Trauma type = exposed to IPV

U.S.

8

Random assignment

  • 64 TF-CBT

  • 60 CCT (usual care)

TF-CBT participants in IPV shelter had significantly greater decreases in anxiety symptoms, hyperarousal and avoidance symptoms as compared to CCT

N/A

McMullen, O’Callaghan, Shannon, Black, & Eakin, 2013

Former child soldiers and other war affected boys ages 13–17 years

Democratic Republic of Congo

15

Random assignment

  • 25 TF-CBT

  • 25 WL

Compared to WL control, TF-CBT group had significantly greater reductions in posttraumatic stress, depressive, and anxiety symptoms, overall distress and conduct problems as well as significant increase in prosocial behaviors

Improvements maintained at 3-month follow up

Murray, Familiar, et al., 2013

Orphans and vulnerable children ages 5–18 years who experienced one or more traumatic events

Zambia

8–23 sessions (average = 11)

Pre-post

58 TF-CBT

Post-treatment assessment results showed significant reductions in trauma symptom severity and severity of shame symptoms.

N/A

Dorsey et al., 2014

  • Children in foster care ages 6–15 years and their foster parents

  • Trauma type = multiple

U.S.

  • M = 15.0 for TF-CBT + engagement

  • M = 12.3 TF-CBT

Random assignment

  • 22 TF-CBT

  • 25 TF-CBT + engagement

Significantly greater retention and completion rates among those who received engagement strategies compared to those who did not. Across treatments, all children displayed significant improvements in regard to clinical outcomes at post treatment.

Improvements maintained at 3-month follow up

Greer, Grasso, Cohen, & Webb, 2014

  • Children ages 7–18 years

  • Trauma type = multiple

U.S.

Not reported

Quasi-experimental

  • 90 TF-CBT

  • 90 TAU

Over the course of a year, 2 times more money was spent on low-end mental health services for youth who received TF-CBT as compared to youth who received TAU and five times more money was spent on high-end mental health services for youth who received TAU as compared to youth who received TF-CBT

N/A

Jensen et al, 2014

  • Youth ages 10 to 18 years

  • Trauma type = different traumas

Norway

12–15

Random assignment

  • 79 TF-CBT

  • 77 TAU

Compared to TAU group, TF-CBT group had significantly lower posttraumatic stress symptoms, depression, and general mental health symptoms as well as significantly greater improvements in functional impairment

N/A

O’Donnell et al., 2014

  • Children ages 7 to 13 years

  • Trauma type = death of parent(s)

Tanzania

12 weekly sessions plus 3 individual visits

Pre-post

64 TF-CBT

Pre to post-treatment improvements in symptoms of grief, posttramatic stress, and depression as well as overall behavioural adjustment

Improvements maintained 3 and 12 months post-treatment

Webb, Hayes, Grasso, Laurenceau, & Deblinger, 2014

  • Youth ages 7 to 16 years with history of trauma and symptoms of PTSD

  • Trauma type = different trauma

U.S.

M = 10

Pre-post

72 TF-CBT

Assessments at 3- and 6- months post intake indicate that participants showed a significant decrease in PTSD symptoms, as well as internalizing and externalizing problems.

Decreases in PTSD symptoms and internalizing and externalizing problems were maintained at the 9- and 12-month follow up assessments.

Diehl, Opmeer, Boer, Mannarino, & Lindauer, 2015

  • Children ages 8-18 years and their nonoffending parents

  • Trauma type = different traumas

Netherlands

Maximum of 8 weekly sessions

Random assignment

  • 25 EMDR

  • 22 TF-CBT

Both conditions showed pre- to post-treatment reductions in symptoms of posttraumatic stress. Parents of children in TF-CBT condition only reported significant reductions in symptoms of depression and hyperactivity.

N/A

Murray et al., 2015

  • Children ages 8–18 years

  • Trauma type = multiple

Zambia

10–16 sessions

Random assignment

  • 131 TF-CBT

  • 126 TAU

TF-CBT led to more positive outcomes than TAU

N/A

Note. TF-CBT, Trauma-Focused Cognitive Behavioral Therapy; PTSD, Posttraumatic Stress Disorder; NST, Nondirective Supportive Therapy; WL, Waitlist; CCT, Child Centered Therapy; TN, Trauma Narrative; CTG, Childhood Traumatic Grief; CPP, Child-Parent Psychotherapy; SPARCS, Structured Psychotherapy for Adolescents Responding to Chronic Stress; EMDR, Eye Movement Desensitization and Reprocessing; TAU, Therapy as Usual; CBITS, Cognitive-Behavioral Intervention for Trauma; IPV, Intimate Partner Violence; CCT, Child Centered Therapy.

TF-CBT continues to evolve as a result of the ongoing research designed to enhance its overall implementation and effectiveness. However, to date, it should be noted that TF-CBT has been given the highest ratings for its efficacy and dissemination based on extensive reviews conducted by the U.S. Department of Health and Human Services (www.nrepp.samhsa.gov), the California Evidence Based Clearinghouse for Child Welfare (www.cebc4cw.org), and the Kaufman Best Practices Task Force Final Report (Chadwick Center for Children and Families, 2004). Finally, the results of a recent systematic review of evidence-based treatments for children exposed to maltreatment suggest that TF-CBT is the best-supported treatment and thus it is often recommended as the treatment of choice for this population (Leenarts, Diehle, Doreleijers, Jansma, & Lindauer, 2013). (p. 42)