Received: 12 June 2018 Revised: 12 November 2018 Accepted: 12 November 2018

DOI: 10.1002/cpp.2345

COMPREHENSIVE REVIEW

Effectiveness of metacognitive interventions for mental

disorders in adults—A systematic review and meta‐analysis
(METACOG)
Rebecca Philipp1 | Levente Kriston1 | Jana Lanio1 | Franziska Kühne2 |

Martin Härter1 | Steffen Moritz3 | Ramona Meister1

1
Department of Medical Psychology,
University Medical Center Hamburg‐ Abstract
Eppendorf, Hamburg, Germany We evaluated the effectiveness and acceptability of metacognitive interventions for
2
Department of Psychology, Clinical
mental disorders. We searched electronic databases and included randomized and
Psychology and Psychotherapy, University of
Potsdam, Potsdam, Germany nonrandomized controlled trials comparing metacognitive interventions with other
3
Department of Psychiatry and treatments in adults with mental disorders. Primary effectiveness and acceptability
Psychotherapy, University Medical Center
Hamburg‐Eppendorf, Hamburg, Germany
outcomes were symptom severity and dropout, respectively. We performed
Correspondence random‐effects meta‐analyses. We identified Metacognitive Training (MCTrain),
Rebecca Philipp, Department of Medical Metacognitive Therapy (MCTherap), and Metacognition Reflection and Insight
Psychology, University Medical Center
Hamburg‐Eppendorf, Martinistrasse 52, Therapy (MERIT). We included 49 trials with 2,609 patients. In patients with schizo-
Hamburg 20246, Germany. phrenia, MCTrain was more effective than a psychological treatment (cognitive reme-
Email: r.philipp@uke.de
diation, SMD = −0.39). It bordered significance when compared with standard or
Funding information
German Federal Ministry of Education and other psychological treatments. In a post hoc analysis, across all studies, the pooled
Research, Grant/Award Number: 01KG1511
effect was significant (SMD = −0.31). MCTrain was more effective than standard
treatment in patients with obsessive–compulsive disorder (SMD = −0.40). MCTherap
was more effective than a waitlist in patients with depression (SMD = −2.80), post-
traumatic stress disorder (SMD = −2.36), and psychological treatments (cognitive–
behavioural) in patients with anxiety (SMD = −0.46). In patients with depression,
MCTherap was not superior to psychological treatment (cognitive–behavioural). For
MERIT, the database was too small to allow solid conclusions. Acceptability of
metacognitive interventions among patients was high on average. Methodological
quality was mostly unclear or moderate. Metacognitive interventions are likely to be
effective in alleviating symptom severity in mental disorders. Although their add‐on
value against existing psychological interventions awaits to be established, potential
advantages are their low threshold and economy.

KEY W ORDS

mental disorders, meta‐analysis, metacognition, psychotherapy, systematic review

1 | I N T RO D U CT I O N the shortcomings of more traditional psychotherapies (Hamm,

Hasson‐Ohayon, Kukla, & Lysaker, 2013; Keller & Boland, 1998;
In the last 20 years, the concept of metacognition, a term first intro- Ludvik & Boschen, 2015; Richter, 1999; Wells & Purdon, 1999). Some
duced by John H. Flavell in the 1970s (Flavell, 1979), was used to authors understand metacognitive interventions as an extension of
develop new psychological interventions in order to address some of cognitive–behavioural therapy (CBT; Hofmann & Asmundson, 2008;

Clin Psychol Psychother. 2019;26:227–240. wileyonlinelibrary.com/journal/cpp © 2018 John Wiley & Sons, Ltd. 227
228 PHILIPP R. ET AL.

Moritz et al., 2014), whereas others classify them as third wave

therapies (Gordon‐King, Schweitzer, & Dimaggio, 2018; Hunot et al., Key Practitioner Message
2013) or use metacognition in the context of integrative psychother-
apies (Dimaggio et al., 2017; Inchausti et al., 2017; Lysaker et al., • Our meta‐analyses suggest that Metacognitive Training

2011). In general, metacognitive interventions include specific thera- and Metacognitive Therapy are likely to be effective in

peutic elements that target patients' “knowledge and cognition alleviating symptom severity in mental disorders.

about cognitive phenomena” (Flavell, 1979). Still, there is a variety of • Metacognitive Training and Metacognitive Therapy
publications in which interventions are called “metacognitive” but are accessible interventions that can easily be adapted
which differ in their definition of the term. While working on this sys- to various clinical settings.
tematic review, we developed a working definition of metacognitive • Large and independent multicentre trials investigating
interventions. We define them as “treatments that explicitly target short‐ and long‐term effects that are relevant to
metacognitive content—characterized by the awareness and under- patients are needed to strengthen the evidence base.
standing of one's thoughts and feelings as well as the thoughts and
feelings of others—as the key element.” Also, they are goal‐oriented
and aim to alleviate disorder‐specific and individual symptoms by whether metacognitive interventions are effective. Second, we
specifically enhancing metacognitive capacities in order to gain more investigate whether effectiveness within these interventions varies
flexibility in the attention, monitoring, control, and regulation of across mental disorders. Third, we explore the acceptability of
cognitive processes. According to this definition, we included three different metacognitive interventions.
metacognitive interventions in our systematic review: Metacognitive
Training (MCTrain), first developed by Steffen Moritz and Todd
Woodward for patients with schizophrenia (Moritz & Woodward, 2 | METHODS
2007); Metacognitive Therapy (MCTherap), first developed by Adrian
Wells and Gerald Matthews for patients with generalized anxiety dis- This review was registered with the PROSPERO international
order (GAD; Wells & Matthews, 1994); and metacognitively oriented prospective register of systematic reviews (CRD42016051006).
integrative psychotherapies that are based on a narrative approach A detailed review protocol has been published in an open access
and were developed for patients with personality disorders and journal (Kühne et al., 2017). For deviations from the protocol and fur-
schizophrenia (Dimaggio & Semerari, 2001; Lysaker & Lysaker, 2001; ther specifications, see (Section S1 of the supplement). We conducted
Semerari et al., 2003). In this review, the latter group of interventions this systematic review and the meta‐analyses in accordance with
is represented by Metacognition Reflection and Insight Therapy (MERIT) current guidelines (Moher, Liberati, Tetzlaff, Altman, & The PRSIMA
as introduced by Lysaker and Klion (2017), because to date, there are Group, 2009; Shea et al., 2007; The Cochrane Collaboration, 2009).
no results of randomized controlled trials (RCTs) or non‐RCTs (NRCTs)
available for other conceptualizations like Metacognitive Interpersonal
Therapy (Dimaggio et al., 2017). Despite a number of differences, all 2.1 | Eligibility criteria
three metacognitive interventions share the assumption of a metalevel
of cognition that affects emotions and behaviour through giving We included RCTs and NRCTs that were conducted in adults
attention to and reflecting on thoughts and beliefs (for a review, see (≥18 years) with any mental disorder. Diagnoses needed to be based
Moritz & Lysaker, 2018). on a formal classification (e.g., ICD, World Health Organization,
Metacognitive interventions were disseminated for a variety of 1992; DSM, American Psychiatric Association, 2000) or on reliable
mental disorders. Their evidence base is constantly growing, and evalu- and validated disorder‐specific questionnaires. We included studies
ation studies report improved psychological symptoms. As described in regardless of patients' co‐morbidity (including any physical disorder)
our review protocol (Kühne et al., 2017), previous narrative and system- and treatment setting.
atic reviews conclude positive effects for MCTherap (Normann, van We only included studies that investigated metacognitive inter-
Emmerik, & Morina, 2014; Wells, 2013) and MCTrain (Eichner & Berna, ventions meeting our working definition. Comparators were other
2016; Liu, Tang, Hung, Tsai, & Lin, 2018; Moritz et al., 2014), but their specific active treatments (psychological, pharmacological, or com-
results are limited by methodological shortcomings, especially in regard bined psychological and pharmacological treatment) and nonactive
to the search and selection of the primary studies, the investigated treatments (e.g., standard treatment, placebo, and waitlist). We
mental disorders, and the systematic evaluation of quality of evidence defined psychological treatment as any form of treatment that uses
and risk of bias. Also, meta‐analyses report inconsistent findings psychological methods to alleviate symptoms. Thus, they include com-
(Jiang, Zhang, Zhu, Li, & Li, 2015; van Oosterhout et al., 2015). Thus, prehensive psychotherapeutic treatments that are based on scientific
a comprehensive and methodologically sound systematic review theories and consider patients' personal needs as well as single psy-
that covers the existing evidence including RCTs and NRCTs of chological techniques like psychoeducation or relaxation techniques,
metacognitive interventions in different mental disorders is needed. supportive treatments, and treatments that foster cognitive function-
In this systematic review, we aim to assess the effects of ing. We defined standard treatment as inpatient or outpatient
metacognitive interventions for adult patients with mental disorders. treatments including pharmacotherapy, contacts to case workers,
Therefore, the purpose of this study is threefold. First, we investigate other psychosocial support, or occupational therapy.
PHILIPP R. ET AL. 229

The primary effectiveness outcome was symptom severity at the 2.5 | Assessment of methodological quality
end of intervention measured with a disorder‐specific questionnaire
or symptom rating scale. Secondary effectiveness outcomes were We used Cochrane's tool (The Cochrane Collaboration, 2009) for
treatment response, improvement in overall symptomatology, changes assessing the risk of bias of the included RCTs and the ROBINS‐I tool
in metacognitive processes, satisfaction with treatment, and quality of (Sterne et al., 2016) for assessing the risk of bias in the included
life. In a patient involvement workshop, applicability of metacognitive NRCTs. Two out of five reviewers (R. P., F. K., R. M., and two more
interventions, autonomy, self‐perception, empowerment, and emotion scientific employees) independently judged risk of bias. If we
regulation were identified as further secondary effectiveness out- disagreed on the methodological quality, we discussed criteria until
comes mainly relevant to patients (Brütt et al., 2017). The primary we reached a consensus or consulted a third reviewer. In case a report
acceptability outcome was the number of patients who dropped out missed data for adequate judgement of risk of bias, we searched the
of the treatment due to any reason. Secondary acceptability outcomes associated study protocol or trial registration or contacted the
were the number of patients with treatment‐related adverse events. corresponding author.

2.2 | Search 2.6 | Data synthesis

We conducted an electronic database search in MEDLINE, ISI Web For the primary effectiveness outcome symptom severity, we ranked
of Science, BIOSIS, CINAHL, PsycINFO, and the Cochrane Central the administered scales for each disorder according to psychometric
Register of Controlled Trials (CENTRAL) on April 28, 2017, and an criteria and frequency of application. We preferred observer‐rated
updated MEDLINE search on March 7, 2018. We searched clinical trial outcomes over patient‐reported outcomes as they are more likely to
registries through the World Health Organization's trials portal be blinded (for details, see Section S4). When studies reported data
(ICRTP) and ClinicalTrials.gov and then contacted the principle investi- for more than one time of measurement, we extracted all available
gators of unpublished and ongoing trials. We checked reference lists data but only synthesized data for the time of primary measurement,
of the included studies as well as other systematic reviews and were which was the end of intervention in all of the studies.
in contact with key authors of metacognitive interventions (Adrian The secondary effectiveness outcome response rate was defined
Wells and Steffen Moritz) for more information regarding published depending on the mental disorder investigated in the study. We
and unpublished studies. To identify grey literature, we searched used a minimum decrease of 30% compared with the score at base-
ProQuest Dissertations, Open Grey, and Google Scholar. For the line for positive symptoms of schizophrenia (Howes et al., 2017); of
complete search strategies, see Section S2. 35% for obsessive–compulsive disorder (OCD; Farris, McLean, van
Meter, Simpson, & Foa, 2013; Lewin et al., 2011); and of 50% for
depression (Keller et al., 2000), anxiety (Loerinc et al., 2015), and
2.3 | Study selection posttraumatic stress disorder (PTSD; Bryant et al., 2008). If response
rates were not reported, we estimated the number of responders
One reviewer (R. P.) screened titles and abstracts of all identified according to Suissa's formula (Meister, von Wolff, & Kriston, 2015;
studies to identify potentially eligible studies. Two out of three Suissa, 1991).
reviewers (R. P., F. K., and R. M.) independently screened the full texts For continuous outcomes, we summarized the outcomes by
of these studies for inclusion. If we found studies to be ineligible, we calculating standardized mean differences (SMD) for studies that
documented the reasons for exclusion. In case we disagreed on the utilized different questionnaires or scales. For dichotomous outcomes
eligibility of a study, we discussed criteria until we reached a consen- (response and dropout rates), we calculated odds ratios with corre-
sus or consulted a third reviewer. If there was more than one report sponding 95% confidence intervals. We calculated odds ratios on the
for a study, we subsumed them because the units of interest were basis of the intention‐to‐treat sample as defined by the authors.
the studies rather than the reports (Table S5.1). However, for the calculation of SMD, we needed to use the sample
size reported by the authors. We calculated and reported SMD,
when the sample size was ≥5 in each group. We combined studies
2.4 | Data collection process and data items for meta‐analyses, when at least two studies reported data for the
same comparison and outcome.
Two out of five reviewers (R. P., F. K., R. M., and two scientific We conducted separate meta‐analyses for the different types of
employees) independently extracted study characteristics including metacognitive interventions and mental disorders. Also, we only com-
intervention characteristics, sample characteristics, metacognitive pared metacognitive interventions that were similar with regard to
intervention and comparators, and outcome data using a structured therapist guidance (e.g., full psychotherapy, major or minor therapist
Microsoft Excel sheet (Section S3). In case the extracted data differed, support, and unguided) or delivery mode (e.g., face to face and online).
we reached consensus through discussion or a third reviewer. If We calculated meta‐analyses using a random‐effects model, because
outcome data or study characteristics were unclear or not reported, we assumed that included studies will show considerable heterogene-
we contacted the corresponding author, documented correspondence, ity (Kriston, 2013). We tested statistical heterogeneity between study
and marked the added data. results using Cochran's Q test and the I2 statistic (Higgins, Thompson,
230 PHILIPP R. ET AL.

Deeks, & Altman, 2003). To test for possible reporting bias and small‐ 3.2 | Description of included studies
study effects, we used Egger's test (Egger, Smith, Schneider, & Minder,
1997) and examined funnel plots visually. We did not perform sub- 3.2.1 | Study characteristics
group analyses in case of categorical predictors or meta‐regression
analyses in case of metric predictors (Section S1). We performed We included 44 RCTs and five NRCTs that were published between

sensitivity analyses excluding studies without randomization (NRCTs). 2007 and 2018 (Table S5.1). More than half of the studies were

Results were contrasted to those acquired with data from all conducted in European countries (n = 31), followed by Iran (n = 8),

studies in order to control for possible effects of study design on Australia and New Zealand (n = 3), China (n = 2), India (n = 2), the

pooled effects. United States (n = 2), and Indonesia (n = 1). Six RCTs were conducted
as three‐arm studies.

3.2.2 | Sample and intervention characteristics

3 | RESULTS
The overall number of included patients was 2,609 and ranged from
3.1 | Study selection four to 154, with an average age ranging from 24 to 46 years.
Although gender distribution varied between 0% and 96% of females,
Altogether, we screened 4,404 records. Figure 1 shows the flow dia- Table S5.1 shows that only 21 of the 49 studies had more than 50% of
gram, including reasons for exclusion of full‐text articles. Additional female participants. Evidence for each of the metacognitive interven-
information on excluded studies and identified ongoing studies are tions was primarily reported for the mental disorders which the inter-
listed in the Table S5.2. We included 49 primary studies reported in ventions were originally developed for. Twenty‐eight studies tested
58 publications in our review. A total of 39 studies with 2,179 patients the effectiveness of MCTrain, most of them in patients with schizo-
provided data for quantitative analyses. phrenia (21/28). MCTherap was tested in 19 of the 49 studies, mostly

FIGURE 1 Flow diagram of identified and included records

PHILIPP R. ET AL. 231

in patients with depression (7/19). MERIT was tested in two studies in 3.3.1 | MCTrain
patients with schizophrenia. Treatment was conducted in individual
(27/49), group (20/49), or mixed sessions (2/49) and lasted between We conducted meta‐analyses for 23 of the 28 studies that investi-
2 and 52 weeks in inpatient and outpatient settings. Comparators gated MCTrain in addition to standard treatment or psychological
were other psychological treatments (n = 24), standard treatment treatment. Standard treatment always included pharmacotherapy.
(n = 12), waitlist (n = 11), pharmacotherapy alone (n = 3), or combined For patients with schizophrenia, we calculated meta‐analyses
treatment (n = 2). for 15 RCTs and four NRCTs that compared MCTrain with standard
treatment or with psychological treatments (Supportive Therapy and
Psychoeducation, Newspaper Discussion, and Cognitive Remediation
Tasks) in a total sample of 1,127 patients. MCTrain was statistically
3.2.3 | Outcome data
more effective when compared with Cognitive Remediation Tasks

Most studies reported symptom severity based on rating scales at but only bordered significance when compared with standard treat-

the end of treatment (45/49). Response rates, however, were only ment or other psychological treatments (Table 1). In a post hoc analy-

reported in nine studies and needed to be estimated in the majority sis, we pooled data across all studies to make our meta‐analysis

of studies (34 studies). Because synthesized results on response comparable with earlier meta‐analyses (Eichner & Berna, 2016). For

rates may be of reduced reliability, we only report them in Sections this analysis, patients in the MCTrain groups reported significantly less

S6 and S7. In regard to secondary outcomes, overall symptomatology symptoms on average than those in the control groups (Figure 2).

was reported in 13/49 studies, changes in metacognitive processes These results corresponded with responder and sensitivity analyses

in 14/49 studies, and quality of life in 10/49 studies (Table S5.1). (Figures S6.1 and S6.2, respectively). The included studies were

Moreover, the outcomes identified as clinically relevant in the clinically heterogeneous with regard to the control groups, patient

patient involvement workshop (Brütt et al., 2017) were not reported characteristics, and the type of MCTrain that was tested (see Table

in any of the studies. As for acceptability, five studies did not S5.1 for detailed characteristics). Accordingly, statistical heterogeneity

report any data on the number of patients that dropped out of the was substantial across studies. Fewer patients dropped out of the

treatment. Most of the studies (45/49) did not report treatment‐ MCTrain groups than out of the groups that performed cognitive

related adverse events. The remaining four studies reported no remediation tasks. Dropout rates between the MCTrain groups

adverse events or effects. and the other control groups did not differ (Figure S6.3). When
we examined the funnel plots visually, we found no indication for
publication bias (Figure S6.4). Accordingly, Egger's test for publication
bias was not significant (p = 0.14).
3.2.4 | Methodological quality
For patients with OCD, MCTrain was developed as an unguided
online self‐help intervention (myMCT). Three studies with a total
In 26 out of 44 RCTs, the sequence for random allocation was gener-
sample of 245 patients (Moritz et al., 2016; Moritz et al., 2018;
ated adequately. The allocation of patients to the study arms was ade-
Moritz, Jelinek, Hauschildt, & Naber, 2010) tested the effectiveness
quately concealed in half of the studies (21/44). Because we
of myMCT against standard treatment alone. Patients in the myMCT
investigated psychotherapy studies, it was not possible to blind
groups showed significantly less severe obsessive–compulsive
patients or therapists adequately in any of them. Blinding was adequate
symptoms than those in the control groups (Table 1). Groups did not
for assessment of the primary outcome in more than half of the studies
differ in the number of patients who responded to the treatment
(23/44), and almost half (22/44) reported outcome data completely.
(Figure S6.5). There was no substantial statistical heterogeneity
Twelve out of 44 studies were registered prior to start and reported
(Figure 2). Although treatment dropout was not determined for these
the outcome data as planned. In half of the studies, it was ensured that
online interventions, two studies specified that 14% to 21% of the
the intervention was implemented as conceptualized (22/44) and that
patients did not read the self‐help manual (Moritz et al., 2010; Moritz
patients in all study arms were attended to equally (26/44). For 12
et al., 2018). Compared with the control groups, fewer patients in
studies, we rated that the authors had no conflict of interest. Table
the myMCT group completed postassessment (Figure S6.6). One study
S5.3 shows the results of the methodological quality assessment of
with a sample of 128 patients (Hauschildt, Schröder, & Moritz, 2016)
the RCTs in more detail. The results of the methodological quality
compared myMCT with a psychological treatment (psychoeducation).
assessment for the five included NRCTs are presented in the Table
Groups did not differ in symptom severity (Table 1). Of the 90% of
S5.4.
the patients who completed postassessment in both groups, half of
the patients stated that they did not read the myMCT manual
thoroughly. A post hoc analysis, across all four studies, showed
3.3 | Quantitative analyses that patients receiving myMCT reported significantly less symptoms
on average than those in the control groups (Table 1). The number of
Table 1 shows the results of the quantitative analyses. Figures 2–4 studies was too small to make a conclusive statement on publication
show the forest plots for the primary outcome of symptom severity. bias (Figure S6.7).
Additional forest plots and funnel plots are presented in Sections S6 MCTrain for patients with depression was tested in one study
and S7. with a sample of 84 patients (Jelinek et al., 2016). Patients in the
232 PHILIPP R. ET AL.

TABLE 1 Results of quantitative analyses

No. of Data available for Heterogeneity Significance of

Comparison studies % of ITT sample SMD 95% CI (I2, p for Q statistic) treatment effect

MCT Control
MCTrain for positive symptoms in schizophrenia
Standard treatment 11 88% 93% −0.27 −0.59 to 0.05 60.7%, p = 0.00 z = −1.68, p = 0.09
Psychological treatment: 2 73% 73% −0.28 −0.81 to 0.25 57.3%, p = 0.13 z = −1.03, p = 0.31
Supportive
Therapy and Psychoeducation
Psychological treatment: 2 85% 89% −0.41 −1.00 to 0.18 35.9%, p = 0.21 z = −1.36, p = 0.17
Newspaper Discussion
Psychological treatment: 4 95% 88% −0.39 −0.67 to −0.10 35.2%, p = 0.23 z = −2.63, p = 0.01
Cognitive Remediation
Tasks
Post hoc: 19 87% 88% −0.31 −0.50 to −0.12 51.0%, p = 0.01 z = −3.23, p = 0.001
Any other treatment
MCTrain for severity of obsessive–compulsive symptoms
Standard treatment 3 63% 81% −0.40 −0.70 to −0.09 0.0%, p = 0.80 z = −2.58, p = 0.01
Psychological treatment: 1 100% 100% −0.10 −0.45 to 0.25 NA z = −0.57, p = 0.57
Psychoeducation
Post hoc: 4 75% 88% −0.27 −0.50 to −0.04 0.0%, p = 0.56 z = −2.32, p = 0.02
Any other treatment
MCTrain for other mental disorders
Depression 1 100% 100% −0.63 −1.07 to −0.19 NA z = −2.81, p = 0.01
Psychological treatment:
Health Training
BPD 1 74% 72% −0.12 −0.65 to 0.41 NA z = −0.43, p = 0.66
Psychological treatment:
Progressive Muscle
Relaxation
MCTherap for severity of depressive symptoms
Nonactive treatment: Waitlist 3 100% 100% −2.80 −5.30 to −0.30 95.0%, p = 0.00 z = −2.19, p = 0.03
Psychological treatment: 3 100% 100% 0.02 −0.40 to 0.43 0.0%, p = 0.46 z = 0.07, p = 0.94
Cognitive–behavioural
treatments
Pharmacological Treatment 1 100% 100% −3.21 −4.68 to −1.74 NA z = −4.28, p ≤ 0.001
MCTherap for severity of anxiety symptoms
Other psychological 4 86% 89% −0.46 0.76 to −0.16 0.0%, p = 0.81 z = −3.00, p = 0.003
treatment: Cognitive–
behavioural treatments
Psychological treatment: 1 100% 100% −1.25 −2.21 to −0.29 NA z = −2.56, p = 0.01
Applied Relaxation
Nonactive treatment: Waitlist 1 92% 92% −1.85 −2.70 to −1.00 NA z = −4.26, p ≤ 0.001
MCTherap for other mental disorders
PTSD 2 100% 100% −2.36 −4.40 to −0.31 82.9%, p = 0.02 z = −2.26, p = 0.02
Nonactive treatment: Waitlist
BDD 1 100% 100% −1.36 −2.34 to −0.39 NA z = 2.75, p = 0.01
Nonactive treatment: Waitlist
MERIT for schizophrenia symptoms in early psychosis
Psychological intervention: 1 80% 100% −0.42 −1.36 to 0.52 NA z = −0.88, p = 0.38
Supportive Therapy

Note. BDD: body dysmorphic disorder; BPD: borderline personality disorder; PTSD: posttraumatic stress disorder.

MCTrain group reported significantly less severe depressive symp- training, PMR). In one study with a sample of 74 patients (Schilling,
toms than those receiving a psychological treatment (health training; Moritz, Kriston, Krieger, & Nagel, 2017), groups did not differ in the
Table 1). Less patients dropped out of the MCTrain group (3/41) than average severity of BPD symptoms (Table 1). More patients dropped
out of the control group (2/43). out of the PMR group (7/36) than out of the MCTrain group (1/38).
MCTrain for BPD was tested in two studies that compared Another study (Schilling, Moritz, Köther, & Nagel, 2015) did not report
MCTrain with a psychological treatment (progressive muscle relaxation data for any of the predefined outcome measures.
PHILIPP R. ET AL. 233

FIGURE 2 Standardized mean differences for Metacognitive Training versus standard or psychological treatment in patients with schizophrenia
(upper figure) and with obsessive–compulsive disorder (lower figure)

3.3.2 | MCTherap MCTherap was superior in alleviating depressive symptoms. Statistical

heterogeneity was considerable (Figure 3). Results corresponded with
We conducted meta‐analyses for 11 of the 19 studies that investi- responder analysis (Figure S7.1). Dropout rates did not differ between
gated MCTherap; in four studies, patients continued pharmacological the groups (Figure S7.2).
treatment (see Table S5.1 for detailed characteristics). Three studies compared a total sample of 98 patients with
Three studies compared a total sample of 99 patients with depression (Ashouri, Atef‐Vahid, Gharaee, & Rasoulian, 2013; Jordan
depression (Hagen et al., 2017; Mami, Sharifi, & Mahdavi, 2015; et al., 2014; Zemestani et al., 2016) receiving MCTherap against
Zemestani, Davoodi, Honarmand, Zargar, & Ottaviani, 2016) receiving patients receiving psychological treatments (CBT or Behavioural Acti-
MCTherap with patients on a waitlist to receive MCTherap. Overall, vation). Overall, treatment groups did not differ in the average severity
234 PHILIPP R. ET AL.

FIGURE 3 Standardized mean differences for Metacognitive Therapy versus nonactive waitlist (upper figure) and psychological or
pharmacological treatment (lower figure) in patients with depression

of depressive symptoms at the end of intervention (Table 1). Neither Two studies compared MCTherap with a nonactive waitlist
did the groups differ in the number of patients who responded to or group in a total sample of 41 patients (Wells & Colbear, 2012; Wells,
dropped out of the treatment (Figures S7.4 and S7.5, respectively). Walton, Lovell, & Proctor, 2015). Results suggest that MCTherap
There was no statistical heterogeneity for any of the outcomes may be superior to a waitlist group in alleviating PTSD symptoms
(Section S7). (Figure 4).
In a study (Ashouri et al., 2013) that tested MCTherap against a Statistical heterogeneity was considerable (Table 1). Compared
pharmacotherapy only condition in patients with depression, the with the patients in the waitlist group, more patients in the MCTherap
MCTherap group reported significantly less severe depressive group responded to the treatment (Figure S7.10). Dropout rates did
symptoms than patients receiving pharmacotherapy (Table 1). not differ between the groups (Figure S7.11).
Four studies compared MCTherap with other psychological treat- Because of the small number of studies and small sample sizes
ments (mindfulness‐based stress reduction or CBT) in a total sample of within the studies, visual examination of funnel plots was inconclusive
182 patients with mixed or co‐morbid anxiety disorders (Capobianco, (Section S7).
Reeves, Morrison, & Wells, 2018; Johnson, Hoffart, Nordahl, & One study tested MCTherap against a nonactive waitlist group in
Wampold, 2017; Kvistedal, 2011; Nordahl, 2009). Overall, MCTherap patients with body dysmorphic disorder (Rabiei, Mulkens, Kalantari,
was superior to another psychological treatment in alleviating symp- Molavi, & Bahrami, 2012). Patients in the MCTherap group reported
toms of anxiety (Table 1). Accordingly, more patients in the MCTherap less symptoms than those in the waitlist group (Table 1).
groups responded to the treatment than in the control groups (Figure
S7.7). Dropout rates did not differ between the groups (Figure S7.8). 3.3.3 | MERIT
Two single studies (Kvistedal, 2011; Wells et al., 2010) included in
the overall analysis (Figure 4) reported results in favour of MCTherap One study (Vohs et al., 2017) compared MERIT for the treatment of
compared with waitlist and psychological treatment (applied relaxa- patients with early phase psychosis (MERIT‐EP) with standard treat-
tion; Table 1). There was no statistical heterogeneity for any of the ment. Results showed that MERIT‐EP did not improve symptoms to
outcomes (Section S7). a greater extent than standard treatment at the end of intervention
PHILIPP R. ET AL. 235

FIGURE 4 Standardized mean differences for Metacognitive Therapy versus psychological or nonactive treatment in patients with mixed anxiety
disorders (upper figure) and versus nonactive waitlist in patients with posttraumatic stress disorder (lower figure)

(Table 1). Another study (Lysaker et al., 2015) compared patients whether the effectiveness of each intervention varied across mental
with schizophrenia who received either MERIT or Supportive Therapy. disorders.
However, the qualitative study did not report data for any of the We analysed the evidence of 34 RCTs and four NRCTs with a
predefined outcome measures. total sample of 2,148 patients quantitatively. More than half of the
included trials studied the effectiveness of MCTrain, whereas
MCTherap and MERIT were investigated less often. For all three
interventions, we found that a large proportion of the studies was
4 | DISCUSSION
conducted and (co‐)authored by the researchers who had developed
Our systematic review assessed the effectiveness of metacognitive the interventions.
interventions for adult patients with mental disorders. We showed
that MCTrain and MCTherap were at least as effective as another 4.1 | MCTrain
psychological intervention and mostly outperformed nonactive treat-
ments in treating patients with schizophrenia, OCD, anxiety disorders, We found that MCTrain was statistically superior to cognitive remedi-
PTSD, and depression. Patients with schizophrenia did not seem to ation tasks in patients with schizophrenia but not to any of the other
benefit from MERIT regarding symptom severity. Acceptability of control groups. One possible explanation may be that the other con-
metacognitive interventions was high among the investigated patient trol groups provided treatments in which patients' ability to reflect
groups. Because interventions were tested for the mental disorders and to think about themselves was fostered to a greater extent than
they were originally developed for, there is only evidence for a limited during cognitive remediation tasks. Moreover, MCTrain in addition
number of mental disorders so far. Hence, we did not investigate to standard treatment improved positive symptoms but only bordered
236 PHILIPP R. ET AL.

statistical significance. In a post hoc analysis across all studies, superior to CBT not only in patients with anxiety but also in patients
MCTrain was more effective in alleviating symptom severity than with depression. Their conclusion is limited by the fact that only one
any of the control groups. The magnitude of the effect was compara- of the primary studies (Nordahl, 2009) investigated a mixed sample
ble with a recent meta‐analysis on CBT for psychosis (Mehl, Werner, & including patients with depression next to GAD and eating or person-
Lincoln, 2015). These results may be due to the small number of stud- ality disorders. The other four primary studies were conducted in
ies that were included in the planned analyses, which may have led to patients with GAD (Kvistedal, 2011; van der Heiden, Muris, & van
a lower statistical power of the meta‐analyses compared with the der Molen, 2012; Wells et al., 2010) or PTSD (Proctor, 2008).
meta‐analysis that included all studies. Another explanation may One strength of the included primary studies on MCTherap is that
be that standard treatment for patients with schizophrenia included they reported data for the number of randomized patients. Neverthe-
pharmacotherapy (mostly antipsychotics), which usually reduces less, findings are limited by the small number of studies that we were
patients' positive symptoms substantially. Therefore, it might be diffi- able to combine for comparisons. Methodological limitations were
cult to show the additional benefits of MCTrain on positive symptoms similar to those found for MCTrain.
beyond standard treatment with antipsychotics. Our results are in line
with recently published meta‐analyses, which synthesized results
4.3 | MERIT
on positive symptoms and pooled data across all included studies
but differed in their selection of studies (Eichner & Berna, 2016;
The included studies indicate that this type of metacognitive interven-
Liu et al., 2018; van Oosterhout et al., 2015). Our meta‐analysis repre-
tion addressed other outcomes than symptom severity. As a long‐term
sents a more comprehensive selection of studies. The overall positive intervention, MERIT seemed to foster patients' insight and reflective-
effect for MCTrain needs to be interpreted carefully because of the
ness but did not alleviate symptom severity. More evidence needs to
differences in patient characteristics, control groups, and study design
be gathered for reliable conclusions.
of the primary studies (Jiang et al., 2015).
The effect for MCTrain compared with a waitlist control group in
patients with OCD was larger than reported in a recent meta‐analysis 4.4 | Other results
on unguided self‐help interventions (Pearcy, Anderson, Egan, & Rees,
Next to symptom severity, which was the main outcome in most of
2016). Our result is compromised by the small number of studies
the primary studies, a focus group with former psychiatric patients
included in the comparison and by the low retention rate. However,
named metacognitive changes and quality of life as relevant outcomes
current research states that low retention and acceptance rates are
(Brütt et al., 2017). Both were only analysed in about a quarter of the
common in these types of interventions. Subjective appraisal of
studies. Other outcomes that patients found highly relevant to their
myMCT was mainly favourable. Although guided interventions
everyday lives were analysed in none of the studies. Especially
seem to be more beneficial, the effectiveness of online interventions
patients with chronic mental disorders like schizophrenia who have
is similar to other psychotherapeutic approaches (Baumeister,
adapted to their symptoms may find interventions more helpful that
Reichler, Munzinger, & Lin, 2014; Karyotaki et al., 2015; Richards &
address other outcomes than symptom severity.
Richardson, 2012).
Further, only few of the included studies reported the predefined
secondary outcomes. Whereas response rates are well defined for
affective disorders (Bryant et al., 2008; Keller et al., 2000; Loerinc
4.2 | MCTherap
et al., 2015), criteria for schizophrenia are less consensual. Also,
adverse events should be reported systematically for psychotherapy
The meta‐analysis suggests that MCTherap outperforms another
studies (Meister et al., 2016). Results may provide helpful information
active psychological intervention in patients with anxiety. Whereas
on how to implement metacognitive interventions in terms of
MCTherap was superior to waitlist control groups in patients with
relevance and applicability.
depression, there was no evidence of a beneficial effect of MCTherap
when compared with cognitive–behavioural psychological treatments
for depression in our meta‐analysis. One theoretical explanation for 4.5 | Methodological evaluation of the systematic
these results might be that MCTherap was conceptualized for GAD review
and later adapted for other mental disorders. Therefore, MCTherap
may target underlying mechanisms of GAD more precisely than CBT, Our systematic review and meta‐analyses address most of the
a benefit that might not apply for depression. Another possible expla- shortcomings that of prior reviews on metacognitive interventions.
nation for these results may be that in one of the studies testing First, this contribution does not only cover evidence on different
MCTherap against CBT for depression (Jordan et al., 2014), the types of metacognitive interventions, it also provides an overview of
patients receiving MCTherap were more often diagnosed with severe results for various mental disorders. Second, by means of a sensitive
co‐morbidities than the ones receiving CBT. Our findings are in line search strategy including grey literature and contacting authors
with other meta‐analyses that reported the effectiveness of of metacognitive interventions, it is likely that we were able to
MCTherap for anxiety and PTSD (Normann et al., 2014; Sadeghi, identify most existing studies. In this context, we offer a working
Mokhber, Mahmoudi, Asgharipour, & Seyfi, 2015). The authors of definition that may enable other researchers to identify studies on
one review (Normann et al., 2014) concluded that MCTherap was metacognitive interventions more easily and encourage them to also
PHILIPP R. ET AL. 237

choose a comprehensive approach when reviewing future evidence in aim of this systematic review, which is why limitations due to indirect-
this field. Third, when preparing and conducting this systematic review ness of comparisons are unlikely (Guyatt et al., 2011).
and the meta‐analyses, we followed methodological standards and
reported results in accordance with current guidelines (Guyatt et al.,
4.7 | Future research
2011; Moher et al., 2009; Shea et al., 2007; The Cochrane
Collaboration, 2009). Two reviewers independently assessed eligibil- Although the number of studies testing the effectiveness of
ity, methodological quality, and extracted data. metacognitive interventions is constantly growing, the existing evi-
Although few primary studies reported results at follow‐up, this dence base does not seem to have become more conclusive within
review focused on the effectiveness of metacognitive interventions the last years. Particularly, trials performed by researchers other than
at the end of intervention. Therefore, the results are compromised the scientists who had developed the interventions would be desirable
and no conclusions about the long‐term effects can be drawn from to rule out allegiance bias. After addressing methodological shortcom-
this systematic review. As metacognitive interventions do not only ings, including small sample sizes, future research could investigate not
incorporate metacognitive elements, it remains unclear whether only whether the effectiveness of the different interventions varies
symptoms were alleviated by metacognitive elements or other ther- across disorders but also to which clinical context they can be applied
apeutic factors. In the primary studies, metacognitive changes were to best. For example, it can be concluded from our systematic review
assessed with a number of different instruments based on different and from the three separate meta‐analyses that MCTrain and
underlying constructs, which is why we were not able to report the MCTherap are short, accessible interventions that can easily be
results in more detail than in the primary studies or make conclusive adapted to various clinical settings. Due to the available evidence,
statements. The contribution of metacognitive elements to treat- MCTrain may be most effective when it is delivered as an add‐on
ment effects still needs to be tested, given that metacognitive treatment in an inpatient setting, and MCTherap and MERIT may be
changes are systematically assessed and analysed in future studies. most beneficial when implemented in outpatient care. Recent promis-
Steffen Moritz developed MCTrain and authored a number of stud- ing results for metacognitively oriented psychotherapies (Gordon‐King
ies that investigated the effectiveness of MCTrain. As a co‐author et al., 2018; Inchausti et al., 2017) suggest that their integrative
of this review, he was involved in drafting of the manuscript and treatment approach complements research in this field, alongside the
contributed to the interpretation of results, which constitutes a con- cognitive–behavioural interventions (Lysaker & Klion, 2017).
flict of interest. Looking at the summarized evidence, metacognitive interventions
are a theoretically founded extension to existing traditional psycho-
therapeutic interventions. The concept of metacognition adds another
4.6 | Quality of evidence perspective to understand underlying mechanisms of mental disorders
that are otherwise hard to capture and, thus, may help patients to
Even though it may be possible that the primary studies were con- broaden their self‐perception. Accordingly, we encourage that each
ducted with higher methodological quality, we rated a lot of domains intervention is further studied within its field in order to be beneficial
as unclear, because necessary information was not reported. Thus, for a wide range of patients. Large and independent multicentre
we were unable to make conclusive statements about the overall trials investigating short‐ and long‐term outcomes that are relevant
methodological quality and assessed it as moderate or unclear. For to patients, including adverse and long‐term effects, are needed to
the primary studies of all three metacognitive interventions, risk of strengthen the evidence base.
bias seemed unlikely with regard to the randomization procedures;
however, results of these primary studies may still be biased due to
possible conflicts of interest, which often were not stated, and inade- 5 | CO NC LUSIO N
quate blinding of patients, personnel, and outcome assessments. Pub-
Unclear to moderate quality evidence suggests that metacognitive
lication bias can be rated as unlikely because of our sensitive search
strategy (Guyatt et al., 2011). Statistical heterogeneity varied across interventions are likely to be effective in the treatment of a number
of mental disorders, particularly when compared with nonactive treat-
comparisons and was different for MCTrain and MCTherap. The esti-
ments. The effectiveness of these interventions, especially MCTrain
mates for MCTrain in patients with schizophrenia varied in their direc-
tion, leading to substantial heterogeneity. This can be explained by and MERIT, when compared with established psychotherapies like

different clinical characteristics and study designs. The estimates for CBT still needs to be studied. Further independent research should
address the methodological shortcomings of existing trials and focus
the MCTherap studies did not vary in their direction, but in their
on finding the ideal place of metacognitive interventions in the mental
extent. Pooled estimates for MCTrain in patients with schizophrenia
and OCD showed narrow confidence intervals. Accordingly, results health care system.

for the primary outcomes are likely to be precise (Guyatt et al.,

2011). Pooled estimates for the comparisons between MCTherap ACKNOWLEDGEMENTS
and waitlists had broad confidence intervals, likely due to the small This work was funded by the German Federal Ministry of Education
number of studies with small sample sizes that were combined for and Research (grant number 01KG1511). We thank Sandra Hischke
meta‐analyses. Therefore, results of these comparisons have to be and Lisa Lebherz for their support with data extraction. We also thank
interpreted as imprecise. All included studies directly addressed the Alessa Jansen who co‐authored the review protocol.
238 PHILIPP R. ET AL.

CONF LICT OF INT E RE ST Egger, M., Smith, G. D., Schneider, M., & Minder, C. (1997). Bias in meta‐
analysis detected by a simple, graphical test. BMJ, 315(7109),
Metacognitive Training was developed and studied by S. M. L. K. 629–634. https://doi.org/10.1136/bmj.315.7109.629
participated in trials of the Metacognitive Training as independent
Eichner, C., & Berna, F. (2016). Acceptance and efficacy of Metacognitive
statistician. All other authors declare no conflict of interest. Training (MCT) on positive symptoms and delusions in patients with
schizophrenia: A meta‐analysis taking into account important modera-
tors. Schizophrenia Bulletin, 42(4), 952–962. https://doi.org/10.1093/
AUTHOR CONTRIBUTIONS
schbul/sbv225
F. K., M. H., and L. K. designed the study, and F. K., R. M., M. H., S. M., Farris, S. G., McLean, C. P., van Meter, P. E., Simpson, H. B., & Foa, E. B.
and L. K. wrote the review protocol. F. K., J. L., and R. P. conducted (2013). Treatment response, symptom remission, and wellness in
literature searches. F. K., J. L., R. M., and R. P. extracted data. R. P. obsessive‐compulsive disorder. The Journal of Clinical Psychiatry,
74(7), 685–690. https://doi.org/10.4088/JCP.12m07789
and R. M. conducted the statistical analyses. R. P. and R. M. wrote
Flavell, J. H. (1979). Metacognition and cognitive monitoring: A new area
the first draft of the manuscript, and all authors revised it critically
of cognitive‐developmental inquiry. American Psychologist, 34(10),
for important intellectual content. We confirm that the manuscript 906–911. https://doi.org/10.1037//0003‐066X.34.10.906
has been read and approved by all named authors and that there are Gordon‐King, K., Schweitzer, R. D., & Dimaggio, G. (2018). Metacognitive
no other persons who satisfied the criteria for authorship but Interpersonal Therapy for personality disorders featuring emotional inhi-
bition: A multiple baseline case series. The Journal of Nervous and Mental
are not listed. We further confirm that the order of authors listed in
Disease. https://doi.org/10.1097/NMD.0000000000000789, 1.
the manuscript has been approved by all of us.
Guyatt, G. H., Oxman, A. D., Kunz, R., Brozek, J., Alonso‐Coello, P., Rind, D.,
et al. (2011). GRADE guidelines 6. Rating the quality of evidence—
ORCID Imprecision. Journal of Clinical Epidemiology, 64(12), 1283–1293.
https://doi.org/10.1016/j.jclinepi.2011.01.012
Rebecca Philipp https://orcid.org/0000-0002-1875-3823
Guyatt, G. H., Oxman, A. D., Kunz, R., Woodcock, J., Brozek, J., Helfand,
Levente Kriston https://orcid.org/0000-0003-0748-264X
M., et al. (2011). GRADE guidelines: 8. Rating the quality of evidence
Franziska Kühne https://orcid.org/0000-0001-9636-5247 —Indirectness. Journal of Clinical Epidemiology, 64(12), 1303–1310.
Martin Härter https://orcid.org/0000-0001-7443-9890 https://doi.org/10.1016/j.jclinepi.2011.04.014
Steffen Moritz https://orcid.org/0000-0001-8601-0143 Guyatt, G. H., Oxman, A. D., Montori, V., Vist, G., Kunz, R., Brozek, J., et al.
(2011). GRADE guidelines: 5. Rating the quality of evidence—Publica-
tion bias. Journal of Clinical Epidemiology, 64(12), 1277–1282. https://
RE FE R ENC E S
doi.org/10.1016/j.jclinepi.2011.01.011
American Psychiatric Association (2000). Diagnostic and statistical manual Guyatt, G. H., Oxman, A. D., Vist, G., Kunz, R., Brozek, J., Alonso‐Coello, P.,
of mental disorders—DSM‐IV‐TR. Washington, DC: American Psychiatric et al. (2011). GRADE guidelines: 4. Rating the quality of evidence—
Association. Study limitations (risk of bias). Journal of Clinical Epidemiology, 64(4),
Ashouri, A., Atef‐Vahid, M.‐K., Gharaee, B., & Rasoulian, M. (2013). Effec- 407–415. https://doi.org/10.1016/j.jclinepi.2010.07.017
tiveness of meta‐cognitive and cognitive‐behavioral therapy in Hagen, R., Hjemdal, O., Solem, S., Kennair, L. E. O., Nordahl, H. M.,
patients with major depressive disorder. Iranian Journal of Psychiatry Fisher, P., & Wells, A. (2017). Metacognitive Therapy for depression
and Behavioral Sciences, 7(2), 24–34. in adults: A waiting list randomized controlled trial with six months
follow‐up. Frontiers in Psychology, 8, 31. https://doi.org/10.3389/
Baumeister, H., Reichler, L., Munzinger, M., & Lin, J. (2014). The impact of fpsyg.2017.00031
guidance on Internet‐based mental health interventions—A systematic
review. Internet Interventions, 1(4), 205–215. https://doi.org/10.1016/ Hamm, J. A., Hasson‐Ohayon, I., Kukla, M., & Lysaker, P. H. (2013). Individ-
j.invent.2014.08.003 ual psychotherapy for schizophrenia: Trends and developments in the
wake of the recovery movement. Psychology Research and Behaviour
Brütt, A. L., Meister, R., Bernges, T., Moritz, S., Härter, M., Kriston, L., & Management, 6, 45–54. https://doi.org/10.2147/PRBM.S47891
Kühne, F. (2017). Patient involvement in a systematic review: Develop- Hauschildt, M., Schröder, J., & Moritz, S. (2016). Randomized‐controlled
ment and pilot evaluation of a patient workshop. Zeitschrift für Evidenz, trial on a novel (meta‐) cognitive self‐help approach for obsessive‐
Fortbildung und Qualität im Gesundheitswesen, 127‐128, 56–61. https:// compulsive disorder (“myMCT”). Journal of Obsessive‐Compulsive
doi.org/10.1016/j.zefq.2017.07.005 and Related Disorders, 10, 26–34. https://doi.org/10.1016/j.
Bryant, R. A., Felmingham, K., Kemp, A., Das, P., Hughes, G., Peduto, A., & jocrd.2016.04.010
Williams, L. (2008). Amygdala and ventral anterior cingulate activation Higgins, J. P. T., Thompson, S. G., Deeks, J. J., & Altman, D. G. (2003). Mea-
predicts treatment response to cognitive behaviour therapy for post‐ suring inconsistency in meta‐analyses. BMJ (Clinical Research Ed.),
traumatic stress disorder. Psychological Medicine, 38(4), 555–561. 327(7414), 557–560. https://doi.org/10.1136/bmj.327.7414.557
https://doi.org/10.1017/S0033291707002231 Hofmann, S. G., & Asmundson, G. J. G. (2008). Acceptance and
Capobianco, L., Reeves, D., Morrison, A. P., & Wells, A. (2018). Group mindfulness‐based therapy: New wave or old hat? Clinical Psychology
Metacognitive Therapy vs. Mindfulness Meditation Therapy in a Review, 28(1), 1–16. https://doi.org/10.1016/j.cpr.2007.09.003
transdiagnostic patient sample: A randomised feasibility trial. Howes, O. D., McCutcheon, R., Agid, O., de Bartolomeis, A., van Beveren,
Psychiatry Research, 259, 554–561. https://doi.org/10.1016/j. N. J. M., & Birnbaum, M. L. (2017). Treatment‐resistant schizophrenia:
psychres.2017.11.045 Treatment Response and Resistance in Psychosis (TRRIP) working
group consensus guidelines on diagnosis and terminology. The Ameri-
Dimaggio, G., Salvatore, G., MacBeth, A., Ottavi, P., Buonocore, L., & can Journal of Psychiatry, 174(3), 216–229. https://doi.org/10.1176/
Popolo, R. (2017). Metacognitive Interpersonal Therapy for personality appi.ajp.2016.16050503
disorders: A case study series. Journal of Contemporary Psychotherapy,
47(1), 11–21. https://doi.org/10.1007/s10879‐016‐9342‐7 Hunot, V., Moore, T. H. M., Caldwell, D. M., Furukawa, T. A., Davies, P.,
Jones, H., et al. (2013). 'Third wave' cognitive and behavioural therapies
Dimaggio, G., & Semerari, A. (2001). Pychopathological narrative forms. versus other psychological therapies for depression. The Cochrane
Journal of Constructivist Psychology, 14(1), 1–23. https://doi.org/ Database of Systematic Reviews, (10), CD008704. https://doi.org/
10.1080/107205301451335 10.1002/14651858.CD008704.pub2
PHILIPP R. ET AL. 239

Inchausti, F., García‐Poveda, N. V., Ballesteros‐Prados, A., Fonseca‐ self reflection in the psychotherapy for schizophrenia: A
Pedrero, E., Ortuño‐Sierra, J., & Sánchez‐Reales, S. (2017). A pilot study conceptual model of the key tasks and processes. Psychology and Psy-
on feasibility, acceptance and effectiveness of metacognitive‐oriented chotherapy, 84(1), 58–69. discussion 98‐110. https://doi.org/
social skills training in schizophrenia. BMC Psychiatry, 17(1), 217. 10.1348/147608310X520436
https://doi.org/10.1186/s12888‐017‐1378‐z Lysaker, P. H., & Klion, R. E. (2017). Recovery, meaning‐making, and severe
Jelinek, L., Hauschildt, M., Wittekind, C. E., Schneider, B. C., Kriston, L., & mental illness: A comprehensive guide to Metacognitive Reflection and
Moritz, S. (2016). Efficacy of Metacognitive Training for depression: A Insight Therapy (1st ed.). New York, NY: Routledge. https://doi.org/
randomized controlled trial. Psychotherapy and Psychosomatics, 85(4), 10.4324/9781315447001
231–234. https://doi.org/10.1159/000443699
Lysaker, P. H., Kukla, M., Belanger, E., White, D. A., Buck, K. D., Luther, L.,
Jiang, J., Zhang, L., Zhu, Z., Li, W., & Li, C. (2015). Metacognitive training for et al. (2015). Individual psychotherapy and changes in self‐experience
schizophrenia: A systematic review. Shanghai Archives of Psychiatry, in schizophrenia: A qualitative comparison of patients in
27(3), 149–157. https://doi.org/10.11919/j.issn.1002‐0829.215065 metacognitively focused and supportive psychotherapy. Psychiatry,
Johnson, S. U., Hoffart, A., Nordahl, H. M., & Wampold, B. E. (2017). 78(4), 305–316. https://doi.org/10.1080/00332747.2015.1063916
Metacognitive therapy versus disorder‐specific CBT for comorbid anx- Lysaker, P. H., & Lysaker, J. T. (2001). Psychosis and the disintegration of
iety disorders: A randomized controlled trial. Journal of Anxiety dialogical self‐structure: Problems posed by schizophrenia for the
Disorders, 50, 103–112. https://doi.org/10.1016/j.janxdis.2017.06.004 maintenance of dialogue. British Journal of Medical Psychology, 74,
Jordan, J., Carter, J. D., McIntosh, V. V. W., Fernando, K., Frampton, C. M. 23–33. https://doi.org/10.1348/000711201160777
A., & Porter, R. J. (2014). Metacognitive therapy versus cognitive Mami, S., Sharifi, M., & Mahdavi, A. (2015). Effectiveness of Metacognitive
behavioural therapy for depression: A randomized pilot study. The Aus- Therapy in reducing psychological and metacognitive factors in
tralian and New Zealand Journal of Psychiatry, 48(10), 932–943. https:// patients with major depressive disorder in a military hospital. The
doi.org/10.1177/0004867414533015 Annals of Military and Health Sciences Research, 13, 112–118.
Karyotaki, E., Kleiboer, A., Smit, F., Turner, D. T., Pastor, A. M., Andersson, Mehl, S., Werner, D., & Lincoln, T. M. (2015). Does Cognitive Behavior
G., et al. (2015). Predictors of treatment dropout in self‐guided Therapy for psychosis (CBTp) show a sustainable effect on delusions?
web‐based interventions for depression: An 'individual patient data' A meta‐analysis. Frontiers in Psychology, 6, 1450. https://doi.org/
meta‐analysis. Psychological Medicine, 45(13), 2717–2726. https://doi. 10.3389/fpsyg.2015.01450
org/10.1017/S0033291715000665
Meister, R., von Wolff, A., & Kriston, L. (2015). Odds ratios of treatment
Keller, M. B., & Boland, R. J. (1998). Implications of failing to achieve suc- response were well approximated from continuous rating scale scores
cessful long‐term maintenance treatment of recurrent unipolar major for meta‐analysis. Journal of Clinical Epidemiology, 68(7), 740–751.
depression. Biological Psychiatry, 44(5), 348–360. https://doi.org/ https://doi.org/10.1016/j.jclinepi.2015.02.006
10.1016/S0006‐3223(98)00110‐3
Meister, R., von Wolff, A., Mohr, H., Nestoriuc, Y., Härter, M., Hölzel, L., &
Keller, M. B., McCullough, J. P., Klein, D. N., Arnow, B., Dunner, D. L., & Kriston, L. (2016). Adverse event methods were heterogeneous and
Gelenberg, A. J. (2000). A comparison of Nefazodone, the Cognitive insufficiently reported in randomized trials on persistent depressive
Behavioral‐Analysis System of Psychotherapy, and their disorder. Journal of Clinical Epidemiology, 71, 97–108. https://doi.org/
combination for the treatment of chronic depression. The New England 10.1016/j.jclinepi.2015.10.007
Journal of Medicine, 342(20), 1462–1470. https://doi.org/10.1056/
Moher, D., Liberati, A., Tetzlaff, J., Altman, D. G., & The PRSIMA Group
NEJM200005183422001
(2009). Preferred reporting items for systematic reviews and meta‐
Kriston, L. (2013). Dealing with clinical heterogeneity in meta‐analysis. analyses: The PRISMA statement. PLoS Medicine, 6(7), 1–6. https://
Assumptions, methods, interpretation. International Journal of Methods doi.org/10.1371/journal.pmed.1000097.g001
in Psychiatric Research, 22(1), 1–15. https://doi.org/10.1002/mpr.1377
Moritz, S., Andreou, C., Schneider, B. C., Wittekind, C. E., Menon, M.,
Kühne, F., Meister, R., Jansen, A., Härter, M., Moritz, S., & Kriston, L. Balzan, R. P., & Woodward, T. S. (2014). Sowing the seeds of doubt:
(2017). Effectiveness of metacognitive interventions for mental disor- A narrative review on Metacognitive Training in schizophrenia.
ders in adults: A systematic review protocol (METACOG). BMJ Open, Clinical Psychology Review, 34(4), 358–366. https://doi.org/10.1016/j.
7(6), e015428. https://doi.org/10.1136/bmjopen‐2016‐015428 cpr.2014.04.004
Kvistedal, D. J. (2011). A randomized controlled trial of metacognitive therapy Moritz, S., Hauschildt, M., Murray, S. C., Pedersen, A., Krausz, M., & Jelinek,
vs. cognitive behavioral therapy vs. waiting list in the treatment of adults L. (2018). New wine in an old bottle?: Evaluation of myMCT as an inte-
with generalized anxiety disorder: A preliminary analysis (Graduate thesis). grative bibliotherapy for obsessive‐compulsive disorder. Journal of
Trondheim, Norway: Norwegian University of Science and Technology. Obsessive‐Compulsive and Related Disorders, 16, 88–97. https://doi.
Lewin, A. B., de Nadai, A. S., Park, J., Goodman, W. K., Murphy, T. K., & org/10.1016/j.jocrd.2017.12.009
Storch, E. A. (2011). Refining clinical judgment of treatment outcome Moritz, S., Jelinek, L., Hauschildt, M., & Naber, D. (2010). How to treat the
in obsessive‐compulsive disorder. Psychiatry Research, 185(3), untreated: Effectiveness of a self‐help metacognitive training program
394–401. https://doi.org/10.1016/j.psychres.2010.08.021 (myMCT) for obsessive‐compulsive disorder. Dialogues in Clinical Neu-
Liu, Y.‐C., Tang, C.‐C., Hung, T.‐T., Tsai, P.‐C., & Lin, M.‐F. (2018). The effi- roscience, 12, 209–220.
cacy of Metacognitive Training for delusions in patients with Moritz, S., & Lysaker, P. H. (2018). Metacognition—What did James H.
schizophrenia: A meta‐analysis of randomized controlled trials informs Flavell really say and the implications for the conceptualization and
evidence‐based practice. Worldviews on Evidence‐Based Nursing, 15(2), design of metacognitive interventions. Schizophrenia Research.
130–139. https://doi.org/10.1111/wvn.12282 https://doi.org/10.1016/j.schres.2018.06.001
Loerinc, A. G., Meuret, A. E., Twohig, M. P., Rosenfield, D., Bluett, E. J., & Moritz, S., Stepulovs, O., Schröder, J., Hottenrott, B., Meyer, B., &
Craske, M. G. (2015). Response rates for CBT for anxiety disorders: Hauschildt, M. (2016). Is the whole less than the sum of its parts?: Full
Need for standardized criteria. Clinical Psychology Review, 42, 72–82. versus individually adapted metacognitive self‐help for obsessive‐
https://doi.org/10.1016/j.cpr.2015.08.004 compulsive disorder: A randomized controlled trial. Journal of
Ludvik, D., & Boschen, M. J. (2015). Cognitive restructuring and detached Obsessive‐Compulsive and Related Disorders, 9, 107–115. https://doi.
mindfulness: Comparative impact on a compulsive checking task. org/10.1016/j.jocrd.2016.04.001
Journal of Obsessive‐Compulsive and Related Disorders, 5, 8–15. Moritz, S., & Woodward, T. S. (2007). Metacognitive Training for schizo-
https://doi.org/10.1016/j.jocrd.2015.01.004 phrenia patients (MCT): A pilot study on feasibility, treatment
Lysaker, P. H., Buck, K. D., Carcione, A., Procacci, M., Salvatore, G., Nicolò, adherence, and subjective efficacy. The German Journal of Psychiatry,
G., & Dimaggio, G. (2011). Addressing metacognitive capacity for 10(3), 69–78.
240 PHILIPP R. ET AL.

Nordahl, H. M. (2009). Effectiveness of brief Metacognitive Therapy versus The Cochrane Collaboration. (2009). Cochrane Handbook for Systematic
cognitive‐behavioral therapy in a general outpatient setting. Interna- Reviews of Interventions Version 5.0.2.
tional Journal of Cognitive Therapy, 2(2), 152–159. https://doi.org/ van der Heiden, C., Muris, P., & van der Molen, H. T. (2012). Randomized
10.1521/ijct.2009.2.2.152 controlled trial on the effectiveness of Metacognitive Therapy and
Normann, N., van Emmerik, A. A. P., & Morina, N. (2014). The efficacy of Intolerance‐of‐Uncertainty Therapy for generalized anxiety disorder.
Metacognitive Therapy for anxiety and depression: A meta‐analytic Behaviour Research and Therapy, 50(2), 100–109. https://doi.org/
review. Depression and Anxiety, 31(5), 402–411. https://doi.org/ 10.1016/j.brat.2011.12.005
10.1002/da.22273 van Oosterhout, B., Smit, F., Krabbendam, L., Castelein, S., Staring, A. B. P., &
Pearcy, C. P., Anderson, R. A., Egan, S. J., & Rees, C. S. (2016). A systematic van der Gaag, M. (2015). Metacognitive Training for schizophrenia
review and meta‐analysis of self‐help therapeutic interventions for spectrum patients: A meta‐analysis on outcome studies. Psychological
obsessive‐compulsive disorder: Is therapeutic contact key to overall Medicine, 46(1), 47–57. https://doi.org/10.1017/S0033291715001105
improvement? Journal of Behavior Therapy and Experimental Psychiatry, Vohs, J. L., Leonhardt, B. L., James, A. V., Francis, M. M., Breier, A.,
51, 74–83. https://doi.org/10.1016/j.jbtep.2015.12.007 Mehdiyoun, N., et al. (2017). Metacognitive Reflection and Insight
Proctor, D. (2008). A randomised controlled‐trial of metacognitive therapy Therapy for early psychosis: A preliminary study of a novel integrative
versus exposure therapy for post‐traumatic stress disorder (Doctoral). psychotherapy. Schizophrenia Research https://doi.org/10.1016/j.
Manchester, United Kingdom: University of Manchester. Retrieved schres.2017.10.041, 195, 428–433.
from http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.491142 Wells, A. (2013). Advances in Metacognitive Therapy. International Journal
of Cognitive Therapy, 6(2), 186–201. https://doi.org/10.1521/
Rabiei, M., Mulkens, S., Kalantari, M., Molavi, H., & Bahrami, F. (2012).
ijct.2013.6.2.186
Metacognitive Therapy for body dysmorphic disorder patients in Iran:
Acceptability and proof of concept. Journal of Behavior Therapy and Wells, A., & Colbear, J. S. (2012). Treating posttraumatic stress disorder
Experimental Psychiatry, 43(2), 724–729. https://doi.org/10.1016/j. with metacognitive therapy: A preliminary controlled trial. Journal of
jbtep.2011.09.013 Clinical Psychology, 68(4), 373–381. https://doi.org/10.1002/jclp.20871
Richards, D., & Richardson, T. (2012). Computer‐based psychological treat- Wells, A., & Matthews, G. (1994). Attention and emotion: A clinical perspec-
ments for depression: A systematic review and meta‐analysis. Clinical tive. Hove, UK: Lawrence Erlbaum.
Psychology Review, 32(4), 329–342. https://doi.org/10.1016/j. Wells, A., & Purdon, C. (1999). Metacognition and cognitive‐behaviour
cpr.2012.02.004 therapy: A special issue. Clinical Psychology & Psychotherapy, 6,
Richter, D. (1999). Chronic mental illness and the limits of the 71–72. https://doi.org/10.1002/(SICI)1099‐0879(199905)6:2<71::
biopsychosocial model. Medicine, Health Care and Philosophy, 2, AID‐CPP186>3.0.CO;2‐G
21–30. https://doi.org/10.1023/A:1009968106317 Wells, A., Walton, D., Lovell, K., & Proctor, D. (2015). Metacognitive
Therapy versus prolonged exposure in adults with chronic post‐
Sadeghi, R., Mokhber, N., Mahmoudi, L. Z., Asgharipour, N., & Seyfi, H.
traumatic stress disorder: A parallel randomized controlled trial.
(2015). A systematic review and meta‐analysis on controlled treatment
Cognitive Therapy and Research, 39(1), 70–80. https://doi.org/
trials of metacognitive therapy for anxiety disorders. Journal of
10.1007/s10608‐014‐9636‐6
Research in Medical Sciences: The official Journal of Isfahan University
of Medical Sciences, 20(9), 901–909. https://doi.org/10.4103/1735‐ Wells, A., Welford, M., King, P., Papageorgiou, C., Wisely, J., & Mendel, E.
1995.170632 (2010). A pilot randomized trial of metacognitive therapy vs applied
relaxation in the treatment of adults with generalized anxiety disorder.
Schilling, L., Moritz, S., Köther, U., & Nagel, M. (2015). Preliminary results
Behaviour Research and Therapy, 48(5), 429–434. https://doi.org/
on acceptance, feasibility, and subjective efficacy of the add‐on group
10.1016/j.brat.2009.11.013
intervention Metacognitive Training for borderline patients. Journal of
Cognitive Psychotherapy: An International Quarterly, 29(2), 153–164. World Health Organization (1992). The ICD‐10 classification of mental and
https://doi.org/10.1891/0889‐8391.29.2.153 behavioural disorders: Clinical descriptions and diagnostic guidelines.
Geneva: World Health Organization.
Schilling, L., Moritz, S., Kriston, L., Krieger, M., & Nagel, M. (2017). Efficacy
of Metacognitive Training for patients with borderline personality Zemestani, M., Davoodi, I., Honarmand, M. M., Zargar, Y., & Ottaviani, C.
disorder: Preliminary results. Psychiatry Research https://doi.org/ (2016). Comparative effects of group metacognitive therapy
10.1016/j.psychres.2017.09.024, 262, 459–464. versus behavioural activation in moderately depressed students.
Journal of Mental Health, 25(6), 479–485. https://doi.org/10.3109/
Semerari, A., Carcione, A., Dimaggio, G., Falcone, M., Nicolò, G., Procacci, 09638237.2015.1057326
M., & Alleva, G. (2003). How to evaluate metacognitive functioning in
psychotherapy?: The metacognition assessment scale and its applica-
tions. Clinical Psychology & Psychotherapy, 10(4), 238–261. https:// SUPPORTI NG INFORMATION
doi.org/10.1002/cpp.362
Additional supporting information may be found online in the
Shea, B. J., Grimshaw, J. M., Wells, G. A., Boers, M., Andersson, N., Hamel, C.,
Supporting Information section at the end of the article.
et al. (2007). Development of AMSTAR: A measurement tool to assess
the methodological quality of systematic reviews. BMC Medical Research
Methodology, 7, 10. https://doi.org/10.1186/1471‐2288‐7‐10
Sterne, J. A., Hernán, M. A., Reeves, B. C., Savović, J., Berkman, N. D.,
How to cite this article: Philipp R, Kriston L, Lanio J, et al.
Viswanathan, M., et al. (2016). ROBINS‐I: A tool for assessing risk of
bias in non‐randomised studies of interventions. BMJ (Clinical Research Effectiveness of metacognitive interventions for mental disor-
Ed.), 355, i4919. ders in adults—A systematic review and meta‐analysis
Suissa, S. (1991). Binary methods for continous outcomes: A parametric (METACOG). Clin Psychol Psychother. 2019;26:227–240.
alternative. Journal of Clinical Epidemiology, 44(3), 241–248. https:// https://doi.org/10.1002/cpp.2345
doi.org/10.1016/0895‐4356(91)90035‐8