Pps923.indd

Clinical Note
Metacognitive Therapy versus Exposure
and Response Prevention for Pediatric
Obsessive-Compulsive Disorder
A Case Series with Randomized Allocation
Michael Simons a Silvia Schneider b Beate Herpertz-Dahlmann a a Department of Child and Adolescent Psychiatry, RWTH Aachen University, Aachen , Germany; b Clinical Child and Adolescent Psychology, University of Basel, Basel , Switzerland Key Words
chotherapeutic alternative to the well-established ERP in Child/adolescent obsessive-compulsive disorder ؒ the treatment of pediatric OCD. Further investigations into the effi cacy of MCT are necessary to answer ques- Cognitive/metacognitive therapy ؒ Narrative therapy tions as to the working mechanisms underlying therapy for OCD. Abstract
Background: Exposure with ritual prevention (ERP) is the
psychotherapeutic treatment of choice for pediatric ob-
Introduction
sessive-compulsive disorder (OCD). In the present study, a new treatment rationale – metacognitive therapy (MCT) Obsessive-compulsive disorder (OCD) is a debilitating for children – was developed and evaluated. Methods:
disorder with an estimated lifetime prevalence rate of Ten children and adolescents with OCD were randomly 2–3% [1] . Besides pharmacotherapy with (selective) sero- assigned to either MCT or ERP therapy condition. Pa- tonin reuptake inhibitors [2] , cognitive-behavioral treat- tients were assessed before and after treatment and at ment (CBT) is the psychotherapy of choice for this disor- the 3-month and 2-year follow-up by means of symptom der, both for adults and young people. The empirically severity interviews. Depressive symptoms were also as- validated method is exposure with ritual prevention sessed. Manualized treatment involved up to 20 sessions (ERP) accompanied by cognitive and – in children and on a weekly basis. Results: We found clinically and sta-
adolescents – by family interventions [3–5] . Until now, tistically signifi cant improvements in symptom severity there have been three controlled studies providing evi- after treatment. At the 3-month and 2-year follow-up, the dence for the effi cacy of CBT (ERP) for pediatric OCD. attained improvements during treatment were retained. De Haan et al. [6] found a slight superiority of ERP over Conclusions: Despite some methodological limitations,
pharmacotherapy (i.e., clomipramine), and Barrett et al. results showed that MCT proved to be a promising psy- [3] found no signifi cant differences between individual Department of Child and Adolescent Psychiatry RWTH Aachen University, Neuenhofer Weg 21 Tel. +49 241 808 8260, Fax +49 241 808 2601, E-Mail msimons@ukaachen.de and group cognitive-behavioral family treatment regard- the contents of thoughts are discussed and examined with ing effi cacy and durability of treatment gains. In a very regard to their truthfulness and probability [24] . In MCT, recent study, the combination of CBT and pharmaco- the focus does not lie on the content of obsessions and therapy (i.e., sertraline) proved to be superior to CBT intrusive thoughts, but on the appraisal and the manage-alone and to sertraline alone [5] . The effectiveness of ERP ment of these thoughts. These thoughts are normalized is limited by high rates of treatment rejection and drop- by simply accepting them. Furthermore, probability rat- outs [7] . This strongly indicates that there is a need for a ings (e.g. ‘How probable do you think it is that you could psychotherapeutic alternative to ERP. Furthermore, the contaminate your parents?’) are seldom useful in OCD effi cacy of ERP cannot be attributed solely to habituation patients. Although they often know that the risk is very [8] . In some cases, symptom reduction could be better low, they are not sure that they can take it. explained by cognitive changes or by changes in self-ef- In psychoeducation, the patient’s specifi c problem- fi cacy. This makes cognitive therapy a promising alterna- maintaining metacognitive appraisals and strategies are tive or additional treatment strategy [9] . to be discovered, while the therapist emphasizes the nor- New cognitive and metacognitive OCD models [10– mality of these processes [10] . Socratic dialogue, thought 12] led to new intervention techniques – at least in the control experiments, and behavioral experiments aim to treatment of adult patients. Until now, there have been change these metacognitive strategies and appraisals. only a few case studies [13, 14] and a case series [15] ap- Cottraux et al. [25] conducted one of the few studies plying these new interventions to children and adoles- that compare cognitive therapy and behavior therapy in adult OCD. In the present study, the effi cacy of MCT for children and adolescents with OCD was investigated. Ac- cording to the Task Force criteria of the American Psy- According to the cognitive model of Salkovskis and chological Association for the identifi cation of empiri- McGuire [16] and Wells’ metacognitive model [11, 12] , cally supported treatments [26] , MCT was tested against obsessional thoughts develop from normal intrusive the already established ERP treatment. Both treatment thoughts that are interpreted and dealt with in special strategies were protocol driven according to two different ways. OCD patients tend to confuse these thoughts with manuals written by the fi rst author [22] . It was hypothe- real actions, or events, or intentions (metacognitive mis- sized that both treatment strategies would be effective at interpretation) [17, 18] . They tend to make use of several post-treatment and would have lasting effects after 3 dysfunctional metacognitive processes, such as increased cognitive self-consciousness (‘too much thinking about thinking’) [19, 20] , thought suppression, and selective at-tention to further intrusive thoughts. Moreover, they often are not sure when to stop the ritual, and make use of dys- functional stop signals like ‘emotional reasoning’ (e.g. they Eleven children and adolescents (7 males and 4 females) with have to repeat the ritual until they ‘feel’ safe or clean). OCD, aged 8–17 years, participated in the study. All were treated These various metacognitive appraisals (e.g. thought- in a child and adolescent psychiatric outpatient setting and were action fusion) and metacognitive processes (thought con- assigned either to narrative ERP or to MCT by simple randomiza-tion. One (male) patient dropped out of MCT, as it demanded too trol strategies, selective attention) can be understood as much self-refl ection on his part. He was successfully treated with dysfunctional solutions which lead to further exacerba- ERP but excluded from further calculations. None of the remaining tions of the vicious cycle of OCD. Recently, Mather and 10 patients (n = 5 in each group) received pharmacotherapy against Cartwright-Hatton [21] have found these metacognitions to be good predictors of obsessive-compulsive symptoms All participants were diagnosed according to the DSM-IV/ICD- 10 diagnosis of OCD based on the German structured clinical in- in adolescents and proposed a more metacognitively en- terview ‘Kinder-DIPS’ [27] with the patients and their families and on the well-established clinical interview Children’s Yale-Brown Obsessive Compulsive Scale (CY-BOCS). Patients were excluded if they had a diagnosis of mental retardation, autism, psychosis, Metacognitive therapy (MCT) aims to change dys- and current treatment using pharmacotherapy for OCD. Five patients had one or more comorbid diagnoses, including functional metacognitive appraisals and strategies. The agoraphobia (n = 2), Tourette’s syndrome (n = 1), chronic tic dis- new term ‘metacognitive therapy’ [22, 23] marks the ma- order (n = 2), and attention defi cit hyperactivity disorder (ADHD) jor differences to standard cognitive therapy. In the latter, Table 1. Details of obsessive-compulsive psychopathology and metacognitive therapy suggestions for the fi ve MCT patients
Behavioral experiment: Use unfl avoured Thought-event fusion (‘thinking this thought increases the risk that it comes true’) Thought-imperative defusion: you cannot do everything that you think ofMetacognitive reframing (you could not suffer from these thoughts unless God is very important to you) ERP = Exposure with ritual prevention.
All instruments were administered by the therapists at pretreat- The primary outcome was a change in OCD symptom severity, ment, post-treatment, 3 months and 2 years after the completion measured with the CY-BOCS. The CY-BOCS is a clinician-rated of the therapy. At the 2-year follow-up, the clinical status regarding scale with two subscales (obsessions and compulsions). In each sub- OCD was assessed using a structured psychiatric interview for the scale, frequency, interference, distress, resistance (to obsessions/ parents and the child (Kinder-DIPS) [27] . At the time of the last compulsions), and control are rated. The total score ranges from 0 assessment, all interviews were performed by two experienced child to 40. Scahill et al. [28] found good reliability and validity for the and adolescent psychiatrists who were blinded to the patient’s treat- CY-BOCS. A recent study [29] found the two resistance items to be unreliable. This is in accordance with the metacognitive theory claiming that resistance to obsessions maintains obsessive-compul- All patients received up to 20 treatment sessions on a weekly The second outcome was a change in severity of depressive basis provided mainly by the fi rst author. Treatment was fi nished symptoms (self-rating), assessed using the German version of the before the twentieth sessions if according to the estimation of both Children’s Depression Inventory (CDI) [30] . The authors of the therapist and patient treatment goals were already met before. Par- German version [31] report good reliability and validity. The 26 ents were generally included in the therapy sessions except in cases, items are scored on a 3-point scale. In the present study, the stan- where the (older) adolescents did not permit it. There were no par- Metacognitive Therapy for Pediatric OCD Table 2. Demographic and clinical
1 Mann-Whitney U-Test.
Alpha level 0.05. Md = median; M = mean; SD = standard deviation; ERP = exposure with ritual prevention; MCT = metacognitive therapy; CY-BOCS = Children’s Yale-Brown Obsessive Scale; CDI = Child Depression Inventory (German version).
The ERP treatment was similar to the published manual by age, duration of complaints, IQ test scores, and the scores of the March and Mulle [32] in combining ERP with narrative therapy. CY-BOCS and CDI before treatment were compared using the In contrast to this manual, however, no anxiety management tech- Mann-Whitney U test. The CY-BOCS total score as the primary niques (such as relaxation training) were applied. In cases where outcome measure and the CDI T score were compared at pretreat- the motivation for ERP began to decrease, home-based contingen- ment, post-treatment, and follow-up by calculating the nonpara- metric Wilcoxon paired rank sum test for every single treatment MCT was based on techniques described by Salkovskis [10] and group. Using the Statistical Package for Social Sciences for Win- Wells [11, 12] , adapted by the fi rst author for children and adoles- dows (SPSS) version 11.0, all hypotheses were examined at an alpha cents. They were grossly comparable with already published cogni- level of 0.05. Furthermore, the percentage of improvement and the tive interventions for adolescents with OCD [13–15] . The main effect sizes were evaluated using Cohen’s d [33] . Due to the small differences between the two manuals were the interpretation of ob- sample size, the effect sizes were corrected as recommended by sessions and the presumed working mechanism: in ERP, obsessions were conceptualized as meaningless and metaphorically described as ‘brain hiccups’ [32] . Therapy was explained to work by habitu-ation. In MCT, obsessions were conceptualized as revealing the personal values and the deepest fears of the patient. Therapy aimed at challenging metacognitive appraisals (metacognitive restructur-ing) and changing metacognitive strategies (e.g. from suppressing Comparability of the Treatment Groups thoughts to permitting/accepting them). Following Wells [12] , in Both treatment groups were comparable regarding this metacognitive framework ERP could be utilized as behavioral their age, duration of complaints, IQ test scores, and de- experiment intended to challenge dysfunctional metacognitions, but not in an intensive, repeated, or habituation-oriented manner. pression scores ( table 2 ). The patients in the MCT condi- Accordingly, MCT patients were not encouraged to do ERP as tion had signifi cantly higher CY-BOCS scores before homework. Details of the 5 MCT patients and their treatment can Statistical Analysis Because of the small sample size only nonparametric tests were All patients showed a signifi cant decrease in OCD used. To examine the comparability of the two treatment groups, symptom severity from pre- to post-treatment assess- Fig. 1. CY-BOCS total scores for each pa-
tient treated with exposure and response
prevention at pre- and post-treatment and
at follow-up 3 months and 2 years after
treatment; range 0–40.
Fig. 2. CY-BOCS total scores for each pa-
tient treated with metacognitive therapy at
pre- and post-treatment and at follow-up 3
months and 2 years after treatment; range
0–40.
ments ( fi g. 1, 2 ). In the ERP condition, the CY-BOCS in the MCT group. Given that an improvement of more total scores decreased from 20 to 1 (z = –2.032, p = 0.042). than 30% is defi ned as success [35] , all 10 patients can be In the MCT condition, the CY-BOCS total scores de- considered as responders in both treatment conditions. creased from 26 to 6 (z = –2.032, p = 0.042). The corrected effect sizes on the CY-BOCS were 2.2 for In both groups, depression scores tended to decrease the ERP condition and 2.92 for the MCT condition. Ac- from pre- to post-treatment assessments, but the differ- cording to the criteria of Cohen [33] , both effect sizes were ences were not statistically signifi cant. CDI T scores de- creased from 50 to 38 (z = –1.512, p = 0.131) in ERP and from 53 to 41 (z = –1.461, p = 0.144) in the MCT condi- Treatment Effects after 3 Months tion. However, the scores were already in the average At the 3-month follow-up, treatment gains were main- tained. No signifi cant differences emerged in the CY- The average treatment duration was 13 sessions for BOCS scores when compared to post-treatment (ERP: ERP (mean = 13.0; SD = 6.04) and 9 sessions for MCT z = –0.535, p = 0.593; MCT: z = –0.677, p = 0.498). All (mean = 10.2; SD = 3.3). There was no signifi cant differ- but one (metacognitively treated) patient were fully re- ence in treatment duration between the two groups (U = At the 2-year follow-up, 4 out of 5 patients in the ERP Computing improvement rates across treatment in the condition and all 5 MCT patients could be assessed. Im- two groups showed a mean improvement in the CY- provements in the CY-BOCS were maintained, i.e., no BOCS total score of 89.6% for the ERP group and of 75% signifi cant differences emerged between post-treatment Metacognitive Therapy for Pediatric OCD and the 2-year follow-up (ERP: z = 0.000, p = 1.000; out of these 13 studies, some children and adolescents MCT: z = –0.816, p = 0.414). received concomitant serotonin reuptake inhibitors/se-lective serotonin reuptake inhibitors pharmacotherapy Diagnostic Status and Utilization of Therapy during without controlling for the specifi c effects of combination therapy. In the present study, no patient received con- Based on the structured interview (Kinder-DIPS), all comitant pharmacotherapy. Hence, improvements in 5 patients in the ERP condition and 4 out of 5 patients symptom severity can be attributed to the psychothera- in the MCT condition did not fulfi l the DSM-IV/ICD-10 peutic interventions. In addition, only a few studies com- prised a follow-up assessment and if they did, the dura- In the ERP group, 1 patient received 6 booster sessions tion of follow-up in previous studies ranged from 3 to 9 of ERP because of recurring OCD symptoms. At the 2-year months. In the present study, follow-up assessments were follow-up, he was fully recovered. Another one received 13 conducted after 3 months and after 2 years, thus demon- booster sessions because of recurring OCD symptoms ac- companied by motor tics and ADHD symptoms. At fol- Since MCT included a metacognitively modifi ed ver- low-up, he showed subclinical compulsive symptoms. sion of ERP one could argue that MCT actually works by In the MCT group, 1 patient obtained 2 MCT booster ERP. However, only 1 MCT patient received this kind of sessions because of recurring obsessions shortly after the ERP in only 1 session. Thus, it can be ruled out that MCT 3-month follow-up. At the 2-year follow-up, she was fully recovered. One further patient received 15 booster ses-sions and additional pharmacotherapy (sertraline 50 mg/ day) because of OCD symptoms and motor tics. At fol- The present study had several limitations. First, thera- low-up, he fulfi lled the DSM-IV/ICD-10 criteria of mild pists’ factors could not be varied (in 9 out of 10 cases the fi rst author was the therapist). Hence, it remains unclear to what degree therapist-specifi c versus treatment-specif-ic factors accounted for the therapy success. Further, Discussion
treatment integrity (i.e., the therapist’s adherence to the treatment protocol) was not assessed. To the authors’ knowledge, this is the fi rst study test- The small sample does not allow generalizations to ing MCT for childhood OCD and the fi rst one applying other children and adolescents with OCD, especially two different psychotherapeutic approaches. Both MCT those with a severer or chronic disorder, or to those of and ERP produced signifi cant and robust reductions in younger age. The youngest patient treated with MCT was obsessive-compulsive symptom severity. These effects 9 years old. It has yet to be clarifi ed at which age MCT were observed during a short time span (13 sessions of can be implemented into treatment. Similar to the pa-ERP and 9 sessions of MCT) and were still observed 2 tients in the previous studies, the patients in the present years after commencement of the therapy. All patients study suffered from moderate OCD. Further studies are treated with ERP and 4 out of 5 patients treated with required to demonstrate that patients with severer and MCT were (nearly) fully recovered 3 months and 2 years chronic OCD can benefi t from ERP and/or MCT. after therapy. Two patients with comorbid Tourette’s Because of the small sample size and higher symptom syndrome/ADHD and with tic disorder were in need of severity (prior to therapy) in the MCT condition, no be- more intensive therapeutic support after the actual ther- tween-group differences were analyzed. Larger-scale apy. This is in accordance with previous fi ndings of a se- comparisons between MCT and ERP are therefore war- verer long-term outcome of OCD in the presence of co- ranted to investigate differences in the effi cacy of the two Regarding depressive symptoms, the scores declined after therapy. The calculated differences failed to reach Clinical Implications and Future Directions statistical signifi cance, which can be attributed to the Notwithstanding these limitations, the present study advocates MCT to be a promising alternative to ERP. Treatment duration, response rate, and treatment ef- Thus, it challenges previous fi ndings and recommen- fects were comparable with previously reported studies dations suggesting that a successful psychotherapy for reviewed elsewhere [3, 38] for both ERP and MCT. In 9 children and adolescents with OCD has to be based on ERP. Williams et al. [15] pointed out that cognitive in- treated with a more cognitively oriented therapy. Our im- terventions – especially the normalizing of intrusive pression was that MCT seems to require a higher level of thoughts – can serve to enable ERP as a treatment strat- self-refl ection, whereas ERP demands a higher level of egy. Insofar, metacognitive interventions can be seen as emotion regulation (especially, coping with anxious arousal). In future studies, longer follow-up intervals Larger-scale randomized controlled trials examining should be conducted. Special attention has to be paid to the short- and long-term effects of MCT are warranted. children with comorbid tic disorder who may show a Furthermore, relative effi cacy, indications, and contrain- poorer outcome (i.e., a need for further interventions/ dications for MCT vs. ERP need to be investigated. Lee and Kwon [39] speculated that (adult) patients with reac- If MCT proves to be effi cacious in OCD treatment, tive obsessions (i.e., obsessions evoked by identifi able then habituation may not be the only working mechanism stimuli) might benefi t more from ERP, whereas patients underlying a successful therapy [40] . The effi cacy of MCT with autogenous obsessions (i.e., obsessions which enter may thus lead to further questions as to ‘what really works consciousness without identifi able stimuli) may be better References
1 Zohar AH: The epidemiology of obsessive- 10 Salkovskis PM: Understanding and treating 21 Mather A, Cartwright-Hatton S: Cognitive pre- compulsive disorder in children and adoles- obsessive-compulsive disorder. Behav Res dictors of obsessive compulsive symptoms in cents. Child Adolesc Psychiatr Clin N Am adolescence: a preliminary investigation. J 11 Wells A: Cognitive Therapy of Anxiety Disor- Clin Child Adolesc Psychol 2004;33:743– 2 Geller DA, Biederman J, Stewart SE, Mullin ders. A Practice and Conceptual Guide. Chi- B, Martin A, Spencer T, Faraone SV: Which 22 Simons M: Exposition mit Reaktionsverhinde- SSRI? A meta-analysis of pharmacotherapy 12 Wells A: Emotional Disorders and Metacogni- rung und metakognitive Therapie bei Kindern trials in pediatric obsessive-compulsive disor- tion. Innovative Cognitive Therapy. Chiches- und Jugendlichen mit Zwangsstörungen; the- der. Am J Psychiatry 2003;160:1919–1928. 3 Barrett P, Healy-Farrell L, March JS: Cogni- 13 Freeston MH: Cognitive-behavioural treat- 23 Wells A, Papageorgiou C: Metacognitive ther- tive-behavioral family treatment of childhood ment of a 14-year-old teenager with obsessive- apy for depressive rumination; in Papageor- obsessive-compulsive disorder: a controlled compulsive disorder. Behav Cogn Psychother giou C, Wells A (eds): Depressive Rumination. trial. J Am Acad Child Adolesc Psychiatry 14 Shafran R, Somers J: Treating adolescent ob- 24 Beck AT: Depression: Causes and Treatment. 4 Chambless DL, Ollendick TH: Empirically sessive-compulsive disorder: applications of Philadelphia, University of Pennsylvania supported psychological interventions: contro- the cognitive theory. Behav Res Ther 1998;36: versies and evidence. Annu Rev Psychol 2001; 25 Cottraux J, Note I, Yao SN, Lafont S, Note B, 15 Williams TI, Salkovskis PM, Forrester EA, All- 5 The Pediatric OCD Treatment Study (POTS) Team: Cognitive-behavior therapy, sertraline, appraisal of responsibility during cognitive be- trolled trial of cognitive therapy versus inten- and their combination for children and adoles- havioural treatment: a pilot study. Behav Cogn sive behavior therapy in obsessive compulsive cents with obsessive-compulsive disorder: the Pediatric OCD Treatment Study (POTS) ran- 16 Salkovskis PM, McGuire J: Cognitive-behav- ioural theory of OCD; in Menzies RG, de Silva 26 Chambless DL, Hollon SD: Defi ning empiri- P (eds): Obsessive-Compulsive Disorder. The- cally supported therapies. J Consult Clin Psy- 6 de Haan E, Hoogduin KA, Buitelaar JK, Keijs- ory, Research and Treatment. Chichester, ers GP: Behavior therapy versus clomipramine 27 Unnewehr S, Schneider S, Margraf J: Diag- for the treatment of obsessive-compulsive dis- 17 Purdon C, Clark DA: Meta-cognition and ob- nostisches Interview bei psychischen Störun- order in children and adolescents. J Am Acad sessions. Clin Psychol Psychother 1999;6:102– Child Adolesc Psychiatry 1998;37:1022–1029. 7 Foa EB, Franklin ME, Kozak MJ: Psychosocial 18 Rachman S, Shafran R: Cognitive distortions: 28 Scahill L, Riddle MA, McSwiggin-Hardin M, treatments for obsessive-compulsive disorder; thought-action fusion. Clin Psychol Psycho- Ort SI, King RA, Goodman WK, Cicchetti D, Leckman JF: Children’s Yale-Brown Obses- Richter MA (eds): Obsessive-Compulsive Dis- 19 Cartwright-Hatton S, Wells A: Beliefs about sive Compulsive Scale: reliability and validity. order. Theory, Research and Treatment. New worry and intrusions: the Meta-Cognitions J Am Acad Child Adolesc Psychiatry 1997;36: Questionnaire and its correlates. J Anxiety 8 Marks I, Dar R: Fear reduction by psychother- 29 McKay D, Piacentini J, Greisberg S, Graae F, apies. Recent fi ndings, future directions. Br J 20 Janeck AS, Calamari JE, Riemann BC, Hef- Jaffer M, Miller J, Neziroglu F, Yaryura-To- felfi nger SK: Too much thinking about think- bias JA: The Children’s Yale-Brown Obses- 9 Bouvard MA, Milliery M, Cottraux J: Manage- ing? Metacognitive differences in obsessive- sive-Compulsive Scale: item structure in an ment of obsessive compulsive disorder. Psy- compulsive disorder. J Anxiety Disord 2003; outpatient setting. Psychol Assess 2003;15: Metacognitive Therapy for Pediatric OCD 30 Kovacs M: The Children’s Depression Inven- 36 Stewart SE, Geller DA, Jenike M, Pauls D, 39 Lee HJ, Kwon SM: Two different types of ob- tory (CDI). Psychopharmacol Bull 1985;21: Shaw D, Mullin B, Faraone SV: Long-term out- session: autogenous obsessions and reactive come of pediatric obsessive-compulsive disor- obsessions. Behav Res Ther 2003;41:11–29. 31 Stiensmeier-Pelster J, Schürmann M, Duda K: der: a meta-analysis and qualitative review of 40 Rachman S, Shafran R: The mechanisms of Depressionsinventar für Kinder und Jugendli- the literature. Acta Psychiatr Scand 2004;110: behavioral treatment and the problem of ther- che (DIKJ), ed 2. Göttingen, Hogrefe, 2000. apeutic failures; in Goodman WK, Rudorfer 32 March JS, Mulle K: OCD in Children and Ad- 37 Wewetzer C, Jans T, Muller B, Neudorfl A, Bu- MV, Maser JD (eds): Obsessive-Compulsive olescents. A Cognitive-Behavioral Treatment cherl U, Remschmidt H, Warnke A, Herpertz- Disorder. Contemporary Issues in Treatment. Dahlmann B: Long-term outcome and progno- 33 Cohen J: Statistical Power Analysis for the Be- sis of obsessive-compulsive disorder with onset 41 Hubble MA, Duncan BL, Miller SD: The havioral Sciences, ed 2. Hillsdale, Erlbaum, in childhood or adolescence. Eur Child Ado- Therapy. Washington, American Psychologi- 34 Hedges LV, Olkin I: Statistical Methods for 38 Simons M, Herpertz-Dahlmann B: Psycho- Meta-Analysis. San Diego, Academic Press, therapy of compulsive disorder in children and 42 Kazdin AE, Nock MK: Delineating mecha- adolescents – An overview. Z Kinder Jugend- nisms of change in child and adolescent thera- 35 Foa EB, Grayson JB, Steketee GS, Doppelt py: methodological issues and research recom- HG, Turner RM, Latimer PR: Success and fail- mendations. J Child Psychol Psychiatry 2003; ure in the behavioral treatment of obsessive- compulsives. J Consult Clin Psychol 1983;51:

Source: http://www.kli.psy.ruhr-uni-bochum.de/kkjp/team/public/Schneider%20Journals/2006/Simons%20et%20al._Metacognitive%20Therapy%20vs.%20Exposure.pdf