As mentioned in the introductory article, cognitive-behavioral therapy (CBT) remains one of the most effective treatments for obsessive-compulsive disorder (OCD). In this article, we define CBT; review the evidence for the efficacy of CBT for OCD; provide a case example and sample treatment plan; and discuss family factors that affect treatment outcome, group, and family-based modalities to administer CBT and special applications of CBT to children with OCD.
Family and couple-based interventions
CBT and medication have been effective treatments of OCD; however, substantial dropout and refractory rates remain. Family members are often excluded from individually based CBT. Perhaps FA and EE are moderators for treatment outcome, and including family in order to target changing these responses could further improve symptom reduction and durability.
Family-based interventions for OCD have been more rigorously and recently investigated in pediatric samples (see below), although reports on the treatment of adults date back four decades. The APA practice guidelines indicate the importance of family factors in OCD treatment outcomes (
86). Lebowitz et al. (
72) searched PubMed and PsycINFO for studies on OCD and family-related terms between 1967 and 2011 and identified 641 articles.
Clearly, the importance of family in the course of the disorder and its treatment has been prominent but not that well developed in controlled studies [for a review, see Steketee and Van Noppen (
87); Renshaw et al. (
88)]. Perhaps the first group to recognize the benefits of including family in patients’ treatment was Marks et al. (
9), who initiated a monthly open-ended educational and support group.
Tynes et al. (
89), Black and Blum (
90), and Cooper (
91) used family psychoeducational groups in the treatment of OCD. In particular, patients with family involvement as a “co-therapist at home” experienced greater reduction in anxiety and improvement in social and occupational functioning (
92).
In an uncontrolled inpatient study, Thornicroft et al. (
93) reported on the use of a family component that focused on reducing relatives' involvement in rituals by training them to monitor patient behavior and encourage self-exposure in a noncritical manner. Findings indicated decreases in OCD symptoms at discharge, indicating good success for this severe inpatient population who scored in the extreme range on disability from OCD symptoms.
Grunes et al. (
94) examined the benefits of a relative’s participation in an 8-week psychoeducation group designed to help reduce accommodation to OCD symptoms. Patients whose family member was involved in the group experienced greater reductions in OCD symptoms and depressed mood compared with patients whose relatives did not participate (YBOCS score decrease of 26% versus 9%). Overvalued ideation, also known as poor insight (
95), and known to be a predictor of treatment outcome in OCD (poor insight), also was reduced to a greater degree in the family intervention group. Gains were maintained at a 1-month follow-up, suggesting that inclusion of family members in treatment is advantageous.
In an uncontrolled study of the effects of 10–12 weekly multifamily behavioral therapy (MFBT) treatments (
96), family members and patients received education about OCD and in vivo ERP. They also were taught family contracting for behavioral change to reduce FA, hostile criticism, and emotional overinvolvement during the family group sessions. YBOCS scores dropped significantly from pretreatment to posttreatment, and gains were maintained at 1-year follow-up [for an abbreviated manual, see Van Noppen (
97)]. MFBT offers distinct advantages by combining the therapeutic components of both family-based and group modalities. It may prove to be more cost effective than individual CBT as well.
In a 16-session couple-based CBT approach, Abramowitz et al. (
98,
99) targeted relationship dynamics to “enhance” the effectiveness of traditional gold-standard CBT. Similar to the therapeutic components in MFBT, this intervention includes partner-assisted ERP, targets changing maladaptive relationship patterns focal to OCD (such as accommodation), and targets non–OCD-related stressors. At posttest, significant improvements in OCD symptoms as measured by the YBOCS, relationship functioning, insight, and depression were reported and were maintained at 1-year follow-up (
99).
Controlled and replicated studies are needed, although findings support the efficacy of family and couple-based CBT for OCD and clinical wisdom suggests that including relatives can only further enhance ERP effects, maintain gains, and improve family functioning. In a recent meta-analysis (
100), “family inclusive treatment (FIT)” for OCD that targeted FA was associated with greater improvement in patient functioning. The investigators stated that the findings “underscore the need for continued momentum in the development, evaluation and dissemination” of family-based treatment of OCD.
Patient-only group behavioral therapy (GBT)
Delivered in a group modality, CBT allows for efficient use of therapist time and potential cost containment. In the 1970s, Hand et al. (
101), Teasdale et al. (
102), and Hafner and Marks (
103) reported, for anxious and phobic patients, not just a reduction in therapist time expenditures, but an increased benefit brought about by group processes such as cohesion and modeling (
104). In a group format, CBT was demonstrated to be effective in a meta-analysis (
105). A small meta-analysis found no difference in efficacy between group and individual formats, although dropout rates were much higher in the individual format in one study (
106).
The first systematic report of behavioral groups for OCD patients (
107) described an elaborate three-phase group process, and like Hand et al. (
101) and Teasdale et al. (
102), Jaurrieta et al. (
106) found that group cohesion provided extra benefits for OCD patients.
With the aim of providing cost-effective treatment that also improves patient compliance, Espie (
108) conducted a 10-session behavioral therapy and treatment focused on behavioral skills, homework, and follow-up group sessions scheduled after the end of intensive therapy. This study reported that the decrease in OCD symptoms was maintained at a 1-year follow-up. Krone et al. (
109) provided a structured GBT, with results that support the assertion that behavioral treatment groups for OCD provide a savings of therapist time and money and are possibly as efficacious as individual treatment.
Fals-Stewart et al. (
11) conducted a controlled trial for OCD, comparing group exposure (imaginal and/or in vivo) and response prevention with individual treatment and individual relaxation therapy (control). Participants in both group and individual treatment conditions showed significant improvement on measures of OCD symptoms, depression, and anxiety at posttest and follow-up, whereas the control condition produced change only on anxiety.
A naturalistic open trial of a time-limited behavioral group treatment with key elements of education, in vivo ERP, therapist and participant modeling, and cognitive restructuring is described by Van Noppen et al. (
110). After 10 sessions of a weekly 90-minute manualized GBT, average YBOCS score dropped from 21.8 to 16.6, similar findings to Krone et al. (
109) and Fals-Stewart et al. (
11). The authors describe distinct therapeutic advantages of group therapy, stemming from Yalom’s (
104) “curative factors of universality, imparting information, instillation of hope, imitative behavior, altruism,” and competition (not a curative factor but a motivator for patients treated in groups).
Benefits of GBT have gained international recognition. Results of GBT are durable and comparable to those of individual-based therapy (
111–
115), and GBT holds promising advantages over individually delivered CBT and may not only be preferable but also produce more robust outcomes for some difficult-to-treat patients. There is a need for further investigation using manualized controlled trials that consider medication effects with consistent study criteria and standard measures that allow comparison and generalizability.
A review of pediatric OCD: CBT, EE, and FA
OCD is estimated to affect approximately 2%−3% of children and adolescents (
116–
118). Pediatric OCD is characterized by a 3:2 male-to-female ratio, and in a National Institute of Mental Health sample, boys were more likely than girls to have an onset before puberty (
119). The mean age of onset of pediatric OCD ranges from 7.5 to 12.5 years, with an average of 2.5 years between age at onset and treatment (
120). Approximately 80% of adults with OCD have their onset during childhood (
118,
121), suggesting the importance of early intervention.
Generally, the types of obsessions and compulsions that children have are similar to those of adults; however, the content differs because it is more developmentally appropriate. Children’s obsessions often center on a fear of a catastrophic family event. The most common obsessions for children are contamination, sexual and somatic obsessions, and excessive scruples (e.g., feeling guilty over small transgressions, whereas the most common compulsions are washing, repeating, checking, and ordering (
122).
Pediatric OCD is associated with significant impairment in social and family functioning and academic difficulties that influence child development (
123,
124). The most common OCD-related problem reported by parents and children is concentrating on schoolwork, with completing homework the second most common (
123,
125). Storch et al. (
126) found greater rates of peer victimization in youth with OCD and noted that peer victimization was positively related to loneliness, child-reported depression, parent-reported internalizing and externalizing symptoms, and clinician-rated OCD severity.
Similar to OCD of adults, with pediatric OCD evidence from meta-analyses indicates that CBT is effective (
127–
129), even for children as young as 3 years of age (
130,
131). The 2004 Pediatric Obsessive-Compulsive Disorder Treatment Study (POTS) trial (
118), using a randomized comparison of CBT, SSRIs, their combination, and a placebo control group, concluded that pediatric OCD benefits from CBT alone or in combination with an SSRI. Further research has supported and extended these findings and deems CBT an effective treatment of pediatric OCD with respect to its safety and durability of response (
132–
136). The results from POTS II (2011) (
136) found that therapist-guided CBT-ERP is more effective than simply giving instructions on CBT (
136,
137), relaxation strategies (
138), or waitlist controls (
139). Kircanski and Peris (
140), in a pilot randomized controlled study of family-based CBT (FCBT) for youth, investigated the impact of variables such as self-reported distress during ERP and the quantity and quality of ERP tasks on treatment outcome. They found that when the variability of distress was greater (i.e., that the distress changes) during ERP tasks and when more than one obsessive-compulsive symptom was targeted, the treatment outcome was better, supporting the reduction of anxiety during ERP as a core ingredient of CBT.
An extension of the Nordic Long-Term OCD Treatment Study compared sertraline with continued CBT in a sample of children who were considered CBT nonresponders and found no difference between groups (
141). For the interested reader, Ung et al. (
142) outlined the optimal treatment of pediatric OCD, including a detailed description of the specific evidence for CBT and pharmacotherapy with SSRIs. Additionally, a section on pediatric acute-onset neuropsychiatric syndrome (PANS) is included.
PANS refers to a subgroup of children who experience OCD symptoms because of an infectious trigger, environmental factors, or other possible triggers (e.g., metabolic disorders) that affect the immune system and result in an inflammation of the child’s brain. Because of this inflammation, a sudden onset of severe OCD symptoms or severely restricted food intake occurs and does not follow the typical, more gradual onset of OCD symptoms seen in pediatric OCD. In addition, neuropsychiatric symptoms such as anxiety, mood disruptions, irritability, frequent urination, academic regressions, and more may be present (
143). The interested reader is referred to Chang et al. (
143) for a consensus statement (derived from the First PANS Consensus Conference) for a clinically distinct presentation of PANS criteria and for expert recommendations on evaluating PANS. These experts outlined the core components of a thorough diagnostic evaluation of suspected PANS cases.
Common assessment tools for pediatric OCD include the Children’s Yale-Brown Obsessive Compulsive Scale (CY-BOCS) (
144), the Children’s Florida Obsessive-Compulsive Inventory (C-FOCI) (
145), the Obsessive-Compulsive Inventory–Revised (OCI-R) (
146), and the Child OCD Impact Scale–Revised (COIS-R) (
147). However, because of space limitations, a review of assessment tools for pediatric OCD is beyond the scope of this article. Readers are referred to other sources for a more in-depth review of assessment measures (
132,
148–
150).
Kircanski et al. (
148) outlined numerous developmental considerations that complicate treatment when working with children, such as 1) challenges on the part of the child in being able to describe specific obsessions and their role in triggering compulsions; 2) less insight; 3) present orientation that makes future symptom relief less of a motivating factor for the child; 4) lower frustration tolerance and attention sustainability; and 5) poorer coping abilities. Child-oriented CBT packages address these issues by incorporating age-appropriate metaphors (e.g., false alarms, “bossing back” OCD), naming OCD, using fear thermometers, and creating incentive programs. The study by Geller and March (
132) provides practice parameters based on a thorough review of the literature, outlines best practices, discusses how treatment differs with children and adolescents with OCD, and provides references for therapists interested in mastering child-friendly techniques.
As with adult OCD, in pediatric OCD FA is a salient feature, with a large percentage of families engaging in modifying family routines and/or participating in a youth’s rituals (
58,
124,
138,
151–
153). FA is associated with poorer treatment outcome (
153,
154). The most common FA behaviors with pediatric OCD are providing reassurance, participating in rituals, and assisting in avoidance of feared situations or objects (
58). Research has found that FA is significantly related to parents’ reports of functional impairment at home, OCD symptom severity, and internalization and externalization of behaviors. Also, children reported greater impairment in school when families accommodated their symptoms. FA mediated the relationship between OCD symptom severity and parent-rated functional impairment of the child (
58). Hence, FA is associated with increased obsessive-compulsive symptom severity because it causes greater family dysfunction and negative family dynamics (
58,
124,
153,
154). For an in-depth review of FA, the interested reader is referred to Lebowitz et al. (
72).
Numerous studies have investigated FA in pediatric samples and noted the following results: child internalizing symptoms mediated the relationship between parent anxiety and FA (
151), family cohesion and externalizing symptoms predicted FA (
155), coercive/disruptive behaviors often significantly contributed to increased levels of FA (
156), and higher levels of accommodation were related to poorer results irrespective of modality (CBT or medication) (
157). These results are consistent with POTS findings (
118).
FA interferes with treatment goals of CBT-ERP by reinforcing the fear-and-avoidance behaviors; thus, FA is an important focus for intervention. Merlo et al. (
153) examined the relationship between FA and treatment outcome by providing FCBT (which directly targets FA) for 14 sessions. They found that participation in FCBT was associated with a significant reduction in FA and that FA reduction was related to less symptom severity and better treatment outcome. Poor insight has predicted greater FA and severity of symptoms, and has been associated with diminished CBT response in children and adolescents (
124,
158,
159). Low insight, higher levels of FA, and greater obsessive-compulsive symptom severity were linked to poorer in-session compliance and willingness to participate in treatment (
160). Adelman and Lebowitz (
161) outlined specific strategies to improve insight in pediatric OCD.
As with adults with OCD, when children have OCD, EE predicts poorer treatment outcome (
162,
163) and is prevalent among the parents. Also, high EE results in poorer overall child adjustment, including increased anxiety. Studies looking at maternal EE and how well children were functioning after treatment found that high baseline EE was a significant predictor of poorer treatment outcome (
164,
165). Studies have shown that mothers and fathers with OCD displayed higher rates of EE than controls (
166,
167). Przeworski et al. (
165) found that mothers had higher EE (driven by criticism rather than emotional overinvolvement) when describing a child with OCD (16%) compared to her child’s sibling without OCD (2.6%), indicating that EE may be state dependent. Peris et al. (
164) found that maternal criticism correlated with parental blame and the degree of responsibility of obsessive-compulsive symptoms attributed to the child but that maternal emotional overinvolvement correlated with parental anxiety, depression, and OCD. These studies underscore the high rate of blame, hostility, criticism, and emotional overinvolvement that correlates with OCD in children. Preliminary evidence that family emotional responses (e.g., conflict, blame, and low levels of support) are related to treatment outcome was found in a separate study by Peris et al. (
154), suggesting the importance of targeting EE in treatment.
In a thorough review (2007–2012) of the psychosocial treatment literature, Freeman et al. (
168), using an evidence-based, five-level evaluation criterion (see the article by Freeman et al. for definitions), evaluated the studies and classified the treatment methods as “well established,” “probably efficacious,” “possibly efficacious,” “experimental,” and “of questionable efficacy.” Using this rigorous methodological criterion, the authors reported that although no pediatric OCD treatment was designated as “well established,” both individually focused CBT and family-focused individual CBT (FCBT) are considered “probably efficacious” treatments. Group FCBT and group individual CBT, as well as alternative non–face-to-face deliveries (e.g., Internet), are now considered “possibly efficacious” treatments.
More recently, research has focused on investigating FCBT in a more systematic manner. Piacentini et al. (
138) contrasted FCBT with psychoeducation and relaxation training (PRT), finding faster and greater decrease in OCD severity and functional impairment for the FCBT compared with the PRT group. The POTS Jr study (
169), a multisite randomized controlled trial (RCT), found that FCBT was significantly more effective than relaxation therapy for young children (ages 5–8 years) with OCD. Consistent with these results, Lewin et al. (
131) in a pilot RCT found that children as young as 3 years of age responded better to treatment with family-based ERP (65%) compared with continuing with previous treatment, as outlined by their providers (7%).
Research has extended the efficacy of FCBT to intensive programs. Storch et al. (
58) compared intensive (daily for 3 weeks) and weekly (once per week) FCBT and found high remission rates for both groups (75% versus 50%, respectively). In an open trial of intensive FCBT, for children who were either medication partial responders or nonresponders, Storch et al. (
170) found that FCBT was successful, with 80% improved after treatment, including significant reductions in obsessive-compulsive–related impairment and FA. Rudy et al. (
171) found that 14 FCBT sessions at 3 weeks had 88.5% children as treatment responders. Whiteside and Jacobsen (
172) found that, even in a 5-day intensive program, FCBT significantly improved symptoms.
In a series of five case studies, Comer et al. (
173) found that after a full treatment course of Internet-delivered FCBT (including ERP), 60% of children no longer met the diagnosis for OCD. These results were consistent with the preliminary findings of Storch et al. (
174) with webcam-delivered FCBT. O’Leary et al. (
175) examined group versus individual FCBT and found a significant remission rate for both forms of FCBT (95% for group based and 79% for individual based) at 7-year follow-up. A pilot study by Farrell et al. (
176) in 2012 found that group FCBT was effective for youth with OCD and comorbid diagnosis.
In conclusion, exposure-based CBT, with the proper developmental considerations and adaptations, is highly effective as treatment of pediatric OCD. FA is prevalent in pediatric OCD and affects symptom severity, treatment outcomes, and functional impairment. The emotional climate in a home, including EE, also affects treatment outcome. Given these findings, treatment with children increasingly involves addressing these behaviors through the use of FCBT, and research increasingly supports its effectiveness. Future research should use the more methodologically rigorous RCT design in order to continue to address and compare the effectiveness of these various treatment modalities for pediatric OCD.