Social Anxiety

Available online at www.sciencedirect.com

ScienceDirect Behavior Therapy 52 (2021) 465–477

Cognitive Behavioral Therapy for Social Anxiety Disorder: Predictors of Treatment Outcome in a Quasi-Naturalistic Setting

Rachel M. Butler Emily B. O’Day Michaela B. Swee Arielle Horenstein

Richard G. Heimberg Temple University

www.elsevier.com/locate/bt

We examined the outcomes of individual cognitive behav- ioral therapy (CBT) for social anxiety disorder (SAD) in a sample of 93 adults seeking treatment in a university outpatient clinic specializing in CBT for SAD. Treatment followed the structure of a manual, but number of sessions varied according to client needs. After approximately 20 weeks of therapy, patients’ social anxiety had decreased and their quality of life had increased. Patients with more severe SAD or comorbid major depressive disorder (MDD) at pretreatment demonstrated higher levels of social anxiety averaged across pre- and posttreatment. However, clinician- rated severity of SAD, comorbid MDD, or comorbid generalized anxiety disorder did not predict treatment outcome. Higher pretreatment scores on measures of safety behaviors and cognitive distortions were associated with higher social anxiety averaged across pre- and posttreatment and predicted greater decreases from pre- to posttreatment on multiple social anxiety outcome measures. We found no predictors of change in quality of life. Those with high levels of safety behaviors and distorted cognitions may benefit more from CBT, perhaps due to its emphasis on targeting avoidance through exposure and changing distorted think- ing patterns through cognitive restructuring methods. Our study lends support to the body of research suggesting that manualized CBT interventions can be applied flexibly in

Funding Source: There was no external funding source for this study. Address correspondence to Richard G. Heimberg, Ph.D., Adult

Anxiety Clinic, Department of Psychology, Temple University, Weiss Hall, 1701 N. 13th St., Philadelphia, PA 19122-6085; e-mail: [email protected].

0005-7894/© 2021 Association for Behavioral and Cognitive Therapies. Published by Elsevier Ltd. All rights reserved.

clinical settings with promising outcomes for patients over a relatively short course of therapy.

Keywords: cognitive behavioral therapy; social anxiety disorder; avoidance; cognitive distortions

SOCIAL ANXIETY DISORDER (SAD), a psychological disorder characterized by fear of negative evalua- tion and avoidance of a variety of social situations, causes marked distress or impairment (American Psychiatric Association [APA], 2013). Individuals with SAD hold negative beliefs about themselves and others’ perceptions of them and frequently try to escape or avoid situations in which they might be viewed negatively (Clark & Wells, 1995; Rapee & Heimberg, 1997). Due to these fears and patterns of avoidance, individuals with SAD can experience elevated rates of social rejection and isolation, which then perpetuate their social fears (Rapee & Heimberg, 1997). SAD is a highly prevalent anxiety disorder; approximately 7% of individuals in the United States meet diagnostic criteria for SAD in any 12-month period (Kessler et al., 2005). In addition to its high prevalence, SAD is a particu- larly important target of psychological interven- tion, as it is imperative for individuals to be able to interact and connect with their social world. A strong body of literature supports cognitive

behavioral therapy (CBT) as an efficacious and effective nonpharmacological treatment for SAD (see meta-analyses by Acarturk et al., 2009; Barkowski et al., 2016; Mayo-Wilson et al., 2014; Powers et al., 2008). Research has shown that CBT for SAD leads to both immediate and long-lasting

466 bu t le r et a l.

reductions in social anxiety both in short-term and long-term follow-ups, underscoring that its effects are maintained across time (for a narrative systematic review of CBT for SAD, see Kaplan, Swee, & Heimberg, 2018). One approach to individual CBT for SAD is a manualized, workbook-driven treatment protocol, Managing Social Anxiety, originally devel- oped by Hope, Heimberg, Juster, and Turk (2000). Within this treatment protocol, clients complete readings and worksheets that correspond with topics covered in each session (e.g., psychoeducation, cognitive restructuring, and in-session and in vivo exposure). Ledley et al. (2009) evaluated this treat- ment protocol in a randomized controlled trial (RCT) for SAD; those in the immediate treatment condition experienced substantial improvement, including de- creases in social anxiety and social-anxiety-related impairment compared to the waitlist condition. Similarly, a second RCT examining Managing Social Anxiety demonstrated that social anxiety decreased across treatment and these gains were maintained at a 1-year follow-up (Goldin et al., 2012). Thus, RCTs support the efficacy of this specific manualized individual treatment of SAD compared to a waitlist condition. Despite the breadth of controlled research support-

ing the efficacy of CBT for SAD, there are relatively few reports ofnaturalisticor quasi-naturalistic trials of CBT for SAD. Beyond efficacy, it is also important to consider how effective a treatment is when imple- mented in less controlled clinical settings. Clinical settings, by nature, introduce variability that can affect outcomes, including factors related to the patient (e.g., comorbidity profile, symptom severity, complexity of psychosocial stressors, engagement in treatment), clinician (e.g., level of training, experience, or supervision), and treatment (e.g., use of a manual). It hasbeenargued thatindividualsinnaturalistic settings are much more heterogeneous than those included in RCTs and that RCTs are not representative of treatment-seeking individuals in an average clinical setting (Jacobson & Christensen, 1996). Furthermore, although CBT may be manualized, the CBT protocol may be less closely followed, monitored, or supervised in clinical settings (Stewart & Chambless, 2009). Despite these concerns, some research demonstrates that group CBT for SAD is equally effective in a community setting as in a controlled research setting (McEvoy, Nathan, Rapee, & Campbell, 2012). However, research on CBT for SAD in community settings remains limited and it is critical to further evaluate whether a more flexible CBT approach utilized in such settings is still effective. In general, CBT has been found to be effective for

the treatment of anxiety disorders. A meta-analysis of 56 studies supports CBT’s effectiveness in treating

anxiety disorders across clinical settings, including outpatient clinics, anxiety specialty clinics, private practices, and primary care, concluding that individ- uals’ levels of anxiety and depression improved significantly over the course of treatment (Stewart & Chambless, 2009). Furthermore, studies that included a waitlist or treatment-as-usual condition found that using CBT led to greater improvements compared to controls. Post hoc analyses suggested that effect sizes increased when individuals were not randomized and when medications were permitted and decreased when therapists were not trained, did not use manuals, and were not monitored. In a relatively severe and comorbid sample at a university-based outpatient clinic, Asnaani, Benhamou, Kaczkurkin, Turk-Karan, and Foa (2020) demonstrated that patients experi- enced clinically significant and reliable improvement across all measures of anxiety symptom severity (including social anxiety), depression, quality of life, and overall functioning over the course of treatment. Other studies have examined CBT for SAD in non-

RCT settings and found it to demonstrate effect sizes comparable to those found in RCTs (Gaston, Abbott, Rapee, & Neary, 2006; McEvoy, 2007; Mörtberg, Berglund, & Sundin, 2005). A study of individual CBT for social anxiety in outpatient clinics found reductions in social anxiety, depression, and general anxiety following treatment (Lincoln et al., 2003). Additionally, McEvoy et al. (2012) compared group CBT for SAD in a research setting to group CBT in a community setting, and found that group CBT in both settings produced substantial and equivalent decreases in social anxiety, despite more severe symptoms and life interference at pretreatment, greater likelihood of comorbid disorders, and lower educational attain- ment among patients in the community setting. Based on these findings, it is reasonable to conclude that manualized CBT for SAD can be applied effectively in a number of clinical settings. To date, there has yet to be an effectiveness study of the Managing Social Anxiety protocol. It also remains somewhat unknown who will

improve during the course of treatment or what predicts treatment outcome in clinical, non-RCT settings. A few predictors of treatment outcome have been identified among those who have participated in group CBT for SAD, including severity of social anxiety symptoms at pretreatment (Otto et al., 2000), pretreatment level of depression (Chambless, Tran, & Glass, 1997), homework compliance (Edelman & Chambless, 1995; Leung & Heimberg, 1996), num- ber of negative thoughts experienced during social interaction, and treatment expectancy (Chambless et al., 1997; Safren, Heimberg, & Juster, 1997). Research has also indicated that individuals with a comorbid mood disorder evidenced greater

Table 1 Demographic Characteristics and Descriptive Statistics for Outcome Variables (N = 93)

Mean or n SD or %

Age (M, SD) 28.13 9.99 Sex (n, %)

Female 37 39.8 Male 56 60.2

Race (n, %) Black 6 6.5 Asian/Pacific Islander 6 6.5 White 72 77.4 Other 6 6.5 Missing or not reported 3 3.2

Ethnicity (n, %) Hispanic 10 10.8 Non-Hispanic 80 86.0 Missing or not reported 3 3.2

Years of education (M, SD) 14.60 2.4 Marital status (n, %)

Single (never married) 80 86.0 Married/living with partner 12 12.9 Divorced/separated 1 1.1

Comorbid Generalized Anxiety Disorder (n, %) 29 30.1 Comorbid Major Depressive Disorder (n, %) 29 30.1 Average number of additional diagnoses (M, SD) 1.72 1.49

467c o gn i t i v e be ha v i o ra l t h e ra p y f or s ad

impairment before and after treatment than individ- uals with a comorbid anxiety disorder or no comorbid disorders, but still benefited from treatment (Erwin, Heimberg, Juster, & Mindlin, 2002). More research is needed to determine whether having a common comorbidity with SAD, such as major depressive disorder (MDD) or other anxiety disorders, may affect the course or outcome of treatment (Brown & Barlow, 1992; Koyuncu, İnce, Ertekin, & Tükel, 2019). It is also worth considering whether other clinical charac- teristics, such as an individual’s degree of behavioral avoidance or severity of cognitive distortions when entering treatment, affect treatment outcome, as these are the primary targets of CBT interventions. A recent study found that public speaking avoidance moderat- ed treatment outcome, suggesting that those who engage in more avoidance may experience added benefit from CBT, which targets these avoidance strategies (Mesri et al., 2017). Overall, It is important to identify predictors and moderators of treatment outcome in order to better understand factors that enhance or hinder CBT’s efficacy across clinical settings and provide more justification for using evidence-based treatment outside of RCTs.

the present study In the current study, we examine the effectiveness of individual CBT for social anxiety disorder using the Managing Social Anxiety protocol in a quasi- naturalistic clinical setting and identify predictors of treatment outcomes. We add to the growing

literature of non-RCT outcome studies of CBT for anxiety-related disorders by examining CBT for SAD in an outpatient anxiety clinic, using a variety of outcome measures, and assessing a number of clinical characteristics that may affect outcomes. We predicted that (a) social anxiety would decrease and quality of life would increase from pre- to posttreatment, (b) cognitive distortions and avoid- ance would moderate the effect of treatment such that those with higher initial levels would improve to a greater extent over the course of treatment, and (c) clinician-rated severity of SAD and existence of comorbid diagnoses would moderate the effect of treatment such that those with more severe SAD and those with comorbid generalized anxiety disorder (GAD) or MDD would benefit to a lesser degree.

Method participants

Participants (N = 93) were adults seeking treatment at the Adult Anxiety Clinic of Temple who met criteria for a principal diagnosis of SAD based on either the Anxiety Disorders Interview Schedule for the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV; APA, 1994)– Lifetime Version (ADIS-IV-L; Di Nardo, Brown, & Barlow, 1994) or the ADIS-5-L (T. Brown & Barlow, 2014) based on DSM-5 (APA, 2013). Individuals were included in the current study if they attended at least one session of treatment for

468 bu t le r et a l.

SAD in the clinic. Patients (Mage = 28.13) were majority male (60.2%) and White (77.4%). See Table 1 for further description of patients’ demo- graphic characteristics. All patients provided in- formed consent, and the study was approved by the Institutional Review Board at Temple University.

procedure

Individuals seeking treatment were recruited through the clinic website, clinical referrals, and community listings. As part of routine procedure, individuals first completed a telephone screening for eligibility prior to the in-person semistructured diagnostic interview. Interviews were completed by clinical psychology doctoral students. Individuals assessed between 2011–2014 (n = 48)

were eligible for treatment if they met DSM-IV criteria for a principal diagnosis of generalized SAD based on the ADIS-IV-L. To meet criteria for generalized SAD, individuals had to endorse greater than moderate fear in five or more distinct social situations. Individuals assessed after 2014 (n = 45) were eligible for treatment if they met DSM-5 criteria for SAD based on the ADIS-5-L. Both the ADIS-IV-L and ADIS-5-L are semistructured clinician-administered diagnostic in- terviews that assess for various anxiety disorders, mood disorders, obsessive-compulsive and related disorders, trauma disorders, somatoform disorders, substance use disorders, and also include a medical and psychiatric treatment history. The most promi- nent change in the diagnosis of SAD in DSM-5 was describing the main criterion as a fear of negative evaluation by others, which is a broader description of fears experienced by individuals with SAD than was the case in DSM-IV (focus on humiliation and embarrassment). Additionally, DSM-5 removed the “generalized” specifier and added the “performance only” specifier (Heimberg et al., 2014). These changes resulted in few corresponding changes to the ADIS interview, and we do not believe they significantly altered the patient population eligible to be seen at the clinic. Clinicians assign severity ratings for each diagnosis on a scale from 0 to 8. The ADIS-IV-L has demonstrated good interrater reliability (T. Brown, Di Nardo, Lehman, & Campbell, 2001). Doctoral students in clinical psychology reviewed 20% of the interviews to assess interrater reliability. There was 100% agreement with the principal diagnosis, and severity ratings assigned by reliability assessors were all within one point of ratings by the primary clinician. After completion of each diagnostic interview,

clinicians met together with the clinic director (RGH) to discuss assigned diagnoses and to determine whether or not the individual was appropriate for treatment. If ineligible, appropriate resources and referrals were offered. If eligible, patients consented to

participate in ongoing research assessments through- out their treatment. Patients then completed pretreat- ment self-report questionnaires using a secure online platform before they initiated treatment. Measures were collected again after approximately 20 sessions or at termination if it occurred earlier than 20 sessions. In the context of the current study, the assessment conducted after approximately 20 sessions of therapy (M = 20.59, SD = 2.69, range from 11 to 28) was considered “posttreatment.” However, some patients continued treatment in the clinic following this assessment based on individual needs.

treatment

Individual CBT was administered using Managing Social Anxiety: A Cognitive Behavioral Therapy Approach, a manualized protocol that includes a therapist guide (Hope, Heimberg, & Turk, 2006) and a client workbook (Hope et al., 2000), which was updated to a second edition in 2010 (Hope, Heimberg, & Turk, 2010a, 2010b). This treatment program begins with psychoeducation about social anxiety before moving to a focus on cognitive restructuring and exposure. Exposure exercises were often completed with the therapist or a role- player (i.e., graduate students, undergraduate research assistants, the clinic director). The manual is based on a 16-session protocol; however, it was applied flexibly in the clinic in that clinicians were able to use clinical judgment to determine if more time was needed on a given topic. At times, therapists focused on pressing concerns or comor- bid conditions outside of SAD (e.g., depression, general anxiety). Additionally, the course of ther- apy was allowed to continue past 16 sessions to allow for continued work on social anxiety or comorbid conditions. Therapists were clinical psychology doctoral students and participated in weekly group supervision with RGH.

measures Measures of Treatment Outcome We included three measures of social anxiety in the current study. First, the Liebowitz Social Anxiety Scale-Clinician Administered Version (LSAS; Heimberg et al., 1999; Liebowitz, 1987) is a widely used 24-item measure of social anxiety. The LSAS includes 11 items that assess reactions to social situations (i.e., going to a party, making conversation) and 13 items that assess reactions to performance situations (i.e., acting, performing, or giving a talk in front of an audience). Fear and avoidance of social situations during the past week are rated for each of the situations, using a 4-point scale ranging from 0 (none/never) to 3 (severe/usually). The ratings are summed to compute a total score, which ranges from 0 to 144. In our study, the LSAS was administered by

469c o gn i t i v e be ha v i o ra l t h e ra p y f or s ad

the ADIS interviewer. The LSAS has demonstrated strong psychometric properties (Fresco et al., 2001; Heimberg et al., 1999). In our sample, the LSAS demonstrated excellent internal consistency at pre- treatment (α = .93) and posttreatment (α = .96). The Social Interaction Anxiety Scale-

Straightforwardly Worded Items (SIAS-S; Rodebaugh, Woods, & Heimberg, 2007) includes the 17 straightforwardly worded items from the original 20-item Social Interaction Anxiety Scale (SIAS; Mattick & Clarke, 1998), which assesses fears of social interactions in dyads and groups (i.e., “I have difficulty making eye-contact with others,” “When mixing socially, I am uncomfortable”). The instrument uses a 5-point Likert-type scale rating ranging from 0 (not at all characteristic of me) to 4 (extremely characteristic of me). The ratings of the straightforwardly worded items are summed to compute a total score, which ranges from 0 to 68. Research has shown that when only the straight- forwardly worded items are considered, it is a significantly stronger predictor of criterion mea- sures than the full scale across both undergraduate and clinical samples (Rodebaugh et al., 2007; Rodebaugh et al., 2011). The authors suggested using only the 17 straightforward items to calculate the total score, creating the 17-item SIAS-S, which we have done here. The SIAS-S has demonstrated excellent internal consistency and strong construct validity (Rodebaugh et al., 2007, 2011). In our sample, the SIAS-S demonstrated excellent internal consistency at pretreatment (α = .89) and posttreat- ment (α = .95). The Social Phobia Scale (SPS; Mattick &

Clarke, 1998) is a 20-item self-report measure that assesses an individual’s level of anxiety in situations in which they may be observed by others (i.e., public speaking, eating in public, using a public restroom). Items are assessed on a 4-point Likert scale, from 0 (not at all characteristic or true of me) to 4 (extremely characteristic or true of me). The ratings are summed to compute a total score, which ranges from 0 to 80. The SPS was created as a companion scale to the SIAS and has demonstrated strong psychometric properties (E. Brown et al., 1997; Mattick & Clarke, 1998). In our sample, the SPS demonstrated excellent internal consistency at pretreatment (α = .90) and posttreat- ment (α = .93). The Quality of Life Inventory (QOLI; Frisch, 1994)

is a questionnaire that assesses well-being and satisfaction with life in 16 domains of life, including love, work, and recreation. Respondents rate how important each domain is to their overall happiness and satisfaction as 0 (not at all important), 1 (important), or 2 (very important), followed by a rating of how satisfied they are in the area from -3

(very dissatisfied) to 3 (very satisfied). Importance and satisfaction ratings for each item are multiplied to form weighted satisfaction ratings ranging from -6 to 6. The items with importance ratings greater than 0 are averaged for a single total score, with higher scores denoting greater quality of life. The QOLI has shown validity in assessment and treatment across settings (Frisch et al., 2003; Frisch et al., 2005). In our sample, the QOLI demonstrated good internal consistency at pretreatment (α = .76) and posttreatment (α = .85).

Hypothesized Treatment Predictors The Subtle Avoidance Frequency Examination (SAFE; Cuming et al., 2009) is a self-report question- naire assessing an individual’s use of subtle avoidance and safety behaviors to alleviate distress in social situations. It includes 32 items that rate how frequently individuals use a given avoidance strategy or safety behavior commonly associated with social anxiety (e.g., avoiding eye contact, rehearsing sen- tences in one’s mind), on a scale from 0 (never) to 4 (always). Total scores range from 0 to 128, with higher scores suggesting greater use of subtle avoid- ance strategies. The SAFE has demonstrated strong internal consistency, good construct validity, and is able to discriminate between nonclinical and clinical populations (Cuming et al., 2009). In our sample, the SAFE demonstrated excellent internal consistency at pretreatment (α = .93). The Cognitive Distortions Questionnaire (CD-

Quest; de Oliveira, 2015) is a clinical tool used to help clinicians and patients track common cognitive distortions that occur in day-to-day life. It assesses the intensity and frequency of 15 cognitive distortions (e.g., dichotomous thinking, fortune telling, labeling, mind reading). For each cognitive distortion, intensity was rated “A little (Up to 30%),” “Much (31–70%),” or “Very much (More than 70%),” and frequency was rated “No (It did not occur),” “Occasional (1–2 days during the past week),” “Much of the time (3–5 days during the past week),” or “Almost all the time (6–7 days during the past week).” Frequency, intensity, and total scores can be generated from responses. In the current study, we used the frequency and intensity scales only. The CD-Quest has demon- strated good psychometric properties in undergradu- ate (de Oliveira et al., 2015; Morrison et al., 2015) and clinical samples (Kaplan et al., 2017). The CD-Quest frequency subscale demonstrated excellent internal consistency at pretreatment (α = .90) as did the intensity subscale (α = .91).

data analytic plan

Item-level missing data were handled by person-mean insertion. If more than 10% of item-level data were missing for a given individual’s response on a measure, we did not compute a total score. Linear mixed-effects

470 bu t le r et a l.

models (LMMs) were used to test changes in outcome measures from pre- to posttreatment as well as moderators of changes across treatment. Random intercepts were included in the models, and maximum likelihood estimation was used to address missing data, so cases with missing data were included. The random effect covariance structure was declared using a scaled identity matrix. Within-group effect sizes were calculated as Cohen’s d (Cohen, 1988), using the differences in estimated marginal means divided by the pooled within-group standard deviation (Dunlap, Cortina, Vaslow, & Burke, 1996). We controlled for number of sessions completed by posttreatment assessment in all analyses.

Results demographic and descriptive infor- mation

Descriptive statistics for the outcome variables at pre- and posttreatment and for the predictor variables at pretreatment are presented in Table 2.

treatment outcomes

To examine treatment outcomes, we ran separate LMMs for each outcome variable with Time as the predictor variable. Scores on the LSAS significantly decreased from pretreatment to posttreatment, B = -17.27, SE = 3.57, p b .001, d = 0.63. SPS scores significantly decreased from pretreatment to posttreat-

Table 2 Descriptive Statistics for Predictor and Outcome Variables

Mean SD

LSAS Pretreatment 75.07 23.16 Posttreatment 57.86 27.14

SPS Pretreatment 33.41 14.48 Posttreatment 20.76 14.03

SIAS-S Pretreatment 44.01 11.38 Posttreatment 30.78 13.98

QOLI Pretreatment -0.20 1.62 Posttreatment 0.61 1.82

CD-Quest Frequency Pretreatment 16.74 10.39

CD-Quest Intensity Pretreatment 18.39 11.19

SAFE Pretreatment 61.10 21.43

SAD Clinician Severity Rating Pretreatment 5.96 0.82

Note. CD-Quest = Cognitive Distortions Questionnaire; LSAS = Liebowitz Social Anxiety Scale; SAFE = Subtle Avoidance Frequency Examination; SPS = Social Phobia Scale; SIAS-S = Social Interaction Anxiety Scale – Straightforwardly worded items; QOLI = Quality of Life Inventory.

ment, B = -12.55, SE = 1.92, p b .001, d = 0.86. SIAS-S scores significantly decreased from pretreatment to posttreatment, B = -14.02, SE = 1.90, p b .001, d = 0.98. Scores on the QOLI significantly increased from pretreatment to posttreatment, B = 0.63, SE = 0.17, p = .001, d = 0.50.

moderators of treatment outcome ADIS Clinician Severity Rating for SAD We tested whether clinician-rated severity of SAD predicted changes in social anxiety and quality of life from pre- to posttreatment using LMMs as described above. There was a main effect of Severity such that higher severity ratings at pretreatment predicted higher LSAS scores aver- aged across pre- and posttreatment, B = 14.32, SE = 3.51, p = .001, d = 0.54. There was no Severity x Time interaction, B = 2.21, SE = 4.36, p = .61. There was a main effect of Severity on SPS scores

such that higher severity ratings at pretreatment predicted higher SPS scores averaged across pre- and posttreatment, B = 3.83, SE = 2.03, p = .003, d = 0.26. There was no Severity x Time interaction, B = 2.91, SE = 2.24, p = .20. There was a main effect of Severity on SIAS-S

scores such that higher severity ratings at pretreat- ment were associated with higher SIAS-S scores averaged across pre- and posttreatment, B = 4.75, SE = 1.80, p = .01, d = 0.37. There was no Severity x Time interaction, B = -0.08, SE = 2.15, p = .97. There was a main effect of Severity on QOLI

scores such that higher severity ratings at pretreat- ment were associated with lower QOLI scores averaged across pre- and posttreatment, B = .59, SE = 0.23, p = .01, d = 0.33. There was no Severity x Time interaction, B = 0.26, SE = 0.19, p = .17.

Comorbid Diagnoses We examined the rates of ADIS-diagnosed GAD and MDD in the sample. Twenty-nine patients had a diagnosis of GAD, and 29 patients had a diagnosis of MDD (11 had both diagnoses and are included in both groups). For the current analyses, we compared those with a given comor- bidity to those without that specific comorbidity. Further information is presented in Table 1.

MDD We examined the effects of a comorbid diagnosis of MDD1 on LSAS scores. There was a main effect of

1 We examined the effects of a broader comorbid mood disorder diagnosis on outcome variables and found an identical pattern of results as for MDD. Then, we examined the moderating effects of Beck Depression Inventory-II scores on outcome variables, and we found the same pattern of results as for MDD. Thus, we only report results for a comorbid MDD diagnosis here. Reports of these analyses are available upon request.

FIGURE 1 Trajectory of change in social anxiety (SIAS-S) from pre- to posttreatment for those with low and high levels of safety behaviors (SAFE) at pretreatment.

471c o gn i t i v e be ha v i o ra l t h e ra p y f or s ad

MDD such that those with MDD had higher LSAS scores, B = -12.98, SE = 6.44, p = .047, d = 0.26. There was no MDD x Time interaction, B = -9.17, SE = 7.47, p = .23. We examined the effects of a comorbid diagnosis of

MDD on SPS scores. There was a main effect of MDD such that those with MDD had higher SPS scores, B = -9.36, SE = 3.70 p = .01, d = 0.32. There was no MDD x Time interaction, B = -2.07, SE = 4.19, p = .62. We examined the effects of a comorbid diagnosis

of MDD on SIAS-S scores. There was a main effect of MDD such that those with MDD had higher SIAS-S scores, B = -7.03, SE = 3.54, p = .049, d = 0.25. There was no MDD x Time interaction, B = -2.65, SE = 4.10, p = .52. We examined the effects of a comorbid diagnosis of

MDD on QOLI scores. There was no main effect of MDD on QOLI scores, B = .66, SE = 0.47, p = .16. There was no MDD x Time interaction, B = -.26, SE = 0.39, p = .51.

GAD We examined the effects of a comorbid diagnosis of GAD on LSAS scores. There was no main effect of GAD on LSAS scores, B = -10.98, SE = 7.14, p = .13. There was no GAD x Time interaction, B = 1.15, SE = 7.91, p = .89. We examined the effects of a comorbid diagnosis

of GAD on SPS scores. There was no main effect of GAD on SPS scores, B = -6.64, SE = 3.92, p = .09. There was no GAD x Time interaction, B = 1.66, SE = 4.22, p = .70. We examined the effects of a comorbid diagnosis of

GAD on SIAS-S scores. There was no main effect of GAD on SIAS-S scores, B = -4.89, SE = 3.70, p = .19. There was no GAD x Time interaction, B = 1.65, SE = 4.16, p = .69.

We examined the effects of a comorbid diagnosis of GAD on QOLI scores. Those with comorbid GAD had lower QOLI scores, B = -1.17, SE = 0.45, p = .01, d = 0.33. There was no GAD x Time interaction, B = -0.09, SE = .37, p = .81.

Safety Behaviors We examined the moderating effect of pretreatment SAFE scores on change in social anxiety and quality of life from pre- to posttreatment. There was a main effect of pretreatment SAFE scores on LSAS scores such that higher SAFE scores at pretreatment were associated with higher LSAS scores averaged across pre- and posttreatment, B = 0.36, SE = 0.14 p = .01, d = 0.34. There was no SAFE x Time interaction, B = 0.32, SE = 0.17, p = .06. There was a main effect of pretreatment SAFE

scores on SPS scores such that higher SAFE scores at pretreatment were associated with higher SPS scores averaged across pre- and posttreatment, B = 0.24, SE = 0.07, p = .001, d = 0.43. The SAFE x Time interaction fell short of statistical significance, B = 0.17, SE = 0.09, p = .06. There was a main effect of pretreatment SAFE

scores on SIAS-S scores such that higher SAFE scores at pretreatment were associated with higher SIAS-S scores averaged across pre- and posttreatment, B = 0.31, SE = 0.07, p b .001, d = 0.31. There was a SAFE x Time interaction, B = -0.18, SE = 0.09, p = .047, d = 0.27, such that those with higher SAFE scores at pretreatment experienced a greater decrease in SIAS-S scores from pre- to posttreatment. See Figure 1 for a depiction of the interaction effect. There was no main effect of pretreatment SAFE

scores on QOLI scores, B = 0.002, SE = 0.01, p =.85. There was no SAFE x Time interaction, B = -0.010, SE = 0.008, p = .24.

FIGURE 2 Graphs of the interactions between cognitive distortion intensity (CD-Quest intensity) and time to predict social anxiety as measured by the LSAS, SPS, and SIAS-S.

472 bu t le r et a l.

Cognitive Distortions We examined the moderating effect of pretreatment CD-Quest intensity and frequency scores on social anxiety and quality of life. There was a main effect of pretreatment CD-Quest intensity scores on LSAS scores, B = 1.25, SE = 0.32, p b .001, d = 0.51, and a main effect of CD-Quest frequency scores on LSAS scores, B = 1.30, SE = 0.35, p b .001, d = 0.48. There was a CD-Quest intensity x Time interaction predicting LSAS scores, B = -.98, SE = 0.34, p =

FIGURE 3 Graphs of the interactions between frequency) and time to predict social anxiety as mea

.005, d = 0.38. The CD-Quest frequency x Time interaction was also significant, B = -.85, SE = 0.38, p = .028, d = 0.30. The interaction effects are depicted in Figures 2 and 3. There was a main effect of pretreatment CD-

Quest intensity scores on SPS scores, B = 0.63, SE = 0.16, p b .001, d = 0.52, and a main effect of CD- Quest frequency scores, B = 0.70, SE = 0.17, p b .001, d = 0.53. There was a CD-Quest intensity x Time interaction, B = -0.45, SE = 0.18, p = .02, d =

cognitive distortion frequency (CD-Quest sured by the LSAS and SPS.

473c o gn i t i v e be ha v i o ra l t h e ra p y f or s ad

0.34. There was a CD-Quest frequency x Time interaction, B = -0.46, SE = 0.20, p = .02, d = 0.32 (see Figures 2 and 3). There was a main effect of pretreatment CD-

Quest intensity scores on SIAS-S scores, B = 0.42, SE = 0.15, p = .005, d = 0.37. There was a main effect of CD-Quest frequency scores on SIAS-S scores, B = 0.48, SE = 0.16, p = .003, d = 0.39. There was a CD-Quest intensity x Time interaction, B = -0.37, SE = 0.18, p = .04, d = 0.28 (see Figure 2). There was no CD-Quest frequency x Time interac- tion, B = -0.35, SE = 0.20, p = .08. There was no main effect of pretreatment CD-

Quest intensity scores on QOLI scores, B = -0.03, SE = 0.02, p = .12. There was a main effect of CD- Quest frequency scores, B = -0.06, SE = 0.02, p = .003, d = 0.40. There was no CD-Quest intensity x Time interaction, B = -0.02, SE = 0.02, p = .21 and no CD-Quest frequency x Time interaction, B = -0.01, SE = 0.02, p = .52.

Discussion We examined the effects of individual CBT for SAD in an open trial of the flexible application of a manualized treatment for SAD. The current study fills a need for more research on the effectiveness of manualized treatment and meets several of the criteria for clinical representativeness laid out by Shadish, Navarro, Matt, and Phillips (2000), including clinically representative problems, a clinically representative setting, clinically represen- tative structure, therapy freedom, and flexible number of sessions. Our clinic meets these criteria by accepting clients from the university and broader community with social anxiety and comorbid disorders, being an outpatient clinic, and allowing clinicians the freedom to address concerns outside of social anxiety as well as extend treatment beyond protocol length. Our clinicians were clinical psy- chology doctoral students, who have been shown to produce treatment outcomes similar to licensed psychologists when supervised by a licensed clini- cian and whose efforts may therefore be generaliz- able to other clinical settings (Öst, Karlstedt, & Widén, 2012). Thus, our study adds to the literature on the effectiveness of manualized CBT treatments in naturalistic or quasi-naturalistic settings. Foremost, patients’ social anxiety reduced after

approximately 20 weeks of individual CBT, as evidenced by decreases in scores on the LSAS, SPS, and SIAS-S from pre- to posttreatment. At post- treatment, mean scores on the LSAS dropped below the recommended cutoff for a diagnosis of gener- alized SAD (LSAS b 60; Mennin et al., 2002). Scores on the SPS fell below the cutoff of 24 for a

diagnosis of SAD proposed by Heimberg, Mueller, Holt, Hope, and Liebowitz (1992), whereas the SIAS-S scores did not fall below the cutoff of 28 proposed by Rodebaugh et al. (2011). The effect sizes for pre- to posttreatment decreases in social anxiety (d’s = 0.63, 0.86, and 0.98, respectively) were in the same range as those found in an RCT of the MSA program (d = 1.05; Goldin et al., 2012). Overall, these findings are particularly encouraging given the flexible application of manualized CBT in the clinic. Patients in the current sample were diagnostically heterogeneous, as they tended to have more than one diagnosis in addition to SAD (M = 1.72, range from 0 to 9), and comorbid conditions varied. Nonetheless, patients were still able to attain a reduction in social anxiety over the course of treatment. Additionally, patients reported significant improve-

ments intheirquality oflifefrompre-toposttreatment. This finding aligns with prior research demonstrating that quality of life, particularly in areas such as achievement and social functioning, improves following group CBT (Eng, Coles, Heimberg, & Safren, 2005). Presumably, CBT reduces symptoms of social anxiety, and this mitigation of symptoms allows for improvement in important domains of life satisfaction such as work, relationships, friendships, and community. Perhaps through exposure and cognitive restructuring, individuals are able to reen- gage in aspects of life they had previously been avoiding (e.g., social gatherings, work responsibilities), thus garnering more satisfaction in those areas. In addition toprimary outcomes, wewere interested

in whether specific baseline factors would predict treatment outcome. Clinician-rated severity of SAD was associated with higher social anxiety and poorer quality of life averaged across pre- and posttreatment but was not predictive of change across treatment. Patients experienced comparable decreases in social anxiety and increases in quality of life, regardless of severity at baseline. This corroborates findings by McEvoy et al. (2012) demonstrating that those who engaged in group CBT at a community clinic had more severe SAD than those at the research clinic but made comparable improvements through treatment. This finding is encouraging, as it suggests that patients with more severe SAD at baseline are not precluded from making gains through CBT. We also examined whether having comorbid

MDD or GAD would affect the degree of change across treatment. Those with MDD had higher scores on measures of social anxiety, but did not differ in quality of life from those without MDD. Those with MDD had similar changes in social anxiety and quality of life from pre- to posttreat- ment as those without, suggesting that individuals

474 bu t le r et a l.

with comorbid MDD are still able to benefit from CBT for SAD. This aligns with past research by Erwin et al. (2002) showing that those with comorbid depression started and ended group CBT with higher levels of social anxiety than those without. Despite this, a comorbid diagnosis of MDD did not affect ability to improve during treatment, as has been demonstrated in prior research on group CBT for SAD (Marom, Gilboa-Schechtman, Aderka, Weiz- man, & Hermesh, 2009). Compared to those without GAD, patients with GAD had poorer quality of life but experienced comparable levels of social anxiety to those without comorbid GAD, which corroborates prior research (Erwin et al., 2002). Those with and without comorbid GAD experienced similar decreases in social anxiety through treatment. Overall, these findings suggest that common comorbid diagnoses such as MDD and GAD do not affect individuals’ ability to experience a reduction in social anxiety and an increase in quality of life through CBT. Higher scores on the measure of safety behaviors

at pretreatment were associated with higher levels of social anxiety averaged across pre- and post- treatment. Additionally, those who reported engag- ing in more safety behaviors prior to starting treatment experienced a greater decrease in social anxiety (i.e., SIAS-S) from pre- to posttreatment compared to those with lower levels of safety behaviors. Teaching those with SAD to identify and resist engaging in safety behaviors may be respon- sible for an important portion of the decrease in social anxiety during treatment (Piccirillo, Dryman, & Heimberg, 2016). CBT specifically targets avoidance by encouraging individuals to approach rather than avoid, so those engaging in more avoidance at the outset may have the most to gain from this type of therapy (Mesri et al., 2017). Whereas this interaction was only evident for the SIAS-S, it trended towards significance for the other two measures of social anxiety (p = .06). Of course, we are limited in that we cannot determine the directionality of these effects, as the SAFE was administered only at pretreatment and thus we cannot know whether decreases in safety behaviors lead to decreases in SA or vice versa. This preliminary evidence suggests that patients who utilize more subtle avoidance strategies before treatment can make considerable improvements and end treatment with levels of social anxiety comparable to those who are less initially avoidant. It will be important to investigate whether other types of avoidance, such as complete behavioral avoidance or avoidance of emotional experiences, have the same implications for treatment outcomes. As expected, higher frequency and intensity of

cognitive distortions at pretreatment were associated

with higher levels of social anxiety and poorer quality of life. Additionally, those reporting more frequent cognitive distortions prior to treatment experienced greater decreases in social anxiety (i.e., LSAS, SPS) from pre- to posttreatment. Similarly, those with higher intensity of cognitive distortions at pretreat- ment experienced greater decreases in social anxiety (i.e., LSAS, SPS, SIAS-S) from pre- to posttreatment. In other words, those who started treatment with higher intensity and frequency of cognitive distortions made more improvements in social anxiety through CBT than those who began with less intense or frequent cognitive distortions. CBT specifically targets cogni- tive distortions by teaching individuals to identify thinking errors and to challenge thoughts rather than accepting them as facts. Changes in negative thinking patterns are believed to relate to changes in social anxiety symptoms during treatment (Chambless et al., 1997), but, as above, the direction of these changes cannot be determined based on the moderation analyses. Based on our findings, CBT may be a particularly useful approach for those who present for treatment of SAD with more extremely distorted thinking. A limitation of the current study is that the

sample was less diverse than might be found in a community outpatient clinic; the majority of patients in our sample were White (77.4%) and Non-Hispanic (86%), and, due to our setting within the university, more patients tended to be college students than would likely be the case in a community clinic. This affects our ability to draw sound conclusions about whether individual CBT for SAD in a quasi-naturalistic setting may benefit a more diverse population. In addition, we are only able to report information about patients’ sex rather than gender, which is an important limita- tion given the clear distinction between the two. Our sample size may have been smaller then ideally suited for our moderation analyses, and although we attempted to collect 3-month follow-up data from patients, we were unable to use those data due to excessive missingness. This and our lack of clinician competency and fidelity ratings limit the reader’s ability to interpret the generalizability of the current study’s findings. Although our research demonstrates that indi-

viduals make substantial improvements in just 20 weeks of flexible, manualized treatment for SAD, we do not know the relative efficacy of this approach compared to strict adherence to the protocol, as in an RCT. Future research should directly compare strict, protocol treatment to a more flexible approach that allows for work on other presenting problems within the context of the treatment. It will be critical for us to determine

475c o gn i t i v e be ha v i o ra l t h e ra p y f or s ad

whether manualized treatment for SAD is more effective when adhered to strictly, or whether allowing therapists to work on other presenting problems within the framework of the therapy results in better overall outcomes.

conclusions

Despite CBT’s strong evidence base, many thera- pists are still reluctant to utilize and implement exposure-based therapies, such as CBT for SAD, due to the concern that “real world” clients will not benefit from the treatments that are studied in RCTs (Asnaani et al., 2020). Continued efforts must be made to demonstrate CBT’s effectiveness for treating anxiety-related disorders in more-or- less naturalistic settings, to make evident that “real patients” can benefit from CBT. The current study adds to the growing evidence base suggesting that, in fact, CBT is applicable and viable for patients in less structured settings. Our findings show that clinician-rated severity of SAD and existence of comorbid MDD or GAD do not affect treatment outcome, but those with high levels of avoidance and distorted cognitions may benefit more from CBT than those with lower levels of avoidance and maladaptive thinking. Future research must con- tinue to investigate predictors and moderators of treatment outcome so that patients and clinicians can develop greater understanding of who benefits most from CBT and what factors may affect treatment gains.

Conflict of Interest Statement Richard G. Heimberg is a coauthor of the commercially available treatment manual used in this study. The authors have no other conflicts of interest to report.

References Acarturk, C., Cuijpers, P., van Straten, A., & de Graaf, R.

(2009). Psychological treatment of social anxiety disorder: a meta-analysis. Psychological medicine, 39, 241–254. https:// doi.org/10.1017/S0033291708003590

American Psychiatric Association (1994). Diagnostic and statistical manual of mental disorders(4th ed.; DSM-IV).Washington, DC: Author. https://doi.org/10.1007/springerreference_179660

American Psychiatric Association (2013). Diagnostic and statistical manual of mental disorders (5th ed.; DSM-5). Washington, DC: Author. https://doi.org/10.1176/appi. books.9780890425596.744053

Asnaani, A., Benhamou, K., Kaczkurkin, A. N., Turk-Karan, E., & Foa, E. B. (2020). Beyond the constraints of an RCT: Naturalistic treatment outcomes for anxiety-related disor- ders. Behavior Therapy, 51, 434–446. https://doi.org/10. 1016/j.beth.2019.07.007

Barkowski, S., Schwartze, D., Strauss, B., Burlingame, G. M., Barth, J., & Rosendahl, J. (2016). Efficacy of group psychotherapy for social anxiety disorder: A meta-analysis of randomized-controlled trials. Journal of Anxiety Disor- ders, 39, 44–64. https://doi.org/10.1016/j.janxdis.2016.02. 005 Chicago

Brown, T. A., & Barlow, D. H. (1992). Comorbidity among anxiety disorders: Implications for treatment and DSM-IV. Journal of Consulting and Clinical Psychology, 60, 835–844. https://doi.org/10.1037/0022-006X.60.6.835

Brown, T. A., & Barlow, D. H. (2014). Anxiety and Related Disorders Interview Schedule for DSM-5 (ADIS-5-L): Lifetime Version. Client Interview Schedule. New York, NY: Oxford University Press. https://doi.org/10.4135/ 9781483365817.n81

Brown, T. A., Di Nardo, P. A., Lehman, C. L., & Campbell, L. A. (2001). Reliability of DSM-IV anxiety and mood disorders: Implications for the classification of emotional disorders. Journal of Abnormal Psychology, 110, 49–58. https://doi.org/ 10.1037//0021-843x.110.1.49

Brown, E. J., Turovsky, J., Heimberg, R. G., Juster, H. R., Brown, T. A., & Barlow, D. H. (1997). Validation of the Social Interaction Anxiety Scale and the Social Phobia Scale across the anxiety disorders. Psychological Assessment, 9, 21–27. https://doi.org/10.1037/1040-3590.9.1.21

Chambless, D. L., Tran, G. Q., & Glass, C. R. (1997). Predictors of response to cognitive-behavioral group therapy for social phobia. Journal of Anxiety Disorders, 11, 221–240. https:// doi.org/10.1016/s0887-6185(97)00008-x

Clark, D. M., & Wells, A. (1995). A cognitive model of social phobia. In R. G. Heimberg M. R. Liebowitz D. A. Hope & F. R. Schneier (Eds.), Social phobia: Diagnosis, assessment, and treatment (pp. 69–93). New York, NY: Guilford Press.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences. Hillsdale, NJ: Lawrence Earlbaum Associates, Inc.

Cuming, S., Rapee, R. M., Kemp, N., Abbott, M. J., Peters, L., & Gaston, J. E. (2009). A self-report measure of subtle avoidance and safety behaviors relevant to social anxiety: Development and psychometric properties. Journal of Anxiety Disorders, 23, 879–883. https://doi.org/10.1016/j. janxdis.2009.05.002

de Oliveira, I. R. (2015). Introducing the Cognitive Distortions Questionnaire. In I. R. de Oliveira (Ed.), Trial-based cognitive therapy: A manual for clinicians (pp. 25–40). New York: Routledge.

de Oliveira, I. R., Seixas, C., Osório, F. L., Crippa, J. A. S., De Abreu, J. N., Menezes, I. G., … Wenzel, A. (2015). Evaluation of the psychometric properties of the Cognitive Distortions Questionnaire (CD-Quest) in a sample of undergraduate students. Innovations in Clinical Neurosci- ence, 12, 20–27.

Di Nardo, P. A., Brown, T. A., & Barlow, D. H. (1994). Anxiety Disorders Interview Schedule for DSM-IV: Lifetime version (ADIS-IV-L). New York, NY: Oxford University Press.

Dunlap, W. P., Cortina, J. M., Vaslow, J. B., & Burke, M. J. (1996). Meta-analysis of experiments with matched groups or repeated measures designs. Psychological Methods, 1, 170–177. https://doi.org/10.1037/1082-989X.1.2.170

Edelman, R. E., & Chambless, D. L. (1995). Adherence during sessions and homework in cognitive-behavioral group treatment of social phobia. Behaviour Research and Therapy, 33, 573–577. https://doi.org/10.1016/0005-7967 (94)00068-u

Eng, W., Coles, M. E., Heimberg, R. G., & Safren, S. A. (2005). Domains of life satisfaction in social anxiety disorder: Relation to symptoms and response to cognitive-behavioral therapy. Journal of Anxiety Disorders, 19, 143–156. https:// doi.org/10.1016/j.janxdis.2004.01.007

Erwin, B. A., Heimberg, R. G., Juster, H., & Mindlin, M. (2002). Comorbid anxiety and mood disorders among persons with social anxiety disorder. Behaviour Research and Therapy, 40, 19–35. https://doi.org/10.1016/s0005- 7967(00)00114-5

476 bu t le r et a l.

Fresco, D.M., Coles, M. E., Heimberg, R. G., Liebowitz, M. R., Hami, S., Stein, M. B., & Goetz, D. (2001). The Liebowitz Social Anxiety Scale: A comparison of the psychometric properties of self-report and clinician-administered formats. Psychological Medicine, 31, 1025–1035. https://doi.org/10. 1017/s0033291701004056

Frisch, M. B. (1994). Quality of Life Inventory (QOLI). Minneapolis, MN: National Computer Systems.

Frisch, M. B., Clark, M. P., Rouse, S. V., Rudd, M.D., Paweleck, J., & Greenstone, A. (2005). Predictive and treatment validity of life satisfaction and the Quality of Life Inventory. Assessment, 12, 66–78. https://doi.org/10.1177/ 1073191104268006

Frisch, M. B., Clark, M. P., Rouse, S. V., Rudd, M. D., Paweleck, J., Greenstone, A., & Kopplin, D. A. (2003). Predictive validity and sensitivity to change in quality of life assessment and life satisfaction: Further studies of the Quality of Life Inventory or QOLI in mental health settings. In M.J. Sirgy, D. Rahtz, & A.C. Samli (Eds.), Advances in Quality-of-Life Theory and Research (pp. 191–210). Dor- drecht: Springer Science+Business Media. https://doi.org/ 10.1007/978-94-017-0387-1_12

Gaston, J. E., Abbott, M. J., Rapee, R. M., & Neary, S. A. (2006). Do empirically supported treatments generalize to private practice? A benchmark study of a cognitive– behavioural group treatment programme for social phobia. British Journal of Clinical Psychology, 45, 33–48. https:// doi.org/10.1348/014466505X35146

Goldin, P. R., Ziv, M., Jazaieri, H., Werner, K., Kraemer, H., Heimberg, R. G., & Gross, J. J. (2012). Cognitive reappraisal self-efficacy mediates the effects of individual cognitive-behavioral therapy for social anxiety disorder. Journal of Consulting and Clinical Psychology, 80, 1034–1040. https://doi.org/10.1037/a0028555

Heimberg, R. G., Hofmann, S. G., Liebowitz, M. R., Schneier, F. R., Smits, J. A. J., Stein, M. B., Hinton, D. E., & Craske, M. G. (2014). Social anxiety disorder in DSM-5. Depression and Anxiety, 31, 472–479. https://doi.org/10.1002/da. 22231

Heimberg, R. G., Horner, K. J., Juster, H. R., Safren, S. A., Brown, E. J., Schneier, F. R., & Liebowitz, M. R. (1999). Psychometric properties of the Liebowitz Social Anxiety Scale. Psychological Medicine, 29, 199–212. https://doi.org/ 10.1017/s0033291798007879

Heimberg, R. G., Mueller, G. P., Holt, C. S., Hope, D. A., & Liebowitz, M. R. (1992). Assessment of anxiety in social interaction and being observed by others: The Social Interaction Anxiety Scale and the Social Phobia Scale. Behavior Therapy, 23, 53–73. https://doi.org/10.1016/ S0005-7894(05)80308-9

Hope, D. A., Heimberg, R. G., Juster, H. A., & Turk, C. L. (2000). Managing social anxiety: A cognitive-behavioral therapy approach (Client Workbook). New York: Oxford University Press.

Hope, D. A., Heimberg, R. G., & Turk, C. L. (2006). Therapist guide for Managing social anxiety: A cognitive-behavioral therapy approach. New York: Oxford University Press.

Hope, D. A., Heimberg, R. G., & Turk, C. L. (2010a). Mana- ging social anxiety: A cognitive-behavioral therapy ap- proach (Client Workbook, 2nd edition). New York: Oxford University Press. https://doi.org/10.1093/med:psych/ 9780195336696.001.0001

Hope, D. A., Heimberg, R. G., & Turk, C. L. (2010b). Therapist guide for Managing social anxiety: A cognitive- behavioral therapy approach (2nd ed.). New York: Oxford University Press. https://doi.org/10.1093/med:psych/ 9780195336689.001.0001

Jacobson, N. S., & Christensen, A. (1996). Studying the effectiveness of psychotherapy: How well can clinical trials do the job? American Psychologist, 51, 1031–1039. https://doi. org/10.1037/0003-066x.51.10.1031

Kaplan, S. C., Morrison, A. S., Goldin, P. R., Olino, T. M., Heimberg, R. G., & Gross, J. J. (2017). The Cognitive Distortions Questionnaire (CD-Quest): Validation in a sample of adults with social anxiety disorder. Cognitive Therapy and Research, 41, 576–587. https://doi.org/10. 1007/s10608-017-9838-9

Kaplan, S. C., Swee, M., & Heimberg, R. G. (2018). Psychological treatments for social anxiety disorder. In O. Braddick (Ed.), The Oxford Research Encyclopedia of Psychology New York: Oxford University Press. https://doi.org/10.1093/acrefore/ 9780190236557.013.98 Retrieved from https://psychology. oxfordre.com/view/10.1093/acrefore/9780190236557.001. 0001/acrefore-9780190236557-e-98

Kessler, R. C., Berglund, P., Demler, O., Jin, R., Merikangas, K. R., & Walters, E. E. (2005). Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. Archives of General Psychiatry, 62, 593–602. https://doi.org/10.1001/archpsyc.62.6.593

Koyuncu, A., İnce, E., Ertekin, E., & Tükel, R. (2019). Comorbidity in social anxiety disorder: Diagnostic and therapeutic challenges. Drugs in Context, 8, 212573. https:// doi.org/10.7573/dic.212573

Ledley, D. R., Heimberg, R. G., Hope, D. A., Hayes, S. A., Zaider, T. I., Van Dyke, M., … Fresco, D. M. (2009). Efficacy of a manualized and workbook-driven individual treatment for social anxiety disorder. Behavior Therapy, 40, 414–424. https://doi.org/10.1016/j.beth.2008.12.001

Leung, A. W., & Heimberg, R. G. (1996). Homework compliance, perceptions of control, and outcome of cognitive-behavioral treatment of social phobia. Behaviour Research and Therapy, 34, 423–432. https://doi.org/10. 1016/0005-7967(96)00014-9

Liebowitz, M. R. (1987). Social phobia. Modern Trends in Pharmacopsychiatry, 22, 141–173. https://doi.org/10.1159/ 000414022

Lincoln, T. M., Rief, W., Hahlweg, K., Frank, M., von Witzleben, I., Schroeder, B., & Fiegenbaum, W. (2003). Effectiveness of an empirically supported treatment for social phobia in the field. Behaviour Research and Therapy, 41, 1251–1269. https://doi.org/10.1016/S0005-7967(03) 00038-X

Marom, S., Gilboa-Schechtman, E., Aderka, I. M., Weizman, A., & Hermesh, H. (2009). Impact of depression on treatment effectiveness and gains maintenance in social phobia: A naturalistic study of cognitive behavior group therapy. Depres- sion and Anxiety, 26, 289–300. https://doi.org/10.1002/da. 20390

Mattick, R. P., & Clarke, J. C. (1998). Development and validation of measures of social phobia scrutiny fear and social interaction anxiety. Behaviour Research and Therapy, 36, 455–470. https://doi.org/10.1016/s0005-7967(97)10031-6

Mayo-Wilson, E., Dias, S., Mavranezouli, I., Kew, K., Clark, D. M., Ades, A. E., & Pilling, S. (2014). Psychological and pharmacological interventions for social anxiety disorder in adults: A systematic review and network meta-analysis. The Lancet Psychiatry, 1(5), 368–376. https://doi.org/10.1016/ S2215-0366(14)70329-3

McEvoy, P. M. (2007). Effectiveness of cognitive behavioural group therapy for social phobia in a community clinic: A benchmarking study. Behaviour Research and Therapy, 45, 3030–3040. https://doi.org/10.1016/j.brat.2007.08.002

McEvoy, P. M., Nathan, P., Rapee, R. M., & Campbell, B. N. (2012). Cognitive behavioural group therapy for social

477c o gn i t i v e be ha v i o ra l t h e ra p y f or s ad

phobia: Evidence of transportability to community clinics. Behaviour Research and Therapy, 50, 258–265. https://doi. org/10.1016/j.brat.2012.01.009

Mennin, D. S., Fresco, D. M., Heimberg, R. G., Schneier, F. R., Davies, S. O., & Liebowitz, M. R. (2002). Screening for social anxiety disorder in the clinical setting: using the Liebowitz Social Anxiety Scale. Journal of Anxiety Disorders, 16, 661–673. https://doi.org/10.1016/S0887-6185(02)00134-2

Mesri, B., Niles, A. N., Pittig, A., LeBeau, R. T., Haik, E., & Craske, M. G. (2017). Public speaking avoidance as a treatment moderator for social anxiety disorder. Journal of Behavior Therapy and Experimental Psychiatry, 55, 66–72. https://doi.org/10.1016/j.jbtep.2016.11.010

Morrison, A. S., Potter, C. M., Carper, M. M., Kinner, D. G., Jensen, D., Bruce, L., … Heimberg, R. G. (2015). The Cognitive Distortions Questionnaire (CD-Quest): Psycho- metric properties and exploratory factor analysis. Interna- tional Journal of Cognitive Therapy, 8, 287–305. https:// doi.org/10.1521/ijct.2015.8.4.287

Mörtberg, E., Berglund, G., & Sundin, Ö. (2005). Intensive cognitive behavioural group treatment of social phobia: A pilot study. Cognitive Behaviour Therapy, 34, 41–49. https://doi.org/10.1080/16506070510010657

Öst, L. G., Karlstedt, A., & Widén, S. (2012). The effects of cognitive behavior therapy delivered by students in a psychologist training program: An effectiveness study. Behavior Therapy, 43, 160–173. https://doi.org/10.1016/j. beth.2011.05.001

Otto, M. W., Pollack, M. H., Gould, R. A., Worthington III, J. J., McArdle, E. T., Rosenbaum, J. F., & Heimberg, R. G. (2000). A comparison of the efficacy of clonazepam and cognitive- behavioral group therapy for the treatment of social phobia. Journal of Anxiety Disorders, 14, 345–358. https://doi.org/10. 1016/s0887-6185(00)00027-x

Piccirillo, M. L., Dryman, M. T., & Heimberg, R. G. (2016). Safety behaviors in adults with social anxiety: Review and future directions. Behavior Therapy, 47, 675–687. https:// doi.org/10.1016/j.beth.2015.11.005

Powers, M. B., Sigmarsson, S. R., & Emmelkamp, P. M. (2008). A meta-analytic review of psychological treatments for social anxiety disorder. International Journal of Cognitive Thera- py, 1, 94–113. https://doi.org/10.1521/ijct.2008.1.2.94

Rapee, R. M., & Heimberg, R. G. (1997). A cognitive- behavioral model of anxiety in social phobia. Behaviour Research and Therapy, 35, 741–756. https://doi.org/10. 1016/S0005-7967(97)00022-3

Rodebaugh, T. L., Heimberg, R. G., Brown, P. J., Fernandez, K. C., Blanco, C., Schneier, F. R., & Liebowitz, M. R. (2011). More reasons to be straightforward: Findings and norms for two scales relevant to social anxiety. Journal of Anxiety Disorders, 25, 623–630. https://doi.org/10.1016/j.janxdis.2011.02.002

Rodebaugh, T. L., Woods, C. M., & Heimberg, R. G. (2007). The reverse of social anxiety is not always the opposite: The reverse-scored items of the Social Interaction Anxiety Scale do not belong. Behavior Therapy, 38, 192–206. https://doi. org/10.1016/j.beth.2006.08.001

Safren, S. A., Heimberg, R. G., & Juster, H. R. (1997). Clients' expectancies and their relationship to pretreatment symptom- atology and outcome of cognitive-behavioral group treatment for social phobia. Journal of Consulting and Clinical Psychol- ogy, 65, 694–698. https://doi.org/10.1037/0022-006x.65.4.694

Shadish, W. R., Navarro, A. M., Matt, G. E., & Phillips, G. (2000). The effects of psychological therapies under clinically representative conditions: A meta-analysis. Psy- chological Bulletin, 126, 512–529. https://doi.org/10.1037/ 0033-2909.126.4.512

Stewart, R. E., & Chambless, D. L. (2009). Cognitive– behavioral therapy for adult anxiety disorders in clinical practice: A meta-analysis of effectiveness studies. Journal of Consulting and Clinical Psychology, 77, 595–606. https:// doi.org/10.1037/a0016032

RECEIVED: February 16, 2020 ACCEPTED: June 8, 2020 AVAILABLE ONLINE: 15 June 2020