Diagnosis-Specific Group CBT Treating Social Anxiety in Adolescents: A Feasibility Study

Preparing for an upcoming randomized controlled trial, the main objective of this study was to examine the feasibility (i.e. attrition rate, session attendance, questionnaire completion rates, and treatment satisfaction) and the preliminary clinical outcome of the Danish version of the CK-E for adolescents with SAD conducted as group treatment.

The study took place at Centre for Psychological Treatment of Children and Adolescents (CEBU), a teaching and research facility at the Department of Psychology and Behavioral Sciences, Aarhus University, Denmark. Thirteen adolescents aged 12-17 with a primary diagnosis of SAD were included. We decided that running 3 groups with 4-5 adolescents in each group would be sufficient to get experience with running the group treatment program.

Inclusion criteria were: (a) adolescents between 12 and 17 years of age; (b) SAD as primary disorder according to the DSM-IV (APA, 1994). Exclusion criteria were: (a) diagnosed Autism Spectrum Disorder; (b) untreated Attention Deficit Hyperactivity Disorder (ADHD); (c) psychotic symptoms; (d) current severe self-harm or suicidal ideation; (e) current eating disorder; (f) more than 50% school refusal.

Participants were self-referred, based on information on websites, newspaper advertisements, and handouts to local general practitioners and educational institutions. Interested families sent in a brief description of the adolescent’s major problems. Families judged suitable from their description were offered baseline assessment. Baseline assessment included self-reported questionnaires as well as diagnostic interviews. Families included based on the baseline assessment were contacted and informed of the study processes.

All questionnaires were administered electronically via e-mail, using an online data collection platform (SurveyXact), and assessed at baseline, post treatment, 3-month after treatment and 1-year after treatment. Diagnostic interview were conducted with the Anxiety Disorder Interview Schedule for DSM-IV, Child and Parent Version (ADIS-IV C/P) (30) at baseline, post treatment, and 3-month after treatment. The ADIS-IV C/P were conducted by clinical psychologists experienced in using the interview.

The study has obtained approval from the Institutional Review Board at Aarhus University (j.nr. 2019-616-000013) and the Institutional Data Protection Agency at Aarhus University (j.nr. 2016-051-000001).

The Cool Kids Anxiety Program, Social Enhanced (CK-E), is a manualized CBT program specifically designed to treat adolescents with SAD, developed at the Centre for Emotional Health at Macquarie University, Sydney, Australia. The program is based on the generic Cool Kids Anxiety Program with standard CBT techniques, but including techniques focusing specifically on SAD. These additional techniques are based on the theories of cognitive and behavioral processes maintaining SAD by Clark and Wells (28) and Rapee and Heimberg (29). The additional techniques are: 1. Training in task-focused attention (to reduce self-focus) and in-session behavioral experiments focusing on self-focus and reduction hereof. 2. In-session behavioral experiments used to investigate and reduce safety-seeking behaviors. 3. Cognitive restructuring with focus on overestimation of costs including exercises where participants deliberately makes mistakes to investigate consequences. 4. Interpersonal exposure tasks with video-feedback used to alter adolescents’ distorted negative self-images. 5. Post-event cognitive restructuring used after challenging exposure tasks.

The CK-E has been translated to Danish and adapted to a group intervention by staff at CEBU. This implies change in timeframe (the individual program includes 10 1-h sessions, whereas the group program includes 10 2-h sessions) and setting (groups of 4-5 adolescents in each youth session with a duration of 1 hour, and parents work independently with the therapist typically 30 minutes and the rest of the time the whole group work together and with the therapist). The adaption did not include changes in the content of the manual. The manual comprises a therapist manual and workbooks for both adolescent and parents which were handed out at session one.

The treatment consisted of nine sessions with the adolescents and parents together and one parents-only session (session 5). Sessions 1 through 8 were held weekly and session 9 and 10 with two-week intervals. Three months after ending treatment participants were offered a 1-h booster group session. Table 1 presents an overview of the intervention program.

The treatment was conducted by clinical psychologists from CEBU with 2-8 years of experience from treating adolescent with anxiety disorders with the Cool Kids Anxiety program. As part of an educational program at CEBU three graduate psychology students assisted the adolescents during the in-session assignments.

The feasibility of the social intervention program was evaluated by participant dropout rates, proportion of completed questionnaires, as well as treatment satisfaction. The preliminary efficacy was evaluated by diagnostic status and change in levels of functioning assessed by the questionnaires.

Experience of Service Questionnaire (ESQ) measures participants’ satisfaction with the intervention at post. ESQ is adjusted by CEBU from the Experience of Service Questionnaire (31). There are separate items for parents (10 items) and adolescents (7 items). It includes positive and negative statements, and items were rated on a 3-point Likert scale (0-2). ESQ includes open questions with possibility for qualitative feedback. The qualitative feedback were used in a qualitative analysis which provided the following themes: 1) The therapeutic techniques and methods, 2) the group format, 3) intensity, 4) satisfaction with the therapists, and 5) inclusion of parents in the program.

Anxiety Disorder Interview Schedule for DSM-IV, Child and Parent Version (ADIS-IV C/P) (31) is a semistructured diagnostic interview conducted with adolescents and parents separately to assess the diagnostic criteria in accordance with DSM-IV for anxiety disorders as well as other disorders often comorbid with anxiety (e.g., depression and ADHD). Severity of the diagnoses are measured on a nine-point Likert scale ranging from not disturbed at all to severely disturbed (0, 1, 2, 3, 4, 5, 6, 7, 8). Scores of 4 or greater indicate a clinical diagnosis. Separate scores are made by adolescents, parents, and the clinician, but only the clinician’s severity rating score will be used as an outcome measure. The most impairing diagnosis, as assessed by the clinician, is considered as the primary diagnosis. Both concurrent validity and test-retest reliability has been established for the anxiety disorder section (30, 32).

Spence Children’s Anxiety Scale (SCAS and SCAS-P) (33) was used to measure adolescent- and parent-rated anxiety symptoms. The adolescent version contains 44 items, and the parent version contains 38 items. Items are rated on a four-point Likert scale (0, 1, 2, 3). Higher scores indicate higher levels of anxiety. It consists of six subscales reflecting symptoms specifically related to SAD, panic disorder and agoraphobia, generalized anxiety disorder, obsessive–compulsive disorder, separation anxiety disorder, and fear of physical injury. Each subscale is scored separately and added together for a total score reflecting overall anxiety symptoms. The Danish version of the SCAS has shown good to excellent internal consistency in clinical and non-clinical samples and good test-retest reliability in a nonclinical sample (34).

The Child Anxiety Life Inference Scale (CALIS) (35) was used to measure the impact of adolescent’s anxiety on various areas of life functioning including friends, school, extracurricular, and family. The impact is evaluated separately by adolescents (9 items) and their parents (16 items). Items are evaluated on a five point Likert scale (0, 1, 2, 3, 4). Higher scores indicate a higher degree of life interference. CALIS has showed satisfactory internal consistency and moderate test-retest reliability (36).

The Short version of the Mood and Feelings Questionnaire (S-MFQ) (37) was used to measure depressive symptoms within the last two weeks. The symptoms are evaluated independently by adolescents and parents. The short version includes 13 items rated on a three-point Likert scale (0, 1, 2) with higher scores indicating a higher degree of depressive symptoms. The Danish version of the long version of MFQ has showed good internal consistency and test-retest reliability (37).

Self-Efficacy Questionnaire for Children (SEQ-C) (38) was used to measure adolescents’ self-efficacy. It is a self-rated measure completed by the adolescents. It has three subscales: social self-efficacy, academic self-efficacy and emotional self-efficacy. It consists of 24 items rated on a 5-point likert scale, with higher scores indicating higher self-efficacy. SEQ-C has shown good internal consistency (39).

Subtle Avoidance Frequency Examination (SAFE) (40) is a self-rated measure designed to assess safety behaviors. Adolescents completed the questionnaire. SAFE is designed to incorporate both active safety behaviors, subtle restrictions of behavior, and behaviors aimed at avoiding or concealing physical symptoms. SAFE consists of 32 items ranging on a five point Likert scale (0, 1, 2, 3, 4), with higher scores indicating a higher degree of safety-seeking behaviors. SAFE has showed good internal consistency and test-retest reliability (40, 41). SAFE was originally designed for adults but can be used reliably and validly to assess safety behaviors in adolescents (42).

The Children’s Automatic Thoughts Scale (CATS) (41) measures a range of the adolescents’ self-reported negative self-statements. CATS was completed by the adolescents. CATS include four subscales relating to automatic thoughts on social threat, personal failure, hostility, and physical threat. All items were scored on a five-point Likert scale (0, 1, 2, 3, 4), with higher scores indicating higher degrees of negative automatic thought. Only two subscales are included in the present study; social threat and personal failure. These subscales have showed correlations with self-rated SAD in a prior study (42). CATS has demonstrated good internal consistency and test-retest reliability (43, 44).

Focus of Attention Questionnaire (FAQ) (43) is a self-rated measure of focus of attention. It was completed by the adolescents. FAQ is a 10-item scale including two 5-item subscales; self-focused attention and external-focused attention. All items were rated on a five-point Likert scale (1, 2, 3, 4, 5). The two subscales are treated as independent. Subscale scores were calculated by averaging the five items. A higher score on the two subscales indicate a higher degree of self-focused attention or external focus respectively. High internal consistency has been reported (45).

Postevent version of the Thoughts Questionnaire (PTQ) (46) is a self-rated measure to assess the degree of post event processing. Adolescents completed the PTQ. Consistent with prior research (47, 48) only the 15 negatively worded items of the PTQ were used in this study. All items were rated on a five-point Likert scale (0, 1, 2, 3, 4). Higher scores indicate a higher degree of post-event rumination. Studies have shown excellent internally consistency (49, 50, 51).

Depression Anxiety Stress Scales (DASS) (49) is used in the study for parents’ rating of their own symptoms regarding anxiety, depression and stress. DASS consists of three subscales measuring symptoms of anxiety (DASS-A), depression (DASS-D) and stress (DASS-S). DASS contains 42 item and each item is rated on a four-point Likert scale (0, 1, 2, 3) with higher scores indicating a higher degree of distress. DASS has been found to have good internal consistency and test-retest reliability (50).

Background Information at Baseline. Participants completed a background information questionnaire before entering the interventions. Background questions included information regarding parents’ mental and physical health, adolescent’s mental and physical health, family demographics, adolescent’s previous and/or ongoing treatment and school absenteeism.

Descriptive statistics, in form of mean, standard deviation (SD) and percent, were used to describe sample characteristics, dropout, session attendance, completion of questionnaires and treatment satisfaction.

Preliminary longitudinal clinical outcomes were analyzed using multilevel modeling (MLM) in Stata 16.1 with the mixed command. MLM accounts for the dependence structure due to repeated measurements within subjects and allows for the examination of change within subjects across time (baseline, post, 3-month follow-up, 1-year follow-up). Subjects with partly missing data can be accommodated thereby utilizing all available data.

Main effects of time were modelled specifying time as a level 1 variable governed by a piecewise linear trajectory (e.g. 54). Thus, between baseline and post one specific slope applies and between post and 1 year follow-up another slope applies. The two line segments are joined at post. A slope significantly different from zero corresponds to a statistically significant t-test of the hypothesis of equal outcome means at the two time periods identifying the lines, viz. baseline and post as well as post and 1-year follow-up. All models included random intercepts and random slopes between post and 1-year follow-up. The variance/covariance structure at level 1 was modeled as independent.

Effect-sizes (Cohens d) were calculated for all measures by subtracting post-mean from baseline-mean divided by the pooled standard deviation (Mpre-Mpost) / SDpooled

Thirteen adolescents between 12 and 17 year of age (mean age=14.69, SD=1.25) and nine of them were females (69 %) were enrolled in the study. Eight adolescents (61.5 %) lived with both biological parents. The mean household income among the private or municipal setting. Three of the adolescents (23%) were on psychopharmacological medication (Setraline 100-150mg).

There were no drop-outs during treatment. Ten of the thirteen adolescents attended all sessions (77 %), two (15 %) missed one session and one (8 %) missed two sessions. Four adolescents (30 %) received one extra 30 minutes individual session (with participation of their parents) and one parent received a phone call. Nine (69 %) mothers attended all ten session and four (31 %) mothers attended nine families were 772.308 DKK (SD=248.338), which is close to the average household income in the area at 619.418 DKK. All thirteen adolescents had SAD as primary diagnosis in the ADIS-C/P interview with a CSR at 6 or 7 (mean=6.23, SD=0.43). Eight of the adolescents (61.5%) had a comorbid anxiety disorder (generalized anxiety disorder (78 %), agoraphobia (11 %), specific phobia (11 %)). Furthermore two (15 %) also had a dysthymia disorder. All adolescents have had former psychological treatment in either sessions. Four (31 %) fathers attended all ten session, two (15 %) fathers attended seven sessions, one (8 %) father attended six, two (15 %) fathers attended five, one (8 %) father attended one sessions, and three fathers (23 %) did not participate in any sessions (this was due to divorce (n=2) and living abroad (n=1)).

Figure 1

All participating adolescents and mothers (n=13, 100%) and 11 fathers (82 %) completed the baseline survey. The lower response rate for fathers was due to divorce (n=1) and living abroad (n=1). Post treatment 12 (92 %) of the adolescents, 13 (100 %) of the mothers and 10 (77 %) of the fathers completed the survey. At 3-month follow-up and 1-year follow-up the response rates declined for adolescents (n=9, 69%; n=8, 62%), mothers (n=12, 92%), and fathers (n=8, 62%; n=7, 54%).

Twelve adolescents (92 %), thirteen mothers (100 %) and ten fathers (77 %), completed the ESQ on treatment satisfaction. The responses from adolescents and parents on each item appear from table 2. Both adolescents and parents were satisfied with the treatment and scored “true” or “partly true” to most of the positive statements about treatment. All participants reported the program as helpful and would recommend it to others

Both adolescents and parents were able to give qualitative feedback and to comment freely on the satisfaction with the treatment and give suggestions for enhancing the program. Two adolescents (15%), twelve mothers (92%) and four fathers (36%) did give qualitative feedback. The majority of the feedback was positive. Answering the question about positive elements of the therapy, one parent wrote: “Everything… the psychoeducation… that it was group therapy… the psychologists… the openness, support and engagement of the parental group, the youth group, the method…”.

The participants’ response to the open-ended questions regarding feedback about the social program were grouped within the following themes.

Eight participants commented on the usefulness of the therapeutic techniques. One adolescent wrote: “I will always be able to use the tools I have gotten ”. The parents (n=7) also commented positively on the techniques, e.g.: “We got some very concrete tools which helped us a lot. We got help to “push” x in some situations that were difficult for her… it made a huge difference for x!”.

Several parents (n=6) found the introduction to how parents can influence on their adolescents anxiety and the support to the parents from the therapist concerning making demands to their child very helpful. One parent wrote “Got support to the things which are hard to implement/endure as a parent”. As a positive statement, one parent further emphasized: “To get tools to address anxiety. Give words to it [the anxiety] by means of RT [realistic thinking]. Good forms. Get focus on own reaction patterns/parent traps”.

Many of the parents (n=9) commented positively on the parental aspect of the group format, e.g.: “It was really great to meet others with the same difficulties/challenges”, and “Totally amazing with the group treatment and that we as parents had each other to spar with”.

Some of the parents (n=3) thought that the size of the group was too large when everyone were gathered in the same room. As one of them put it: “Too many gathered in the last part of the sessions, when the psychologist, the adolescents and the parents round off the session.”

Another theme from the participants’ feedback was the intensity of the intervention program. Some of the parents (n=2) requested more time during the session to “go in depth with what we have read from session to session”. One parent commented on the initial assessment interview. “The first interview was difficult for our daughter. She was adversely affected several weeks afterwards and had a difficult time attending school. Which we haven’t seen for a long time. Maybe it was meeting with two strangers immediately asking questions – intrusive – concerning the most difficult things in the world”.

One adolescent commented on the number of experiments in the program, she wrote: “…too many experiments a week”. Furthermore, one of the parents recommended conducting an individual session during the program to facilitate idiosyncratic adjustments.

All comments (n=5) regarding the therapists from both adolescents, mothers and fathers were positive, and reflected great satisfaction with the therapists: “It has been great for x to meet some experts on the field and to be understood and see that we as parents now understand her.” One adolescent wrote: “The only thing I didn’t like was that I was the only one who didn’t get a permanent co-therapist assigned, I got the leader of it all, or whatever you call her.”

Many of the parents (n=7) viewed the inclusion of both adolescents and their parents as a positive feature of the program. When asked about what worked well in the program, a parent replied: “That we have got a common understanding for our daughters difficulties and common methods” Some parents (n=3) requested even more time for psychoeducating parents, one parent wrote: “There was too little time to acquaint the parents with their role. Often it feels like “a waste of time” to sit and wait for the psychologist to come in to the parents. When the psychologist had time with the parents it all went a bit too fast”. When asked about suggestions for enhancing the program, another parent emphasized: “Get more education regarding anxiety and more time to education”. Other parents were very satisfied with the level of parental support, one of whom states: “A good thing that I, as a parent, got knowledge about how my own reactions affected my child, and that we got more suitable terms of the way we did”.

Table 3 gives an overview of the descriptive statistics as mean scores, standard deviations and with-in group effect sizes for the outcome measures.

The clinical severity rating (CSR) of the Anxiety Disorder Interview Schedule for DSM-IV, Child and Parent Version (ADIS-IV C/ P) were reduced with a large effect-size (d=1.02). The reduction improved further at 3-month follow-up equivalent to a medium effect size (d=0.58). However no adolescents were free of their SAD diagnosis at post and only two (15 %) were free of their SAD diagnosis at 3-month follow-up.

As presented in table 4 adolescents as well as their parents reported a significant reduction in outcomes related to anxiety symptoms and interference. The reduction were equivalent to medium to large effect-sizes (see table 3). The reduction remained at 1-year follow-up.

Furthermore, the outcomes related to specific mechanism of SAD as safety behavior, self-focused attention and post-event processing showed significant reduction from baseline to post treatment equivalent to large effect-sizes. Generally the change remained at 1-year follow-up.

The self-efficacy outcomes improved significantly with a large effect-size, which remained at 1-year follow-up.

The adolescents reported a significant reduction in depression symptoms at post treatment which remained at 1-year follow-up. No significant change in their children’s depression symptoms was reported by their parents. Parents reported no significant change in their own stress, depression or anxiety levels.