Although it is relatively uncommon for adults with only a specific phobia to seek treatment, it is also one of the most effectively treated disorders. The primary treatment for specific phobias is CBT. Several controlled studies have found that up to 90% of participants experienced significant symptom reduction within 2 to 3 hr of therapy (Öst, 1989; Öst, Ferbee, & Furmark, 1997; Öst, Salkovskis, & Hellstroem, 1991). The effectiveness of CBT is likely the result of exposing the patient to feared stimuli in a supportive environment. This is commonly done systematically, with the therapist using a hierarchy to address the patient's fears. Research has also shown that although exposures tend to be more effective when therapists model the treatment tasks for the patient, direct observation in and of itself is generally not sufficient (e.g., Öst et al., 1997). With the exception of dental phobias (e.g., Jerremalm, Jaansson, & Öst, 1986), cognitive therapy alone has not been examined in the treatment of specific phobias.

Researchers have examined the effectiveness of benzodiazepines, beta-blockers, and SSRIs in treating specific phobias. Some controlled studies have found benzodiazepines to be helpful for patients who are averse to initial exposures to feared stimuli, or who simply need acute symptom relief (e.g., Schmidt, Koselka, & Woolaway-Bickel, 2001). However, both controlled and uncontrolled studies have found long-term use of benzodiazepines to be associated with unsatisfactory outcome and a myriad of other complications (Sanderson & Wetzler, 1993; Westra & Stewart, 1998). These unwanted effects include depression, a higher rate of symptom relapse once drugs are discontinued, and a risk of drug tolerance, dependence, and abuse (Spiegel & Bruce, 1997).

The use of benzodiazepines in conjunction with exposure therapy was found to interfere with the treatment efficacy of exposure (e.g., Marks et al., 1993; Wilhelm & Roth, 1997). Studies have shown benzodiazepines to interfere with both the acquisition and retention of new information (Barbee, Black, & Tordorov, 1992) and to affect the way new information is stored in memory (Gray, 1987). This interference can cause state-dependent learning, whereby information encoded in the presence of the drug may not be properly retrieved in a drug-free state. In addition, uncontrolled studies have reported that the use of benzodiazepines can interfere with the development of tolerance to anxiogenic stimuli on a neurochemical level (Gray, 1987). It has been speculated that these medications interfere with the central nervous system's ability to adapt to persistently elevated levels of neurotransmitters involved in the stress response (Anisman & Zacharko, 1992). More research is needed to further examine these potentially important findings.

Fear is a normal response to active or imagined threat that is characterized by affective, behavioral, and cognitive components. As discussed elsewhere (King et al., 1988; Ollendick, King, & Muris, 2002), nearly all children and adolescents experience some degree of fear. For the most part, these fears are adaptive—they appear to emanate from day-to-day experiences of children and they reflect their emerging cognitive and representational capabilities. Moreover, most of these fears do not involve intense or persistent reactions, and they are short-lived. This is not so with specific phobias.

Using criteria established for empirically supported treatments (see Chambless & Hollon, 1998; Chambless & Ollendick, 2000), Ollendick and King (1998, 2000) have indicated that two psychosocial treatments have attained “well-established” status (participant modeling, reinforced practice) and five treatments have achieved “probably efficacious” status (imaginal desensitization, in vivo desensitization, live modeling, filmed modeling, and verbal self-instruction) in the treatment of phobic disorders in children. The well-established treatments have been shown to be more effective than credible placebo controls in at least two RCTs, whereas probably efficacious treatments have

been shown to be more effective than wait-list control conditions in at least two RCTs or to be more effective than a credible placebo control in at least one RCT. Support for these interventions has come solely from intervention studies with children 12 or 13 years of age and younger. None of the studies treated specific phobia in adolescents. More research is needed with adolescents to determine the efficacy of behavioral treatments for them. Two recent studies conducted have examined cognitive-behavioral interventions in samples that included adolescents.

In the first Type 1 study examining the treatment of both children and adolescents, Silverman et al. (1999b) examined the relative benefits of an operant-based contingency management treatment and a cognitive-based self-control treatment, in comparison to an education-support control group. Graduated in vivo exposure was used in both the self-control and the contingency management conditions but not in the education-support condition. Phobic children (N = 81) between 6 and 16 years of age along with their parents were evaluated using child, parent, and clinician measures. The children were assigned randomly to one of the three 10-week manualized treatment conditions (i.e., self-control, contingency management, or education support). Although all three conditions were found to impart improvement in the child's functioning as measured by the reports of children, parents, and clinicians, clinically significant improvements were noted only in the two cognitive-behavioral treatment conditions. Specifically, on a measure of clinical distress at posttreatment, 80% of the participants in the self-control and 80% of the participants in the contingency management conditions reported very little or no distress, compared to only 25% in the education-support condition. Notably, 88% of the participants in the self-control condition no longer met diagnostic criteria at posttreatment compared to 55% in the contingency management and 56% in the education-support condition.

In the second Type 1 study, Öst, Svensson, Hellström, and Lindwall (2001) evaluated the effects of a cognitive-behavioral procedure referred to as “one-session treatment” on phobic children and adolescents 7 to 17 years old. The hall mark of this one-session treatment is a graduated, systematic, prolonged exposure to the phobic stimulus combined with the active dissuading and repair of faulty cognition. As such, it involves a combination of strategies including cognitive restructuring, in vivo graduated exposure, participant modeling, and social reinforcement. Notably, this treatment has been designed to be maximally effective in one session, and therefore lasts 3 or more hours in length. Results indicated that the treatment was superior to wait-list and produced significant gains both immediately posttreatment and at 1-year follow-up (Öst et al., 2001). Also, in a Type 3 study, one-session treatment has been found to be comparable to other, longer treatments with more sessions (see Muris, Merckelbach, Holdrinet, & Sijsenaar, 1998). Presently, Ollendick and Öst are completing a larger Type 1 study examining the effectiveness of one-session treatment in comparison to educational support and wait-list conditions. Approximately 200 youth with specific phobias are participating in this trial, which is being conducted in the United States and Sweden.

At present, no randomized pharmacological treatment studies of specific phobia in children and adolescents have been completed (Ginsburg & Walkup, 2004). The lack of such studies appears to be related to the common misconception that specific fears are a part of normal experience and not a condition associated with impairment. On the basis of the findings obtained with adults, however, SSRIs may be effective among children and adolescents with specific phobia (in contrast to the benzodiazepines and TCAs). A recent Type 3 open trial suggests their utility. Fairbanks et al. (1997) completed a 9-week trial of fluoxetine in children ages 9 to 18 years with mixed anxiety disorders (N = 16). After not responding to brief psychotherapy, patients were started on low-dose fluoxetine (5 mg/day), which was then increased weekly until side effects or improvement occurred to a maximum of 40 mg/day (children) and 80 mg/day (adolescents). Of the 16 patients enrolled, 6 had specific

phobia and 4 of the 6 responded to fluoxetine. The medication was also generally well tolerated.

To date, no randomized clinical trials have examined the joint efficacy of psychosocial and pharmacologic treatments with children or adolescents. Given the promise of several treatments, however, there is reason to believe that beneficial effects of combined treatment may occur. Research into their combinatorial effects is needed before any conclusions can be drawn.

The efficacy of cognitive, behavioral, and cognitive-behavioral treatment programs has been examined and is now well established for panic disorder. Under a broad theoretical umbrella, the emphasis in treatment has been on disconfirming mistaken beliefs about the meaning of physical sensations (Clark, 1986, 1996) and on exposing the patient to the feared physical sensations (Barlow, 1988). Cognitive-behavioral therapy has proven efficacious even when cognitive therapy is condensed (Clark et al., 1999). A commonly used protocol, panic control treatment, has been found to be efficacious in several studies (e.g., Zarate, Craske, & Barlow, 1990), and in the largest combined treatment study published to date, it was superior to pill placebo at the end of acute treatment but not superior to imipramine alone (Barlow et al., 2000). Notably, although panic control treatment typically includes breathing retraining, a procedure designed to minimize panic sensations, a recent controlled study indicated that the inclusion of breathing retraining did not provide incremental benefit above and beyond CBT alone, and on some outcome measures actually yielded less favorable outcome (Schmidt et al., 2000).

A wide range of medications have demonstrable efficacy for panic in the initial phase of treatment, including TCAs, low-and high-potency benzodiazepines, monoamine oxidase inhibitors (MAOIs), and SSRIs. Because of the comparable efficacy to that of older medications, yet a vastly superior side-effect profile, SSRIs are now considered first-line medications for treating panic. Pa roxetine was the first SSRI approved by the FDA for panic disorder and its antipanic effects have been demonstrated in placebo-controlled studies. A 12-week double-blind, placebo-controlled study of 367 panic patients found that both paroxetine and clomipramine were superior to placebo (Lecrubier, Bakker, Dunbar, & Judge, 1997). Compared to clomipramine, paroxetine acted more quickly to block spontaneous panic attacks, overall was associated with less adverse effects, and for 9 months after the 12-week acute phase demonstrated improvement. In a double-blind, placebo-controlled study, Ballenger, Davidson, Wheadon, and colleagues (1998) compared different doses of paroxetine and concluded that 40 mg daily was the most effective dosage. Sertraline (now also FDA approved), fluvoxamine, and fluoxetine have all shown significant ability to block panic compared to placebo. The general pattern of treatment response is as follows. First, spontaneous panic attacks are blocked, usually within the first 6 to 8 weeks. This is followed by a gradual reduction in anticipatory anxiety and agoraphobic avoidance behavior. A critical issue that has emerged from virtually all studies of SSRI treatment for panic is the vulnerability of these patients to jitteriness, restlessness, and increased anxiety during initial dosing. Starting with low doses is thus recommended (e.g., paroxetine 10 mg/day, sertraline 25 mg/day, fluvoxamine 50 mg/day, fluoxetine 10 mg/day) with gradual increases to the optimal target dose.

High-potency benzodiazepines are also effective for panic disorder. Alprazolam and clonazepam have been the most widely studied. In Phase 1 of the Cross-National Collaborative Panic study of 481 panic patients, alprazolam at a mean dose of 5.7 mg/day was more effective than placebo with a 55% response rate (compared to 32% placebo rate) at the 8-week end point (Ballenger, Davidson, Lecrbier, et al., 1998). Rosenbaum, Moroz, and Bowden (1997) conducted a multicenter, parallel group, placebo-controlled, fixed-dose study of clonazepam in 413 patients. Patients were randomly assigned to one of five fixed doses. The results indicated significant improvement at all doses but a differentiation of the four higher doses from the 0.5-mg dose and placebo. The mini

mum effective dose was 1 mg; beyond this dose all doses were equally efficacious. Reticence to prescribe benzodiazepines for panic has been based on the (unwarranted) assumption that patients will gradually escalate the dose to achieve therapeutic effect in a pattern similar to that seen in substance abusers. However, controlled studies have shown that panic disorder patients are not likely to increase dosages and actually found such increases to be uncomfortable and associated with unwanted effects (Salzman, 1993).

With respect to TCAs, imipramine has been the most studied antipanic agent, with several studies demonstrating superiority over placebo. The serotonergic-acting clomipramine also has strong antipanic effects and in one placebo-controlled trial was actually superior to imipramine and placebo (Modigh, Westberg, & Eriksson, 1992). However, the majority of studies point to equivalent efficacy for TCAs and SSRIs yet a superior side-effect profile with SSRIs (Bakker, Spinhoven, van der Does, van Balkom, & van Dyck, 2002). Although venlafaxine, nefazodone, and mirtazapine have not been studied in a controlled fashion, they have all been shown to exhibit antipanic action, particularly in patients with comorbid depression. For a detailed review of the literature on panic disorder treatment outcome see Roy-Byrne and Cowley (2002).

Panic disorder can be a disabling condition accompanied by psychosocial, family, peer, and academic difficulties (Birmaher & Ollendick, 2004; Moreau & Weissman, 1992; Ollendick, Mattis, & King, 1994). In addition, panic disorder is associated with increased risk for other anxiety disorders, major depressive disorder (MDD), and substance abuse (Birmaher & Ollendick, 2004; Moreau & Weissman, 1992). Moreover, such adverse outcomes are more prevalent in adults whose panic disorder starts early in life (before 17 years of age; see Weissman et al., 1997). Despite this, it takes on average 12.7 years from the onset of reported symptoms for adults to initiate and seek treatment (Moreau & Follet, 1993). Unfortunately, it appears that very few youngsters with panic disorder seek help at all (Essau, Condradt, & Petermann, 1999).

Psychosocial treatments for panic disorder have been based largely on cognitive and cognitive-behavioral theories (Chambless & Ollendick, 2000), and the few studies of cognitive-behavioral treatment of panic disorder in children and adolescents have used a protocol based largely on panic control treatment (Barlow, 1988). This treatment consists of three primary strategies: relaxation training and breathing retraining to address neurobiological sensitivities to stress, interoceptive exposure to address heightened somatic symptoms, and cognitive restructuring to address faulty misinterpretations associated with the somatic symptoms. Adolescent treatment studies based on Barlow's model have been sparse; however, two Type 3 case series studies lend initial support to its use and one randomized, large-scale, Type 1 treatment outcome study appears promising. In the first of the Type 3 studies, Barlow and Seidner (1983) treated three adolescents who had panic disorder with agoraphobia using an early version of panic control treatment. The patients were treated in a 10-session group-therapy format and were accompanied by their mothers, who were enlisted to facilitate behavior change. Following treatment, two of the three adolescents showed considerable improvement in their symptoms; the third did not show significant change. In the second of the Type 3 studies, Ollendick (1995) used a multiple baseline design to illustrate the effects of an adapted version of panic control treatment with four adolescents who had panic disorder with agoraphobia. The adolescents ranged in age from 13 to 17 years; however, their panic attacks began when they were between 9 and 13 years of age, nearly 4 years on average prior to the beginning of treatment. The patients were treated individually but, as in Barlow and Seidner's (1983) study, their mothers were enlisted to facilitate behavior change. Treatment varied in duration but lasted between 10 and 12 sessions for the four adolescents. Treatment was effective for all the patients eliminating panic attacks, reduc

ing agoraphobic avoidance, decreasing accompanying negative mood states, and increasing self-efficacy for coping with previously avoided situations and panic attacks should they occur in the future.

Presently, Mattis and colleagues are in the final stages of a Type 1 RCT of a developmental adaptation of panic control treatment for adolescents (PCT-A) and a self-monitoring control condition in the treatment of panic disorder. Three aspects of panic disorder are addressed in this 11-session, manualized intervention: (1) the cognitive aspect of panic disorder or the tendency to misinterpret physical sensations as catastrophic; (2) the tendency to hyperventilate or overbreathe, thus creating and/or intensifying physical sensations of panic; and (3) conditioned fear reactions to the physical sensations. Although the trial is still under way, initial findings suggest a reduction in severity of panic attacks and panic disorder in patients receiving PCT-A relative to those in the monitoring control group (Mattis et al., 2001).

No RCTs for the pharmacological treatment of panic disorder in children and adolescents have yet been completed (Birmaher & Ollendick, 2004). In children and adolescents, anecdotal case reports suggest that benzodiazepines and the SSRIs (Birmaher & Ollendick, 2004) may be efficacious for panic disorder. For example, in a prospective Type 3 open trial, Renaud, Birmaher, Wassick, and Bridge (1999) treated 12 children and adolescents with panic disorder with SSRIs for a period of 6 to 8 weeks. Nearly 75% of the youth showed much to very much improvement with SSRIs without experiencing significant side effects. At the end of the trial, eight (67%) no longer fulfilled criteria for panic disorder whereas four (33%) continued to have significant and lasting effects.

Panic disorder is accompanied frequently by a variety of other mental health disorders (Biederman, Faraone, et al., 2001; Mattis & Ollendick, 2002). Treatment of these comorbid conditions may be needed to improve the youngster's overall functioning. Fortunately, two of the common comorbid disorders, depression and other anxiety disorders, also respond to CBT and/or SSRIs (Birmaher et al., 1994; Brent et al., 1997; Emslie et al. 1997; Kendall et al., 1997; Ollendick & King, 1998; RUPP, 2001).

To date, no randomly controlled clinical trials have examined the joint efficacy of psychosocial and pharmacologic treatments with children or adolescents with panic disorder. Given the independent promise of both treatments, however, there is reason to believe that synergistic effects could occur. Nevertheless, research into their separate and combined effects, such as that pursued in adult populations (e.g., Barlow et al., 2000), is needed.

Among the psychosocial treatments for OCD, exposure and response (or ritual) prevention (EX/RP) is the most well-established treatment for adults, having been found superior to a variety of active and control treatments (for a review see Franklin & Foa, 2002). In this treatment patients gradually confront situations that evoke their obsessional fears (e.g., sitting on the floor, where they feel contaminated by germs) and at the same time are refrained from reducing their discomfort by performing ritualistic behaviors (e.g., excessive washing and cleaning).

The only medications studied in randomized clinical trials and reported to be efficacious for OCD are serotonin reuptate inhibitors (e.g., clomipramine) and the SSRIs (e.g., fluoxetine, fluvoxamine, sertraline, paroxetine, citalopram). The data suggest a 20% to 40% reduction in symptoms, but many patients classified as responders in these trials still have clinically significant symptoms. Moreover, patients with severe comorbid MDD or significant substance abuse have usually been excluded (for a review see Dougherty, Rauch, & Jenike, 2002).

Several RCTs examining the efficacy of combined treatment have been conducted in OCD (Cottreaux et al., 1990; Foa, Rothbaum, & Furr, 2003; Hohagen et al., 1998; Marks et al., 1988; van Balkom et al., 1998), and their collective re