Failures in Cognitivebehavior Therapy for Children

Failures in Cognitive-Behavior Therapy for Children

Yasmin Rey, Carla E. Marin, and Wendy K. Silverman

Florida International University

This article discusses treatment failures in child therapy, specifically cognitive-behavioral therapy (CBT) for anxiety and its disorders. The theoretical foundations and principles of CBT are discussed first, followed by a summary of the treatment outcome literature. Also discussed is how treatment failure is defined and gauged in CBT, as well as factors implicated in treatment failure. A case illustration highlights these factors, which resulted in the child not advancing positively in treatment. The article concludes with key practice recommendations. © 2011 Wiley Periodicals, Inc. J Clin Psychol: In Session 67:1140-1150, 2011.

Keywords: Adolescent; Anxiety; CBT; Child; Psychotherapy; Treatment failure

Correspondence concerning this article should be addressed to: Wendy K. Silverman, Department of Psychology, Florida International University, 11200 S.W. 8th Street, Miami, Florida, 33199; e-mail:

Published online in Wiley Online Library (wileyonlinelibrary.com/journal/jclp). DOI: 10.1002/jclp.20848

Clinical disorders of childhood and adolescence are highly prevalent, afflicting approximately 10% to 20% of youth in the United States (La Greca, Silverman, & Lochman, 2009). About 70% of these youth do not receive treatment services to alleviate the distress and impairment resulting from these disorders (U.S. Department of Health and Human Services, 1999). Further, many of the youth receiving treatment are not receiving treatments that have been empirically found ‘‘to work’’ (Silverman & Hinshaw, 2008). More importantly, and especially pertinent to this Journal issue, even if a child or adolescent receives an evidence-based treatment, it cannot be assumed that positive treatment response will ensue. A substantial portion (e.g., 20% to 40% of anxious youth) who receive evidence-based treatments fail to respond positively (McKay & Storch, 2009).

In this article, we discuss treatment failure in child cognitive-behavior therapy. Given the broadness of ‘‘child therapy,’’ we focus our discussion in our area of specialization: anxiety and its disorders. As will be apparent, many of the principles we discuss are applicable to other disorders of childhood and adolescence. This is because these disorders, like anxiety, can be reduced with cognitive-behavioral treatment (CBT).

CBT for Youth

When working with children and adolescents with anxiety disorders, a key CBT procedure is exposure (Silverman & Kurtines, 1996). Exposure involves the individual confronting fear or anxiety provoking objects or situations in a gradual fashion, either live (i.e., in vivo) or in imagination. Gradual exposures are typically implemented along a fear hierarchy. A fear hierarchy lists items that represent the specific feared objects or situations, ranked from least to most fearful. Each item on the fear hierarchy is then used in-sessions and out-of-sessions as an exposure task. Typically, exposures begin with low fearful items and over treatment the exposures focus on the more fearful items.

The precise mechanism by which exposure works to reduce anxiety is unclear. However, all explanations involve modification of behavioral, cognitive, and affective processes. For example, exposure to anxiety-provoking situations or objects may lead to anxiety reduction by extinguishing (or removing) the physiological arousal associated with the feared object or situation. Alternatively, exposure may allow the youth to receive incompatible information that perceived catastrophic consequences associated with anxious stimuli do not occur.

To facilitate youths’ learning during exposures, cognitive strategies are used. Most common is self-control strategies, which rely on self-observation, self-modification, self-evaluation, and self-reward. In our work, we teach youth to use the acronym STOP during in and out of session exposures (Silverman & Kurtines, 1996). Youth learn to first identify when they are feeling anxious or Scared (S), then to identify their anxious Thoughts (T). Then they learn to modify or restructure their anxious thoughts by generating other alternative coping thoughts and behaviors (O). Finally, youth learn to evaluate their performance in confronting their fears during exposure tasks, and to reward or Praise themselves for confronting their fears (P). To help modify irrational thoughts (or change Ts’’ to O’s), youth are taught to look for evidence for their anxious thoughts and to then identify a more realistic thought based on the evidence or lack of evidence.

Parental Involvement

When parents appear to be maintaining their child’s anxiety problems, especially avoidant behaviors, it may be helpful to get parents involved in their child’s treatment. To help parents learn how to decrease their child’s avoidant behaviors, a common behavioral procedure used is contingency management. Contingency management emphasizes training parents in the appropriate use of contingencies by contracting. In contingency contracting, the therapist helps the child and parent devise weekly contracts. The contracts specify the details of the exposure task the child is to perform in the coming week and the specific reward the parent is to provide to the child contingent on the child’s completion of the exposure. To further assist parents, parents are trained in reinforcement and extinction. Using these skills, parents are encouraged to provide positive consequences (praise, tangible rewards) following their child’s efforts in facing the anxiety-provoking situations. Parents also are encouraged not to allow (or to extinguish) their child’s avoidance of the anxiety provoking objects or situations.

Outcome Research

A large research literature has accumulated that provides strong and consistent evidence for the efficacy of CBT to reduce anxiety and its disorders in youth (Silverman, Pina, & Viswesvaran, 2008). CBT is efficacious in reducing anxiety disorders whether delivered to the child individually, to the child and parent together, and in a group format. Most of the RCTs involve random assignment of youth participants to a CBT condition versus a waitlist control. A smaller number of studies have randomized youth to CBT versus an active, credible comparison control condition. These studies involve multimethod-multisource assessment procedures to evaluate treatment outcome.

Most studies report recovery rates ranging from about 60% to 80% of youth no longer meeting diagnostic criteria for their targeted anxiety diagnosis at post-treatment (Silverman et al., 2008). Most studies also report statistically significant reductions in youth and parent anxiety symptom ratings. A smaller number of studies also have reported statistically significant reductions in teacher ratings, as well as in behavioral observation ratings of the youths’ anxiety.

Defining and Gauging Failure in CBT

Despite the generally positive outcomes of CBT for youth anxiety disorders, a significant portion of youth fails to respond to treatment: About 20% to 40% of youth continue to meet diagnostic criteria for their anxiety disorder at the conclusion of treatment (Silverman et al., 2008). These youth also fail to show significant reductions in clinician ratings of disorder severity and youth and parent anxiety symptom ratings. They also are likely not to be viewed as significantly improved by teachers and behavioral observations from pre to post-treatment.

As in almost all youth randomized clinical trials, youth anxiety trials use multimethod- multisource assessment to gauge treatment response and treatment failure. To gauge whether the child or adolescent continues to meet diagnoses, diagnostic interview schedules aretypically employed.

In the anxiety area, the Anxiety Disorders Interview Schedule for DSM-IV: Child and Parent Versions (ADIS-IV: C/P) is the most widely used interview. The ADIS-IV: C/P also contains a Severity Rating Scale, which allows for a global assessment of severity of interference or impairment associated with the youth’s anxiety diagnoses. Ratings on the Severity Rating Scale can range from 0 to 8. After treatment, higher ratings (i.e., four or greater) reflects treatment failure because the child’s diagnoses are still resulting in high interference.

Treatment success and failure also is typically gauged using youth-completed and parent- completed symptom rating scales (Silverman & Ollendick, 2005). The most widely used youth symptom rating scales are the Revised Children’s Manifest Anxiety Scale and Multidimensional Anxiety Scale for Children. The most widely used parent symptom rating scale is the Child Behavior Checklist, specifically, the Internalizing Broadband Scale and/or the Anxious/Depressed Narrowband Scale. A companion teacher version, the Teacher Report Form, also has been used.

Factors Implicated in CBT Failure in Youth

Below we summarize what is known about treatment failure in CBT among youth suffering from anxiety disorders (see McKay & Storch, 2009; Silverman & Hinshaw, 2008).

Symptom Severity

Severity of youths’ symptoms before treatment has been implicated in CBT failure among youth with anxiety disorders (Silverman et al., 2008). For example, youth who retained their primary anxiety diagnoses following CBT had higher pretreatment levels of self-rated trait anxiety and self-rated depressive symptoms than youth who were free of their primary anxiety diagnosis at post-treatment. High pretreatment child rated trait anxiety also was associated with high clinician severity ratings of interference of the youth’s primary anxiety diagnoses at post-treatment. Also implicated is high mother ratings of their child’s social withdrawal symptoms, as well as teacher ratings of the child’s anxious/depressed symptoms (Southam- Gerow, Kendall, & Weersing, 2001).

Cognitive Factors

Youths’ self-statements or self-talk prior to treatment also have been implicated in CBT failure (McKay & Storch, 2009). For example, youths’ anxious self statements, such as ‘‘I am very nervous’’ and ‘‘I am going to make a fool of myself,’’ were associated with less pretreatment to post-treatment reductions in youth self ratings of fear, anxiety, and depressive symptoms. Further, youth with more anxious pretreatment self statements than positive self statements (e.g., ‘‘I am a winner,’’ ‘‘I feel good about myself’’) were associated with less pre to posttreatment reductions in youth self ratings of anxiety symptoms.

Comorbidity

Pretreatment comorbidity is also probably involved in CBT failure among youth (Liber et al., 2010; McKay & Storch, 2009). Youth with a comorbid diagnosis of depression, for example, were more likely to retain their primary anxiety diagnosis following CBT than youth without a comorbid depression diagnosis. Additionally, youth with any type of comorbid diagnoses (e.g., anxiety, depressive, disruptive) were more likely to retain an anxiety diagnosis at posttreatment (Liber et al., 2010). However, some studies have not been able to show that pretreatment comorbidity is implicated in youth anxiety CBT outcome (McKay & Storch, 2009).

Parent Psychopathology

The presence of anxiety disorders in the mothers and fathers of child patients increases the risk of treatment failure (Bodden et al., 2008; Cresswell, Willetts, Murray, Singhal, & Cooper, 2008; Gar & Hudson, 2009). Additionally, depression symptoms in mothers and fathers have been implicated (Liber et al., 2008; McKay & Storch, 2009). Some studies, however, have not found an association between parental depression symptoms and treatment failure (Crawford & Mannassis, 2001; Southam-Gerow, Kendall, & Weersing, 2001).

Parenting Behaviors and Other Family Factors

Maternal over involvement has also been shown to contribute to CBT failure among youth with anxiety disorders. Mothers’ expressions of fear, such as being stiff, tense, and fidgety, have been identified in CBT failure (Cresswell et al., 2008). Fathers’ reports of rejecting their child and children rating their mothers as low in warmth also have been shown as significant predictors of CBT failure (Liber et al., 2008). Mothers, fathers, and youth who report high family dysfunction and mothers who report high parenting stress tend to predict, and probably contribute to, CBT failure.

Treatment Processes

Several treatment processes, defined in different ways across studies, predict lower rates of positive response in CBT. For example, the quality of the therapeutic relationship, defined as the perceived bond between therapist and youth client, predicted higher failure rates (McKay & Storch, 2009). Lower client involvement in CBT, defined as youths’ willingness to participate in treatment aspects such as self-disclosure and engagement, has been implicated as well (McKay & Storch, 2009). Specifically, minimal youth involvement at mid treatment, prior to their participation in exposure tasks, predicted higher clinician severity ratings at posttreatment. Decreases in the youth therapist alliance, defined as both the quality of the therapeutic relationship and the degree of youths’ willingness to participate in treatment, also have been implicated in anxiety treatment failure (Chiu, McLeod, Har, & Wood, 2008).

Case Illustration

Presenting Problem and Client Description

Juan, an 8-year-old Hispanic boy in second grade, was referred to our clinic by his school counselor. He lives with both biological parents and his 5-year-old sister. Juan’s mother telephoned the clinic because he had missed about 30 days of school due to his fear of harm befalling himself and his mother when they are not together. Juan’s separation anxiety was so severe that he could never be left alone anywhere in the house. He also could not sleep by himself at night in his own bed. Juan’s school counselor recommended the family seek help because Juan’s grades had severely declined and he was currently not attending school. Juan was at risk of being expelled from school because of his excessive absenteeism.

Juan’s mother reported that, since the first day of the school year, Juan cried excessively and had temper tantrums every morning she tried bringing him to school. Juan’s crying and tantruming were so severe that his mother sometimes turned around and went home with Juan instead of dropping him off at school, or she picked him up early from school. Juan told his mother he was afraid she would not come back to pick him up at the end of the school day. Mother reported that ever since Juan stopped attending school, he appeared sadder and to have lost interest in activities he used to enjoy.

Juan had difficulties being away from his mother since age 2. Mother described him as ‘‘clingy,’’ and even as a toddler, he showed much anxiety about being in new situations. Juan’s mother provided many examples of how he became highly distressed when she left him with her mother (Juan’s grandmother) so that she could run errands. He cried, threw tantrums, and begged his mother not leave him.

Juan’s mother responded by not running her errands; instead, she stayed home to allay Juan’s separation protests. Mother reported, ‘‘Juan is so worried I will leave and never come back that he does not want me to ever leave his side.’’ She also indicated that her son has been sleeping with her and her husband since toddlerhood. Whenever she tried to get him to sleep in his own bed, he would cry and throw tantrums; she thus ‘‘gives in.’’

Mother also reported that Juan showed difficulties attending and staying in school since he began kindergarten. Juan’s mother was a stay at home mom from the time of Juan’s birth until the start of kindergarten. She noted that Juan had difficulty in making the transition from being home all day with her to having to be away from her for the school day. During the first two weeks of kindergarten, mother reported that Juan cried incessantly and pleaded with her not to leave him. His anxiety about separation from her would not subside despite her reassurances that she will definitely return, on time, to pick him up at the end of the school day.

When Juan first began kindergarten, his mother reported that she walked her son to his classroom and she stayed in the room with him for the duration of the school day. After two weeks, however, the vice principal no longer allowed her to stay in the classroom. Thereafter, his mother had to drop him off at the classroom door. Mother reported that when she left him, ‘‘His screams could be heard all over the school.’’ Juan’s fussing and protesting about going to school and the amount of time it took to soothe him subsided after the first month of kindergarten. After the first month, Juan was able to attend and stay in kindergarten with no subsequent significant difficulties for the remainder of the school year.

Upon entering first grade, Juan again experienced difficulties attending school in the mornings at the start of the school year. Again, however, Juan’s difficulties subsided after the first month of first grade. Currently, in the second grade, Juan’s difficulties in attending school were substantially worse than the previous school years. After about a month and a half from the first day of the second grade, Juan refused going to school altogether. His refusal to attend school has persisted for a month and was still ongoing at the time he presented to our clinic.

Juan’s mother further reported that this year also was different than past years because he was ‘‘down and upset.’’ When Juan learned he was failing all his classes he told his mother, ‘‘I can’t do anything right.’’ According to Juan’s mother, hardly anything ‘‘brought him joy.’’

Juan’s mother reported that her son reached all his developmental milestones ‘‘more or less’’ on time. He learned to walk when he was 1-year-old, and he was toilet trained by age 3. He had no significant delays in speech. There was no history of abuse or neglect. Juan’s mother further reported she never sought treatment for Juan’s difficulties separating from her before.