Principal Authors: Robert A. King, M.D., Henrietta Leonard, M.D., and John March, M.D. This Summary was developed by the Work Group on Quality Issues: William Bernet, M.D., Chair, and John E. Dunne, M.D., former Chair; Maureen Adair, M.D., Valerie Arnold, M.D., R. Scott Benson, M.D., Oscar Bukstein, M.D., Joan Kinlan, M.D., Jon McClellan, M.D., and David Rue, M.D. AACAP Staff: L. Elizabeth Sloan, L.P.C. The full text of the Practice Parameters for the Assessment and Treatment of Children and Adolescents With Obsessive-compulsive Disorder is available to Academy members on the World Wide Web (www.aacap.org) and appears in the October 1998 supplement to the Journal. The full text of these parameters was reviewed at the 1997 Annual Meeting of the American Academy of Child and Adolescent Psychiatry. Both the full text and this Summary were approved by the AACAP Council on April 1, 1998. Reprint requests to AACAP, Communications Department, 3615 Wisconsin Avenue, N.W., Washington, DC 20016. [xxxxx]©1998 by the American Academy of Child and Adolescent Psychiatry.

ABSTRACT

This summary provides an overview of the assessment, differential diagnosis, and treatment recommendations contained in the Practice Parameters for the Assessment and Treatment of Children and Adolescents With Obsessive-compulsive Disorder. OCD is a disorder of heterogeneous origin characterized by intrusive thoughts or compulsive urges or behaviors that are distressing, time-consuming, or functionally impairing. In children and adolescents, the disorder often is accompanied by a wide range of comorbidity. Two modalities have been systematically assessed and empirically shown to ameliorate the core symptoms of OCD: cognitive behavioral therapy (primarily exposure/response prevention) and serotonin re-uptake inhibitor medication. Additional individual and family psychotherapeutic, pharmacological, and educational interventions often are necessary. Key Words: children, adolescents, obsessive, compulsive, anxiety, tic, cognitive, behavioral, serotonin, practice parameters, guidelines.

These parameters give the clinician direction in the assessment and treatment of obsessive-compulsive disorder (OCD) in children and adolescents. These parameters are not intended to define the standard of care; nor should they be deemed inclusive of all proper methods of care or exclusive of other methods of care directed at obtaining the desired results. The ultimate judgment regarding the care of a particular patient must be made by the clinician in light of all the circumstances presented by the patient and his or her family, the diagnostic and treatment options available, and available resources. Recommendations are based on extensive review of the scientific literature and clinical consensus among experts in the subject. The literature review, including references, and the rationale for specific recommendations are contained in the complete document (American Academy of Child and Adolescent Psychiatry, 1998).

ASSESSMENT

The diagnostic evaluation of children with OCD includes a careful assessment of current and past obsessive-compulsive (OC) symptoms and comorbid conditions. A comprehensive evaluation of the child's development and psychosocial functioning, including a detailed review of the child's medical, developmental, and family histories, is essential. This evaluation requires interviewing both child and parents and usually requires more than one session. Repetitive, perfectionistic, or ritualistic behaviors and recurrent worries are common in children at various stages of development. Thus, true OC symptoms must be distinguished from childhood rituals and routines and anxious worries. For example, obsessive concerns about sameness or symmetry, "just right" phenomena, insistence on order in certain situations (e.g. bedtime rituals), and upset if thwarted, occur as an apparently normal developmental phenomena in as many as two-thirds of preschoolers. True obsessions and compulsions are recurrent and intrusive thoughts and urges that are distressing, bothersome, and interfere with daily functioning. Although children may experience their repetitive thoughts or urges as senseless or excessive, this is highly variable. Children often are secretive about their symptoms and may deny or minimize them. In addition to identifying specific symptoms, it is important to assess their context, frequency, detailed phenomenology, and degree of associated distress and impairment, as well as the child's attitude and degree of insight into and resistance to the symptoms. One should ask whether the child performs compulsive acts or rituals to relieve obsessions or prevent feared outcomes; what would happen if the child did not perform the compulsion and how does he or she know when it has been done enough? Total time spent on a worry or compulsion is one useful measure of severity; in other cases, where the symptoms are brief but frequent, the usual longest symptom-free period may be a more useful measure of severity. Once the diagnosis of OCD has been established, instruments such Children's Yale-Brown Obsessive Compulsive Scale (CY-BOCS) can be used to rate and record symptom severity, provide a useful guide to assessment, and are sensitive to change. The degree to which parents have become entangled with the child's symptoms is important to assess, since such involvement often seriously impairs family life, interferes with treatment, and helps perpetuate the child's symptoms.

DIFFERENTIAL DIAGNOSIS

Given the close association of tic disorder and OCD, it is important to inquire about the history of motor or phonic tics. Compulsive habits may be difficult to distinguish from complex tics; both may be preceded by premonitory urges that persist until the action is completed. In addition to compulsions, inquiry should be made concerning compulsive habits such as hair-pulling, nail-biting, or face- or scab-picking. While focusing on the child's OC symptoms, it is important also to assess the overall context of the child's personality, functioning, and developmental adaptation to his or her family, school, and social environment (American Academy of Child and Adolescent Psychiatry, 1997). A careful developmental history is necessary to identify areas of delay or difficulty. The presence of depression, anxiety, perfectionism, irritability, impulsivity, aggression, eating or body image concerns, or psychotic symptoms and their relationship to the obsessions and compulsions should be assessed. Family history of OCD and related tic or anxiety disorders should be elicited, as these disorders often are familial. The child's school performance should be assessed to determine if compulsive re-reading or re-writing; pathological perfectionism; co-morbid anxiety, attentional and impulse problems; or associated cognitive impairments have an adverse impact. Acute onset or exacerbation of OC symptoms or tics requires careful consideration of recent medical illnesses, including upper respiratory infections. A throat culture for strep and antistreptolysin O (ASO) and anti-streptococcal DNAase B titres may be considered to assist in diagnosing a Group A Beta-hemolytic streptoccocal (GABHS) infection, which has been speculated to produce such a symptom picture in certain children.

TREATMENT

Children with OCD vary significantly as to the specific nature and severity of their OC symptoms, the range of co-morbid psychopathology, and the impact of the disorder on the child's and family's functioning. These individual features may have important implications for treatment planning in terms of compliance, response to treatment, and exacerbating or ameliorating factors. In addition to targeting the content and severity of the core OCD symptoms, effective treatment planning for children with OCD also must take into consideration the presence of co-morbid difficulties; the child's developmental level, personality, and adaptive functioning; and the family context. To help ensure optimal involvement in the treatment, both the patient and family should participate to the greatest extent possible in the development of the individualized treatment plan. Children with OCD often are secretive about their symptoms. By the time they come to clinical attention, children with OCD and their parents may be caught in a vicious cycle of mutual coercion and irritation. The process of assessment and treatment planning may help destigmatize and reframe the child's symptoms and encourage a treatment alliance. Only two treatment modalities, cognitive-behavioral therapy (CBT) and serotonin-reuptake inhibitors (SRIs), have been studied systematically and empirically shown to have specific efficacy for the core symptoms of OCD. Although some uncontrolled case studies have found psychotherapy useful in treating childhood OCD or OC personality traits, the effectiveness of psychotherapy alone (other than CBT) for the core obsessive and compulsive symptoms of OCD has not been studied systematically. However, psychotherapy may be an important adjunctive component of a comprehensive treatment approach to children with OCD, playing a significant role in teaching coping skills, increasing the child's sense of mastery, addressing other co-morbid diagnoses and family issues, and helping to improve peer and family relationships. Family psychopathology is neither necessary nor sufficient for the onset of OCD. Nonetheless, families affect and are affected by the disorder. For example, high "expressed emotion" may exacerbate OCD; a calm supportive family may improve outcome. Some families become extensively involved in participating in the child's compulsive rituals or reassuring obsessional worries in an effort to avoid anxious or angry blow-ups; other families resist participation but become mired in grueling angry struggles and arguments with their symptomatic child. Work with families on how to manage the child's OCD symptoms and participate effectively in behavioral and pharmacological treatment is thus essential. Interventions with parents to diminish their entanglement with the child's symptoms must go hand in hand with measures that diminish the child's distress. Family support groups also can help families to acquire useful intervention skills and prevent discouragement. In addition to these measures directed specifically at the child's obsessions and compulsions, family work is often needed around the host of adaptational, interpersonal, and co-morbid difficulties that may accompany childhood OCD.

COGNITIVE-BEHAVIORAL PSYCHOTHERAPY

Cognitive-behavioral therapy is regarded by many as the psychotherapeutic treatment of choice for children and adolescents with OCD. This treatment presents a logically consistent and compelling relationship between the disorder, the treatment, and the specified outcome. Furthermore, in contrast to medication, where relapse is common when treatment is withdrawn, CBT has been shown to be a durable treatment, although booster or refresher sessions may be required from time to time. Treatment generally involves a three-stage approach consisting of information gathering, therapist-assisted exposure and response prevention (E/RP), and homework assignments. As with adults, hierarchy-based exposure and response (E/RP) prevention is the central element in treatment. Exposure relies upon the fact that anxiety usually attenuates after sufficient duration of contact with a feared stimulus. Thus a child with fear of germs must confront relevant feared situations until his or her anxiety decreases. Repeated exposure is associated with decreased anxiety across exposure trials, with anxiety reduction largely specific to the domain of exposure, until the child no longer fears contact with specifically targeted phobic stimuli. Exposure is typically implemented in a gradual fashion (sometimes termed graded exposure), preferably with the patient playing an active role in the choice and sequence of situations to be targeted for exposure. Adequate exposure depends on blocking rituals or avoidance behavior, a process termed response prevention. For example, a child with germ worries must not only touch "germy things," but must refrain from ritualized washing until his or her anxiety diminishes substantially. Selection of E/RP tasks by the child from those items where the child is already successfully resisting OCD maximizes the probability that E/RP will be successful. Component analyses of CBT in adult OCD suggest that both exposure and response prevention are active ingredients of treatment, with exposure reducing phobic anxiety, and response prevention, rituals. Clinically, both temperament and developmental stage influence a child's ability to understand the requirements of E/RP and willingness and ability to tolerate the often intense affects associated with E/RP. Beyond adjusting the therapeutic conversation to match the child's to level of understanding, CBT typically includes a "tool kit" the child can use to manage thoughts and feelings before, during, and after E/RP. Typical interventions for children with predominately internalizing symptoms include relaxation training and cognitive training, which have proven helpful in children with other anxiety disorders. While relaxation appears to have no direct beneficial effect on obsessions or compulsions, relaxation may aid children with high levels of physical anxiety symptoms to complete E/RP tasks successfully. Cognitive therapy for OCD, as distinct from response prevention for mental rituals, includes reinforcing accurate information on OCD and its treatment, self-administered positive reinforcement, targeting erroneous "OCD beliefs," and cultivating psychological distance from OCD symptoms. Used alone, cognitive therapy is a less effective treatment for OCD than E/RP. Although positive or negative consequences are not especially helpful by themselves as treatments for OCD per se, the use of a systematic reinforcement plan (such as rewards contingent on compliance with CBT exercises) may be helpful in maintaining compliance in children with oppositional behavior or other disruptive behavior disorder. While clinical observations suggest that a combination of individual and family sessions is best for most patients, determining the extent of family involvement in treatment is a key therapy process consideration. Extensive family involvement in rituals or other family dysfunction that interferes with individual CBT requires a more central family role in treatment. However, abrupt, unilateral, or confrontational interventions, such as when a parent stops participating in OCD rituals without the child's consent, are almost never helpful because: (1) parents often have no workable strategy for managing the child's distress; (2) target symptoms often are inaccessible; and (3) such measures fail to help the child internalize a strategy for coping with current and potential OCD symptoms. Implementation of CBT requires establishing a treatment alliance with the child. Even with children who regard their obsessions and compulsions as ego-dystonic, enlisting the child's cooperation entails helping the child to understand that he or she will not be required to tolerate unbearable anxiety, that the child will play an active role in choosing the sequence of symptoms and situations to be confronted, and that compliance with treatment will help to increase mastery and ultimately diminish anxiety. Most children readily comply with cognitive-behavioral psychotherapy when it is presented correctly, and experience significant symptom relief. Prognostic indicators of good response to CBT include a motivated patient willing to cooperate with treatment, presence of overt rituals and compulsions, ability to monitor and report symptoms, and absence of complicating co-morbidities. Factors contributing to partial or non-response to CBT alone include extensive co-morbidity; family conflict interfering with CBT; and developmental factors, such as very young age, mental retardation, or pervasive developmental disorder. CBT appears less effective in children with obsessions only than in children with compulsions; primary obsessional slowness appears to respond poorly to both behavioral and medication treatment. Such children require other cognitive behavioral techniques, such as modeling and shaping or thought-stopping.

PHARMACOTHERAPEUTIC INTERVENTIONS

Serotonin Reuptake Inhibitors

The availability of the SRI clomipramine and the selective serotonin reuptake inhibitors (SSRIs) fluoxetine (Prozac), fluvoxamine (Luvox), paroxetine (Paxil), and sertraline (Zoloft) have dramatically changed the treatment of OCD.

Clomipramine

Clomipramine is the most extensively studied of the SRIs in children. Clomipramine is significantly superior to placebo and non-serotonergic tricyclic antidepressants (TCAs), with moderate improvement apparent by 5 weeks in 75% of the subjects receiving clomipramine. An adequate therapeutic trial of clomipramine generally consists of dosages up to 3 mg/kg/day for 3 months. Dosages should not exceed 5 mg/kg/day or 250 mg/day because of the risks of toxicity, including seizures and electrocardiogram (ECG) changes. Clomipramine's anticholinergic and antihistaminic side effects are typical of other TCAs. Anticholinergic side effects include dry mouth, somnolence, dizziness, tremor, headache, constipation, stomachache, sweating, and insomnia. As with other TCAs, clomipramine can cause tachycardia and EKG axis changes and interval prolongations; thus many authors recommend baseline and periodic EKG monitoring, with particular attention to the QTc interval.

Selective Serotonin-reuptake Inhibitors

The SSRIs represent a new class of agents with distinct advantages over the TCAs in side effect profile and therapeutic index. Initial studies suggest that the SSRIs are safe, effective, and tolerated well in children and adolescents, with a side effect profile similar to that seen in adults. In general, the most commonly described side effects of the various SSRIs include nausea, headache, tremor, gastrointestinal complaints, drowsiness, insomnia, akithisia, disinhibition, agitation, or even hypomania. In adults on SSRI's, a "frontal-lobe" syndrome, characterized by apathy and/or disinhibition, has been described. These agents also can alter sleep architecture, producing insomnia, daytime sedation, and/or impaired cognitive performance. Although the SSRIs can occasionally exacerbate or precipitate tics, especially at higher doses, they frequently are useful in treating OCD in patients with tics, without exacerbating them in most cases. Although direct comparison data in children are lacking, the SSRIs appear to have clinically significant differences in pharmacokinetic and side effect profile, half-life, and relative ability to alter the metabolism of other medications. Little is known regarding the pharmacokinetics of these agents in children and the special vulnerabilities polypharmacy may pose. The possibility of clinically significant drug interactions is increased with drugs that: (1) induce or inhibit hepatic microsomal enzymes, (2) have a low therapeutic index, (3) have multiple pharmacologic actions, and (4) may be metabolized differently in high-risk populations. Fluoxetine is distinguished from the other SSRIs by the long half-life of the parent compound and active metabolite; steady state is not reached for 2 to 3 weeks, and the drug is not completely eliminated from the system for up to 6 weeks after discontinuation. Fluoxetine has the advantage over other SSRIs of being available in liquid form, permitting initial dosages in children as low as 2.5 mg/day to 5.0 mg/day.

Choice of Agent

Although the serotonergic agents differ in potency and selectivity, these qualities appear unrelated to their clinical antiobsessional efficacy. In the absence of direct comparison and meta-analytic studies, it is not known if one SRI/SSRI is more effective than another for treating childhood OCD. Thus, in practice, the choice of agent may depend on side effect profile and potential for drug interactions. Clomipramine has the most anticholinergic side effect profile, requires EKG monitoring in children, and is the most toxic in an overdose. In contrast, the SSRIs do not require EKG monitoring, but may be associated with more complaints of headaches, nausea, insomnia, or agitation. Sometimes, comorbidity might argue for or against a TCA, although this has not been well studied. The hepatic metabolism and competitive inhibition profile (P450 enzyme system) of each agent should be considered if the concomitant use of other medications is contemplated. In summary, consideration of medical issues, side effects, concomitant medications, comorbid disorders, suicide risk, and previous trials require an individual decision for each patient.

Dose and Duration

A trial of adequate dosage and duration (10 to 12 weeks) is generally necessary to determine whether a child is a responder to a given SRI/SSRI. Many patients do not show symptom improvement until 6 to 10 weeks and may continue to improve during the first 3 months of pharmacotherapy. Some patients may develop a worsening of their OCD symptoms or complain of an agitated feeling ("jitteriness syndrome") in the first 10 days of treatment. If tolerable, the patient should be encouraged to continue the trial, perhaps at a reduced dose, since this phenomenon usually subsides after the initial few days. Systematic dose-response data are not available for children. Fixed dose design studies in adults suggest that lower dosages may be as effective as maximal dosages with fewer side effects. Since a clinical response is unlikely in the first 3 weeks, it is preferable to start with a low initial dosage, which can be increased slowly, thereby allowing a patient to tolerate the medication and avoiding dose-related side effects. The duration of a trial is probably as critical as dosage, suggesting the advisability of targeting a therapeutic dosage and waiting at least 10 to 12 weeks before changing agents or undertaking augmentation regimens.

Responders and Nonresponders

Failure to respond to one SSRI does not necessarily predict failure to respond to another SSRI. Thus, if there is no clinical response after 10 to 12 weeks, then switching to another SSRI or clomipramine is reasonable. Unfortunately, there are no systematic studies that compare switching medications with adding an augmenting agent to the initial medication. Even on adequate trials of different SRIs, however, a substantial number of patients either are non-responders or have significant residual OCD symptoms. For patients who have only a partial clinical response to successive 10 to 12 week trials of various SRI/SSRIs, augmentation strategies may be useful. Of the numerous classes of medications tried as augmentation agents, only clonazepam and neuroleptic have proven superior to placebo in controlled studies in adults. A comorbid tic disorder and schizotypal personality disorder were associated with a positive response to haloperidol augmentation of fluvoxamine. Risperidone addition decreased OCD in adult OCD patients unresponsive to SRI treatment alone; response to risperidone was unrelated to presence or absence of comorbid tic disorder or schizotypal personality disorder. Hence, addition of a low dose of a neuroleptic may be useful especially in children with OCD and a comorbid tic disorder or schizotypal personality disorder. However, careful consideration is required prior to prescribing a neuroleptic in children. Concerns about neuroleptic use include potential side effects of cognitive impairment, sedation, dysphoria, weight gain, extrapyramidal symptoms (including acute and tardive dyskinesias), as well as the increased incidence of side effects with concomitant pharmacotherapeutic agents. In children, concerns about benzodiazepine augmentation include the need to avoid abrupt discontinuation, the potential for drug dependency, and the rare symptoms of disinhibition. The predictors of response are largely unknown. Individuals with comorbid or family history of tic disorder may not respond as well to the SSRIs.

Long-term Maintenance

The optimal time to maintain an individual who has responded to an SRI/SSRI on medication is unclear. Although periodic discontinuation trials are advisable, many responders require ongoing maintenance pharmacotherapy. In a double-blind discontinuation study, 89% of child and adolescent patients withdrawn from long-term clomipramine maintenance relapsed within 2 months. Comparable rates of relapse are observed in adults within 7 to 12 weeks of withdrawal from maintenance clomipramine or fluoxetine. These studies suggest that long-term maintenance may be required for some, although CBT may decrease the need for long-term pharmacotherapy. Many patients on continued SSRI or clomipramine maintenance continue to exhibit some OC symptoms, which may vary in severity over time. Abrupt discontinuation of clomipramine or those SSRIs with shorter half-lives (e.g., sertraline, paroxetine, and fluvoxamine) has been noted in adults to produce withdrawal syndromes consisting of gastrointestinal disturbance, headache, dizziness, malaise, and/or insomnia; hence gradual tapering is advisable.

INVESTIGATIONAL TREATMENTS

Investigational treatments, such as intravenous clomipramine, psychosurgery, and transcranial magnetic stimulation, have been studied in adults with OCD refractory to standard treatments. There are no data regarding the efficacy or safety of such treatments in children or adolescents. The identification of a subgroup of children who develop OCD and/or tic disorders after GABHS pharyngitis has led to investigational trials of prohpylactic antibiotic treatment and immunomodulatory interventions, including plasmapheresis and intravenous immunoglobulin. Double-blind controlled trials of penicillin prophylaxis for children whose OCD or tic disorder symptoms are precipitated or exacerbated by GABHS are ongoing. At present, the practical clinical implications of these studies are that when the onset (or exacerbation) of OCD (with or without associated tics) is abrupt and explosive, inquiry should be made regarding recent streptococcal infections and a throat culture for streptococcus and anti-streptolysin O and anti-streptococcal DNAse antibody testing should be considered. A positive streptococcal throat culture warrants standard antibiotic treatment; evidence of the onset or exacerbation of OCD being associated with streptococcal exposure warrants ongoing monitoring for recurrent streptococcal infection.

COORDINATING INTERVENTIONS

Treatment planning must begin with a comprehensive individualized assessment of the child's obsessive compulsive symptoms; the presence of co-morbid psychopathology; the child's developmental level, personality, and adaptive functioning; and family context. Mild obsessions or compulsions that are not the source of substantial distress or impairment may warrant monitoring without the initiation of specific treatment. Where these appear related to external or developmental stresses, psychotherapy or other psychosocial interventions targeted to these stresses may be useful. Psychotherapy also may be useful for OC personality traits.

INITIAL TREATMENT

If a child's OCD is unaccompanied by additional significant psychopathology or difficulties unrelated to the obsessions and compulsions, the initial treatment of choice is CBT or anti-obsessional medication, either alone or in combination. Where additional difficulties are limited to irritability and struggles with parents over the performance of compulsions, further interventions may not be required. In a large proportion of cases, however, where other significant psychopathology, developmental difficulties, or family problems accompany OCD, additional psychotherapeutic (individual and/or family), school, and/or pharmacological interventions are indicated from the outset. In the absence of direct comparative trials, clear empirical guidelines are lacking for determining whether to begin with CBT, an SRI/SSRI, or both. Many experts favor CBT as the initial treatment of choice for children and adolescents with OCD, especially milder cases without significant comorbidity. The choice of CBT as the initial treatment has the advantage of apparent durability and of avoiding the potential side effects of medication. Further, because the natural history of mild cases of OCD in younger children is unclear, this choice avoids committing the child to an uncertain period of medication. In specific cases, however, after consultation with parents, an anti-obsessional medication may be the initial choice of treatment because of the time, effort, expense, or anxiety associated with behavioral therapy or the unavailability of a behavioral therapist trained in the use of E/RP in children. Other indications for beginning with medication include the child's lack of sufficient cognitive or emotional maturity to cooperate in CBT (including the ability to view the symptoms as ego-dystonic, form an alliance, or tolerate anxiety during exposure) or the lack of family support for the treatment. The presence of co-morbid depression, anxiety disorder, or disruptive behavior disorder may also be an indication for including an SRI/SSRI as part of the initial treatment. The impact of CBT alone on co-morbid difficulties, such as irritability, depression, general anxiety, and impulsivity, remains largely unstudied. To the extent that some such difficulties reflect the burden of OCD symptoms on the child and family, effective cognitive behavioral treatment of the core OC symptoms might be expected to have broadly beneficial effects. In other cases, however, co-morbid symptoms may reflect the influence of neurobiological factors underlying the pathogenesis of both the core OCD symptoms and mood, anxiety, attentional, tic, and/or temperamental difficulties. Whether CBT is as effective as pharmacological interventions with these associated forms of psychopathology, which may warrant medication in their own right, remains to be determined. As various potential subtypes of OCD are elucidated (e.g. tic-related vs. non-tic-related; familial vs. non-familial; prepubertal vs. pubertal vs. adult onset; infection-triggered vs. non-infection related) it will be important to establish whether these subtypes also differ in their relative responsivity to CBT alone or in combination with medication. For example, compulsions that are driven by a premonitory urge or need to repeat an action until it feels "just right" (such as in children with tic disorder) may respond differently than those that serve to reduce anxiety and may benefit from habit reversal techniques using self-monitoring, relaxation, and competing motoric responses that have proven useful in habit disorders such as trichotillomania. Many clinicians believe that a combination of CBT and pharmacotherapy is optimal. When CBT is added to a stable medication regimen, the average magnitude of improvement in some studies is greater than that usually seen with medications alone. Some studies also suggest that CBT may reduce relapse rates in patients withdrawn from medication. Clear data are lacking, however, on whether CBT and medication are equally effective, whether their combined effects are synergistic, and what the effects of CBT are on relapse prevention. There appears to be tentative support, however, for combining both approaches, even in those patients for whom ongoing pharmacotherapy proves necessary, since cognitive-behavioral psychotherapy, including booster treatments during medication discontinuation, may improve both short- and long-term outcome in medication-responsive patients. Children whose OC symptoms are accompanied by other substantial psychopathology, developmental difficulties, and/or family problems usually require the corresponding appropriate individual and/or family psychotherapeutic, educational, and/or pharmacological interventions. Although family or social problems do not appear to cause OCD, improving family functioning and decreasing the child's non-OC adaptive difficulties at home and school also may have non-specific beneficial effects on the child's symptoms and treatment compliance. Where the child's OC symptoms or accompanying difficulties are interfering with the child's school participation, close collaboration with school personnel is required to monitor the child's symptoms and provide specific response guidelines and appropriate educational and curricular adjustments.

ONGOING TREATMENT

To monitor treatment response, it is necessary to assess systematically the child's symptoms using both parent- and self-reports. Ongoing evaluation is necessary to guide treatment planning and to indicate the need for modifications in treatment modality, dosage, or setting. Periodic re-evaluation for comorbid difficulties also is necessary, since the relative contribution of anxiety, depression, or attentional difficulties to the child's adaptive difficulties may become clear only as treatment of OCD symptoms progresses. Conversely, inadequate response to treatment may be explained in part by unrecognized co-morbid difficulties. For children who do not respond or have only a partial response to CBT alone, the next strategy is to alter the CBT technique, intensity, or setting and/or to add an SRI/SSRI. Technical modifications that may be introduced to facilitate E/RP include the use of positive reinforcers to encourage exposure; anxiety management techniques for coping with exposure-related anxiety; thought-stopping or cognitive restructuring techniques; modeling and shaping; and family interventions to improve compliance. Although CBT usually is implemented with 13 to 20 weekly individual or family visits in an office setting and E/RP homework assignments, partial or non-responders may require more frequent visits, out-of-office therapist-assisted E/RP, or even, in severe and unresponsive cases, daily visits in an office, partial hospital, or inpatient setting. Hospitalization may be necessary for safety when self-injurious or aggressive compulsions are prominent. Depending on the severity of symptoms, an SSI or SSRI should be added for compliant patients who have no response after 2 to 4 weeks to a well-delivered trial of CBT, or who have only a partial response after 4 to 7 weeks of CBT. A substantial number of children obtain no improvement or only partial relief of OCD symptoms on SRI/SSRIs alone, even after 10 to 12 weeks at maximum tolerated doses. In such cases, the next strategy is to add CBT, if this has not already been initiated. In addition, it may be useful to switch SRI/ SSRIs. If successive trials of two or three serotonin inhibitors (including clomipramine) are unsuccessful, consideration should be given to augmentation with clonazepam or a low dose of neuroleptic. Addition of a neuroleptic may be particularly useful in children with a comorbid tic disorder.

Duration of Pharmacotherapy

For children whose OCD responds satisfactorily to anti-obsessional medication with or without CBT, an important treatment decision concerns how long to maintain the patient on medication. The decision to maintain a child on anti-obsessional medication indefinitely should be made only if relapse occurs repeatedly following attempts to taper medication. Patients should be maintained on anti-obsessional medication for 12 to 18 months following a satisfactory response before attempting to discontinue medication. Once the decision is made to attempt a reduction or discontinuation of medication, the taper should be gradual, for example, decreasing the medication by 25% and observing for 2 months before making further reductions, depending on the patient's response. Withdrawal studies suggest that a majority of patients relapse within a few months following complete discontinuation of medication. Hence, trials of discontinuing medication should be timed for periods that are relatively low in stress and when potential symptom recurrence would be least disruptive. Abrupt discontinuation of TCAs and those SSRIs with shorter half-lives should be avoided, since it may produce a withdrawal syndrome such as gastrointestinal disturbance, sleep disturbance, or malaise. There is some evidence that CBT may reduce the need for long-term pharmacotherapy. For those patients who experience some symptom recrudescence following medication dose reduction, booster sessions of CBT or the introduction of CBT de novo (for those who have not previously received it) may reduce the need to return to previous medication levels. As with many of the recurrent or persistent psychiatric disorders of childhood, children and adolescents with OCD and their families are best served by continuity of care and the long-term availability of and monitoring by a knowledgeable clinician familiar with the child's unfolding developmental course and needs.

REFERENCES

American Academy of Child and Adolescent Psychiatry (1998), Practice Parameters for the Assessment and Treatment of Children and Adolescents with Obsessive-compulsive Disorder. J Am Acad Child Adolesc Psychiatry, 37(10suppl)

American Academy of Child and Adolescent Psychiatry (1997), Practice Parameters for the Psychiatric Assessment of Children and Adolescents. J Am Acad Child Adolesc Psychiatry, 36:4S-20S