Clinical effectiveness of metacognitive training as a transdiagnostic program in routine clinical settings: A single-group pre-post study

Objective To evaluate the clinical effectiveness of metacognitive training (MCT) as a transdiagnostic program, on a diverse population with mental disorders in routine day-care settings through a single-group pre-post-design. Results Thirty-four participants diagnosed with various mental disorders (schizophrenia = 22, non-schizophrenia = 12) received ten MCT group-sessions. Intent-to-treat analyses revealed signicant improvements in quality of life/global functioning during the intervention period, and these improvements were well-maintained during the follow-up (all p < 0.05). The baseline to follow-up treatment effect sizes for quality of life and global functioning were middle (Hedge's g = 0.44 and 0.47, respectively). Signicant improvements were also found in depressive symptoms during both the intervention and follow-up periods (all p < 0.05), but not in cognitive insight. Overall, participants were highly satised with the MCT content and format. Scores on almost all outcomes (except for depression) at each assessment point were not signicantly different between the schizophrenic and non-schizophrenic sub-groups. The ndings of this study suggest that MCT for a diverse population with mental disorders is a potentially effective approach in improving quality of life/global functioning and other clinical outcomes in routine day-care settings.


Introduction
Diverse non-pharmacological approaches for schizophrenia have been developed as complementary strategies to antipsychotic medication. An example of a novel psychological approach based on cognitive theory is metacognitive training for psychosis (MCT). MCT is theory-driven, standardized, and manualized group training for schizophrenia, targeting common cognitive errors and problem-solving biases [1][2][3]. Several meta-analyses of existing clinical trials, mostly conducted in Western countries, have demonstrated that MCT reduces the positive schizophrenic symptoms, and is particularly effective in reducing delusions [4,5]. A Japanese study group has recently conducted a randomized controlled trial, and also demonstrated the e cacy of MCT among patients with schizophrenia [6].
Although MCT is designed mainly for schizophrenia, the program is not limited to schizophrenic patients [1][2][3]7]. This is because most of the cognitive errors and problem-solving biases addressed in MCT, are also common in other mental disorders (e.g. monocausal attributions, jumping to conclusions, etc.) [8]. In fact, in routine clinical settings, especially in psychiatric day-care services in Japan, patients usually attend the same program regardless of disorder [9]. Thus, MCT is provided not only for schizophrenia but also for a wide range of other disorders in clinical settings. Previous studies of MCT only targeted individuals with schizophrenia, but there is no evidence supporting the clinical effectiveness of MCT as a transdiagnostic program.
Therefore, the purpose of this study was to evaluate the effectiveness of MCT as a transdiagnostic program, on a diverse population with mental disorders in routine Japanese day-care settings through a single-group pre-post study.

Study design
The study employed a single-arm pre-post design, and was carried out from April 2018 to December 2019. Participants were recruited at two psychiatric day-care centers in Miyazaki prefecture, Japan. All participants received up to ten sessions of MCT. Assessments were conducted at baseline (preintervention: Pre), 5 weeks (mid-intervention: Mid), 10 weeks (post-intervention: Post), and 14 weeks (follow-up: FU).
This study was conducted according to the Transparent Reporting of Evaluations with Nonrandomized Designs (TREND) statement.

Participants
In order that the study population re ect routine psychiatric day-care practice in Japan, we set minimal inclusion/exclusion criteria and then recruited a broader, heterogeneous population. The inclusion criteria were: patients diagnosed with any of the mental disorders listed in DSM-IV; aged ≥ 20 years; and patients who were willing to participate. We excluded participants only if their condition was expected to worsen by participating in MCT and/or they posed a threat to group dynamics.

Intervention (MCT)
MCT group sessions were delivered based on the Japanese version of the MCT manual and session materials [7], and consisted of ten 60-minute weekly sessions. Each group had 4-6 participants. More details of MCT programs can be found at the MCT website [7]. All MCT sessions were conducted by one or two mental health care professionals (nurse, occupational therapist, and/or psychiatric social worker) who worked full-time at the study institutions. All therapists had considerable experience as study therapists in a prior randomized controlled trial [6], and were adequately trained via a one day in-person workshop.

Results
Participants' ow and baseline characteristics Thirty-eight patients were recruited for the study, of which four patients were excluded. The remaining 34 patients enrolled in the study. The mean number of MCT sessions attended was 8.44 (SD = 2.02), and 26 patients (76%) completed the program (eight dropped out). Please also see Additional le 1: Figure S1 for ow diagram. We observed no serious adverse events during the study. There was no change in prescribed medications during the study.
Of the 34 subjects, 22 (65%) were diagnosed with schizophrenia, and the other 12 non-schizophrenic patients (35%) were diagnosed with either major depressive disorder (n = 4) or bipolar-II disorder (n = 8). The average age was 52.7 (SD = 11.4) years. See Table 1 for more details about baseline clinical characteristics.
At the end of the MCT sessions, we also evaluated patient satisfaction with treatment (MCT Satisfaction Scale [22,23]). The reliability and validity of the Japanese versions of all measures have been con rmed.

Statistical analysis
The analyses were conducted based on the intention-to-treat (ITT) principle, whereby the last obtained data for dropouts were carried forward until the endpoint assessment. The Pre, Mid, Post, and FU scores for the outcome measures (except for the MCT Satisfaction Scale) were analyzed with single-factor (time) repeated measures analysis of variance (ANOVA). Where the repeated ANOVA indicated signi cant changes, pairwise Bonferroni corrected t-tests were conducted for post-hoc tests. The magnitude of the within-group treatment effect was determined as the effect size based on Hedges' g. As an additional analysis, we compared the scores for transdiagnostic outcomes (GAF, EQ-5D-5L, BDI-II, and MCT Satisfaction Scale) between sub-groups (schizophrenic group vs. non-schizophrenic group) using Welch's t-test.
All statistical tests were two-tailed, and α = 0.05 was employed. Statistical analyses were performed using IBM SPSS Statistics, version 24.0 (IBM, Armonk, New York, USA). Outcomes Table 2 presents changes in all transdiagnostic outcome measures during the study. Within-group treatment effect sizes (Hedges' g) for transdiagnostic outcomes are shown in Table 3. For all subjects, signi cant improvements were observed in EQ-5D-5L and GAF during the intervention period (Pre-Post) (p = 0.014 and 0.002, respectively), and further improvements were observed during the follow-up (Post-FU) (p = 0.021 and 0.039, respectively). The Pre-FU treatment effect sizes for EQ-5D-5L and GAF were middle (Hedge's g = 0.44 and 0.47, respectively). As for cognitive insight, there were no signi cant differences in all BCIS sub-scales. Signi cant improvements were observed in BDI-II during the intervention period (Pre-Post) (p = 0.006), and further improvements were observed during the follow-up    In the schizophrenic sub-group, signi cant differences on PANSS positive scores were observed during both the intervention and follow-up periods (p = 0.002 and < 0.001, respectively). As for cognitive bias, there were no signi cant differences in total scores of CBQp. However signi cant improvements were observed in JTC sub-scale scores of CBQp during the intervention period (p = 0.049), and these improvements were maintained during the follow-up (see Additional le 1: Table S2 for more details). Table 3 shows means and standard deviations of transdiagnostic outcome scores (EQ5D-5L, GAF, BCIS, BDI-II and MCT Satisfaction Scale) at each assessment point for the schizophrenic and nonschizophrenic sub-groups. Scores on BDI-II at Pre and Mid in the non-schizophrenic group were signi cantly higher than those in the schizophrenic group (p = 0.010 and 0.035, respectively). Scores on other outcome measures (GAF, EQ-5D-5L, and MCT Satisfaction Scale) at each assessment point were not signi cantly different between sub-groups.

Discussion
This study aimed to examine the clinical effectiveness of MCT on a diverse population with chronic mental disorders in routine Japanese day-care settings. The key nding of this study is that, regardless of primary disorder (schizophrenia or non-schizophrenia), MCT leads to signi cant improvements in quality of life, global functioning, and depression. Both schizophrenic and non-schizophrenic patients were highly satis ed with the MCT content and format.
This study was designed to recruit patients similar to those seen in routine day-care settings; as a result, 65% had schizophrenia, as is typical (56%) in clinical practice [24]. Also, in this study, patients who had a primary diagnosis of schizophrenia had less severe schizophrenic symptoms at baseline than those observed in our previous randomized controlled trial [6].
Although patients suffering from various mental disorders attended the same MCT program, they obtained signi cant improvements in their primary psychiatric symptoms, supporting the potential effectiveness of MCT as a transdiagnostic program for a wide range of mental disorders. In the schizophrenic sub-group, positive psychotic symptoms and cognitive biases commonly seen in schizophrenia (i.e. jumping to conclusions) were signi cantly improved through receiving MCT; though the Pre-FU treatment effect size of 0.47 (Hedge's g) on PANSS was lower than that of 0.71 in our previous trial [6]. In the non-schizophrenic (mood disorder) sub-group, the baseline severity of depressive symptoms was higher than that in the schizophrenic sub-group, but was signi cantly improved through MCT. As mentioned in the introduction, although MCT was originally developed mainly for schizophrenia [1][2][3], most of the MCT modules address the cognitive errors and problem-solving biases that are also common in other mental disorders; thus, MCT may also reduce depressive symptoms. These improved psychiatric symptoms/biases, then, might have contributed to improving quality of life and global functioning in this study [1,25,26] Our transdiagnostic MCT program led to considerable improvements in the study population; however, other factors unrelated to MCT-speci c effects might also have contributed to this positive outcome. First, the group-format itself might have had a signi cant treatment effect. Based on the MCT Satisfaction Scale, participants were very satis ed with the MCT, but the highly-rated items (e.g. "The training was fun", "I was pleased to go to the training regularly", "I found it bene cial that the training was administered in a group") may also appear in other well-organized group treatments [27][28][29]. As far as we are aware, no previous MCT studies have employed group-based active psychological-treatment as a control condition (e.g. wait-list, treatment as usual, supportive counseling, CogPack [computerized cognitive rehabilitation program], or newspaper discussion group) [4,5]. Further studies should pin down MCT speci c-effects by using more MCT-speci c questionnaires and/or employing other group-based active treatment controls. Second, the therapists in this study already had experience providing MCT [6], and were practitioners who regularly worked at the institution where the study was conducted. Therefore, each therapist may have already established a good relationship with patients. This factor may have also had a positive impact on treatment effects.
Although treatment satisfaction was higher compared to other studies, the dropout rate (24%) in this study was similar to rates in previous studies4. The main reasons given for dropping out were because of schedule con icts, and because the therapy was too di cult. Regarding the former reason, participants who did not regularly attend MCT were considered to be non-completers in this study, even though these patients actually continued irregular participation in MCT (i.e. they did not completely drop out from the treatment). This is because the original MCT protocol allows participants to start with any module at any time; so, for the purposes of the study, drop outs due to schedule con icts do not seem to be a serious problem in a clinical setting. As for patients who said the therapy was too di cult, this may be due to the group-format therapy. In group sessions, therapists sometimes nd it di cult to adjust the pace of sessions even when the therapist observes that a patient is struggling (e.g they did not fully understand the therapy contents of the module). For patients who struggle in group therapy, individual-based MCT (known as MCT+ [30]) should be introduced.
In conclusion, the ndings of this study indicate the potential e cacy of MCT as a transdiagnostic program in routine clinical settings. Future research should employ active psychological controlled conditions and larger sample sizes in order to replicate the study ndings and address the limitations of this study.

Limitations
First, this study employed a single-group design without a control group; therefore, we cannot conclusively state that our MCT was effective. The presence of the Hawthorne effect as well as other non-treatment speci c effects, which are typically observed in uncontrolled trials, cannot be ruled out in our study design. Second, our study had a limited number of participants. Third, diagnosis of samples included in the non-schizophrenic group was not uniform; it may be inappropriate to compare the nonschizophrenic group with the schizophrenic group. Fourth, satisfaction feedback was only obtained from patients completed the MCT. Future studies should obtain feedback from dropouts so that we can gather information towards the prevention of patients dropping out.