ful for social anxiety disorder. There are no controlled trials of patients with nongeneralized social anxiety disorder, and studies containing both generalized and nongeneralized patients include too few of the latter to be definitive. Placebo-controlled trials of patients with the generalized subtype suggest efficacy for phenelzine, clonazepam, paroxetine, sertraline, fluvoxamine, venlafaxine, gabapentin, and pregabelin. The average responder rate in these trials is 40%–55%, defining response as a 1 or 2 on the Clinical Global Impression (CGI) Improvement scale. A majority of the patients classified as responders in these trials still have clinically significant symptoms. Patients with comorbid major depressive disorder (MDD) or significant substance abuse have usually been excluded. Drugs such as phenelzine or sertraline have shown greater acute efficacy than CBT in trials comparing the two modalities. Phenelzine plus CBT, and sertraline plus CBT, showed modestly greater efficacy than the respective drug monotherapies alone in controlled trials. Remission has not been assessed in any trials, and there is also no agreement on the criteria to define remission in social anxiety disorder. Negative findings from controlled trials exist for buspirone, atenolol, moclobemide, and fluoxetine in generalized social anx

iety disorder (for a review see Roy-Byrne & Cowley, 2002).

Increased self-consciousness and preoccupation with social matters are extremely common during late childhood and early adolescence, a developmental stage in which the importance of peers is typically heightened. Accordingly, adult epidemiological studies have also consistently indicated that onset of social phobia is common during late childhood, with onset typically occurring between ages 11 and 12 (Albano, 2003). Differentiating normative adolescent social concerns from the clinical social anxiety disorder poses difficulties for clinicians, parents, teachers, and adolescents alike, and perhaps the best way to parse the two is to examine functional impairment and concomitant symptoms: compared to non-anxious counterparts, youth with social anxiety disorder more often report depressed mood, high trait anxiety, and perceived social incompetence (Albano, Chorpita, & Barlow, 2003), higher levels of loneliness and fewer friends (Beidel, Turner, & Morris, 1999), a wide range of social avoidance (Albano, Detweiler, Logsdon-Conradsen, Walker, & Ollendick, 1999), and, in at least some cases, social anxiety is associated with school refusal (Kearney & Drake, 2002). Given the potentially serious consequences of social anxiety disorder and its sequalae for the life trajectory of adolescents, detection and treatment are especially important.

There is an increasing body of evidence demonstrating the efficacy of CBT approaches for children and adolescents with anxiety disorders, which has included those with social phobia (or with the previous diagnostic term, avoidant disorder). A number of controlled trials targeting anxious children have included those diagnosed with social phobia (Barrett, Dadds, & Rapee, 1996; Barrett, 1998; Cobham, Dadds, & Spence, 1998; Dadds et al., 1999; Dadds, Spence, Hol land, Barrett, & Laurens, 1997; Flannery-Schroeder & Kendall, 2000; Ginsburg & Schlossberg, 2002; Kendall, 1994; Kendall et al., 1997; Last, Hansen, & Franco, 1998; Lumpkin, Silverman, Weems, Markham, & Kurtines, 2002; Silverman et al., 1999a, 1999b).

Three Type 1 randomized, controlled cognitive-behavioral treatment studies specifically for children and adolescents with social phobia have reported positive clinical response. Hayward et al. (2000) randomly assigned 35 female adolescents with social phobia to cognitive-behavioral group therapy (CBGT) (N = 12) or no treatment (N = 23). Eleven of the 12 cases completed the active treatment. After 16 weeks of treatment, significantly fewer treated patients met criteria for social phobia. At 1-year follow-up, however, there was no difference between the two groups, suggesting that the CBGT may result in a moderate short-term effect. Spence, Donovan, and Brechman-Toussaint (2000) randomly assigned 50 children with social phobia, ages 7–14 years, to either child-focused CBT, CBT plus parent involvement, or a wait-list control. The integrated CBT program involved intensive social skills training with graded exposure and cognitive challenging. After treatment, children in both CBT groups had a greater decrease in social and general anxiety and a greater increase in parental ratings of child social skills than did those in the wait-list control group. At 12-month follow-up, both CBT groups had retained their improvement. Beidel, Turner, and Morris (2000) demonstrated the efficacy of social effectiveness therapy for children (SET-C) in comparison to an active but nonspecific intervention (Testbusters): children ages 8–12 treated with SET-C had enhanced social skill, reduced social anxiety, decreased associated psychopathology, and increased social interaction both at posttreatment and at 6-month follow-up.

Researchers have also examined whether the treatment should be delivered individually, with family, or with peers. In the Spence et al. (2000) study described above, superior results seemed to have emerged when parents were involved in CBT treatment, but this effect was not statistically significant. In a Type 3 study, Masia, Klein, Storch, and Corda (2001) piloted a school-based behavioral treatment for adolescents and con

cluded that the school setting is a logical place to deliver the treatment because it is where the individuals with social anxiety endure the most distress.

Current research is attempting to identify the necessary or sufficient components in the psychosocial treatment of social phobia, as well as the optimal context. Based on meta-analyses of adult studies, exposure in some form may be a key ingredient (Beidel et al., 2000). Treatment of the adolescent with social phobia may require different treatment interventions from those used in the younger ages (Beidel et al., 2000; Spence et al., 2000).

As recently as 1999, no definitive data existed on the efficacy of psychopharmacology in the treatment of anxiety disorders in youth (excluding OCD) (Labellarte, Ginsburg, Walkup, & Riddle, 1999). In 2001, fluvoxamine, an SSRI, was studied in children and adolescents 6 to 17 years of age with either social phobia, SAD, or generalized anxiety disorder (GAD; Walkup et al., 2001). Youth (N = 153) were evaluated and enrolled in a 3-week open treatment trial with supportive psychoeducational therapy. Only five children improved with brief psychoeducation and did not go on to medication therapy. One hundred and twenty-eight children were assigned to either fluvoxamine or placebo for 8 weeks. The fluvoxamine dose was increased every week by 50 mg to a maximum of 300 mg/day for adolescents and 250 mg/day for children less than 12 years of age. The average dose of fluvoxamine was 2.9 + 1.3 mg/kg, and the average last dose was 4.0 + 2.2 mg/kg. The medication was generally well tolerated; five children in the fluvoxamine group and one in the placebo group discontinued because of adverse effects. Adverse effects were more common in the medication group; abdominal discomfort was significantly greater in the fluvoxamine group than in the placebo group, and there was a trend toward a greater frequency of increased motor activity in the fluvoxamine group. The youth in the fluvoxamine group had significantly greater reductions in symptoms of anxiety and higher rates of clinical response than did the children in the placebo group. On the CGI scale, 48 of 63 (76%) of children had a response to treatment, in comparison to only 19 of 65 children (29%; p < .001). The study led to the conclusion that fluvoxamine is an effective treatment for children and adolescents with social phobia, SAD, or generalized anxiety.

In follow-up analyses, Walkup and colleagues (RUPP, 2003) examined the data for moderators and mediators of pharmacologic response in these 128 youths. Interestingly, no significant moderators of efficacy were identified, except for lower baseline depression scores, based on “parent” (but not “child”) report, which were associated with greater improvement. Patients with social phobia and greater severity of illness, irrespective of medical assignment, were less likely to improve. Further study will be needed to determine why the diagnosis of social phobia predicted a less favorable outcome in this specific study.

Although there have been few studies examining the effects of treatments that combine CBT and medication in adults, no systematic studies of combined treatment for children and adolescents with social phobia exist. Given the lack of information about adolescents with anxiety disorders, efficacy studies on both CBT and pharmacotherapy and on the relative efficacy of each treatment alone and in combination are very much needed.

Behavioral treatment of selective mutism has generally been the initial and the primary intervention in clinical practice, but there are no randomized controlled trials (RCTs) evaluating behavioral treatment. On the basis of Type 6 case reports and single case designs, contingency management, exposure-based techniques, and self-modeling are the most frequently used behavioral interventions (Anstendig, 1998; Cunningham, Cataldo, Mallion, & Keyes, 1984; Holmbeck & Lavigne, 1992; Kehle, Owens, & Cressy, 1990).

Despite the paucity of controlled trials on the efficacy of medications for the treatment of selective mutism, pharmacotherapy is often used in clinical practice. In one Type 1 report, Black and Uhde (1994) treated 16 children with selective mutism with single-blind placebo for 2 weeks. The 15 placebo nonresponders were then randomly assigned to double-blind treatment with fluoxetine (N = 6) or placebo (N = 9) for 12 weeks. The mean maximum dose of fluoxetine was 21.4 mg/day (range 12–27 mg/day, 0.60 to 0.62 mg/kg/day). Side effects were minimal, and all double-blind patients completed the 12 weeks. Patients on fluoxetine (N = 6) were significantly better than those on placebo (N = 9) on parent's rating of mutism change and global change. However, clinician and teacher ratings did not show any significant differences between those receiving medication and those on placebo. The authors suggested that this could in part be due to the small number of patients or the more severe baseline symptoms of the medication group. Nevertheless, the authors cautioned that despite improvement for some, patients in both groups remained very symptomatic at the end of the study.

Several CBT programs have been developed and empitacally evaluated for GAD. As might be expected given the nature of GAD, some of these protocols place less emphasis on exposure to specific fear cues and instead focus primarily on the development of effective coping strategies, of which exposure-based procedures (e.g., worry time) are among several techniques used. The outcome data for CBT are generally encouraging, with several meta-analytic reports indicating very large within-subjects effects and moderate to large effects in comparison to psychosocial control treatments and to low-dose diazepam treatment (Borkovec & Ruscio, 2001; Gould, Buckminster, Pollack, Otto, & Yap, 1997). Although the efficacy of several CBT programs has been established, outcome studies also suggest that there is significant room for improvement, especially with respect to the percentage of patients who achieve high end-state functioning (Orsillo, Roehmer, & Barlow, 2003). This observation has prompted recent innovations in GAD treatment, including the addition of interpersonal elements into the existing CBT programs (Borkovec, Newman, & Castonguay, 2003).

Several medications have been found effective for acute treatment of GAD. Benzodiazepines, such as alprazolam, have demonstrated short-term efficacy for reducing GAD symptoms, with response rates (usually defined as a 40%–50% improvement in symptoms on the Hamilton Anxiety Scale or a CGI improvement rating of “moderately” or “much” improved) as high as 75%. Benzodiazepines have the advantage of being faster acting than alternative medications. Despite concerns about side effects (e.g., sedation, memory loss, physiologic dependence, withdrawal syndromes), benzodiazepines are still among the most widely used medications for treatment of GAD. Buspirone has also been found effective in controlled trials and its side-