An empirical study of offline and online group metacognitive training in patients with schizophrenia

Abstract

Background

Schizophrenia often has hallucinations, delusions and other complex mental symptoms, metacognitive training (MCT) is a type of low-intensity group training. However, MCT requires a lot of time and expense, increasing the workload of professionals and the financial burden of patients. hybrid offline-online MCT reduces time and economic costs to the greatest extent. This study was designed to validate the interventional effects of a hybrid offline-online MCT model on psychiatric symptoms, delusion severity, and insight in patients with schizophrenia, while concurrently evaluating its clinical applicability.

Methods

Randomized controlled trials comparing hybrid offline-online MCT were conducted in patients with paranoid schizophrenia. The positive and negative syndrome scale (PANSS), the Chinese version of the delusion feature rating scale (C-CDRS), the Chinese version of the psychotic symptom rating scale (PSYRATS), and the insight and treatment attitude questionnaire (ITAQ) were used to evaluate psychiatric symptom improvement and insight in the baseline (T0), 4 weeks after the initiation of training (T1), and upon completion of the training (T2, week 8).

Results

The total scores of PANSS, C-CDRS, PSYRATS, and their subscale scores in the MCT and control groups were lower than baseline scores at weeks four and eight, while the total ITAQ score was higher than baseline scores (P < 0.01). In the MCT group, PANSS positive subscale and total score, C-CDRS total score, and PSYRATS delusion subscale and total score were lower than those in the control group at weeks four and eight, whereas the ITAQ total score was higher than the baseline score (P < 0.01).

Conclusions

Hybrid offline-online group MCT can reduce positive symptoms in patients with schizophrenia, alleviate delusions, and improve patients' insight.

Trial registration

The trial was registered at the Chinese Clinical Trial Registry on 01/24/2015 (www.chictr.org.cn; ChiCTRID: ChiCTR2500096505).

Supplementary Information

The online version contains supplementary material available at 10.1186/s12888-025-06968-0.

Introduction

Schizophrenia is a group of chronic and severe mental disorders with unclear etiology, characterized by cognitive, thought, emotional, and behavioral impairments [17]. Paranoid schizophrenia is the most common subtype of schizophrenia and is characterized by delusions, hallucinations, and corresponding behavioral disorders [35]. According to previous studies, the lifetime prevalence of schizophrenia in China is 0.75% [16]. The World Health Organization identifies schizophrenia as one of the top 10 diseases contributing to the global burden of disease [42]. Approximately 20%–30% of patients with paranoid schizophrenia develop antipsychotic treatment resistance, rendering pharmacological monotherapy insufficient for achieving substantial mental state improvement [3]. Metacognition, introduced by American developmental psychologist Flavell, is vital in cognitive repair in schizophrenia [6]. Based on cognitive behavioral therapy, Moritz developed a metacognitive training manual to improve cognitive bias in patients with schizophrenia [31]. Several studies have revealed that MCT enhances patients' self-awareness and insight into these cognitive distortions, alleviating positive symptoms of mental disorders, particularly in patients with paranoid schizophrenia, while enhancing cognitive function, social functioning, and symptom improvement [10, 13, 41, 44]. Conventional MCT modalities frequently demonstrate suboptimal treatment adherence due to protracted intervention durations and logistical barriers associated with hospital commuting [7]. The hybrid offline-online intervention paradigm effectively addresses these limitations [9]. Concurrently, advancements in mobile internet technologies have expanded the implementation of digital interventions in psychiatric disorder management [2, 23, 40]. Emerging evidence indicates that such integrated approaches not only ameliorate psychopathological symptoms but also reduce patients' socioeconomic burdens and treatment-related psychological distress [32]. Nahum's investigation revealed that web-based cognitive training programs effectively enhance social functioning and quality of life in schizophrenia patients while improving the cost-effectiveness ratio of psychological interventions [34]. Complementing these findings, Rotondi's research has shown that internet-delivered psychoeducation significantly improves positive symptom profiles and illness awareness in this clinical population [38]. Furthermore, Özer's pioneering work on group-based online Acceptance and Commitment Therapy demonstrated dual therapeutic benefits – reducing psychotic symptom severity and enhancing psychosocial functioning in individuals with early psychosis [36]. To our knowledge, the hybrid offline-online MCT intervention for schizophrenia has not yet been implemented in China, and its therapeutic efficacy remains insufficiently validated. This study aims to verify the interventional effects of the hybrid offline-online MCT model on psychiatric symptoms, delusion severity, and insight in patients with schizophrenia, and to explore its clinical applicability. We hypothesize that the hybrid offline-online MCT intervention model can improve patients' psychiatric symptoms, delusion severity, and insight.

Methods

Study design and setting

This was a single-blind, randomized, controlled study conducted from January 2024 to March 2024. All procedures followed the consolidated standards of reporting trial guidelines. We recruited participants from two closed psychiatric wards. Before inclusion in the study, the patients underwent a thorough psychiatric examination conducted by an experienced psychiatrist based on the DMS-5 diagnostic criteria for schizophrenia [43]. Each patient was diagnosed with paranoid schizophrenia by two doctors, exhibiting significant and persistent delusions and hallucinations. All psychiatrist evaluations were supervised, and the study leader directly reviewed all evaluations for reasonableness. The inclusion criteria were as follows: (1) Age 18–60 years, with at least a junior high school education; (2) mini-mental state examination score > 24 points [18], indicating the ability to understand the training content; (3) a total score on the positive and negative syndrome scale (PANSS) ≤ 75 [19]; (4) written informed consent obtained from both the patients and their guardians. The exclusion criteria included the following: (1) High risk of suicide or aggression; (2) severe cognitive impairment; (3) serious physical illness or substance abuse, including alcohol.

Sample size estimation

This study used G-Power 3.1.2 for the sample size calculation. The effect size was 0.25, the test efficacy (1-β) was 0.80, and the significance level (α) was 0.05 for the two-tail test with 2 groups of samples and 3 measurements. The calculated required sample size for each group was 44, and considering the 10% loss of follow-up rate, at least 50 samples were required for each group. Participants were randomly assigned in a 1:1 ratio to the intervention and control groups, with randomly generated numbers placed in sealed, opaque envelopes delivered by the research assistant to patients who met the inclusion criteria.

Measures

The assessments before and after the intervention were conducted by assessors unaware of the participants’ group assignments. All assessors were board-certified psychiatrists who underwent standardized training protocols prior to study commencement. To ensure that raters remained blind throughout the study, the research assistants reminded participants not to disclose information after the intervention began and during follow-up assessments.

PANSS [19]

The PANSS was divided into three subscales—the positive scale (7 items), the negative scale (7 items), and the general psychopathology scale (16 items), for a total of 30 items. Scores range from 1 to 7, with higher scores indicating more severe mental symptoms. The Cronbach's α coefficient of the scale was 0.73–0.83, and the retest reliability was 0.77–0.89, the scale demonstrated good reliability.

Chinese version of characteristics of delusions rating scale (C-CDRS) [20]

A total of 11 entries were scored on a scale of 1–10, with 1 representing "not believing at all" and 10 representing "fully believing." The higher the score, the more severe the delusional symptoms. The retest correlation coefficient of the C-CDRS total score was 0.85, and the Cronbach's α coefficient of the C-CDRS total score was 0.83, the scale demonstrated good reliability.

Psychotic symptom rating scale (PSYRATS) [8]

The PSYRATS comprised 17 items: 11 items assessed auditory hallucinations, and 6 items assessed delusions. Scores ranged from 0 to 4, with higher scores indicating more severe auditory hallucinations or delusions. Cronbach's α value for the auditory hallucination scale was 0.978, while for the delusion scale, it was 0.937. The Cronbach's α coefficient of the total quantity table is 0.943, the scale demonstrated good reliability.

Insight and treatment attitudes questionnaires (ITAQ) [29]

The ITAQ comprised 11 items, where 0 indicates no self-knowledge, 1 denotes partial self-knowledge, and 2 implies complete self-knowledge. A higher total score reflects a more comprehensive self-knowledge of the patient. The retest reliability of the scale was 0.869, and Cronbach's α coefficient ranged from 0.801–0.923, the scale demonstrated good reliability.

Intervention group training program

The intervention group underwent MCT twice a week for eight weeks (45–60 min for each training session). The MCT sessions strictly adhered to the MCT manual for schizophrenia. Each group was led by a psychotherapist and comprised 6–10 participants. Members of the intervention underwent MCT courses (MCT online training is available at www.uke.de/e-dmct). To ensure the quality of the intervention, members of the intervention group were regularly supervised by senior psychotherapists with extensive experience in MCT. The psychotherapist was responsible for implementing two phases of the standard MCT's eight modules (MCT training modules are freely available at www.uke.de/mkt). All MCT modules are presented as slides with all relevant information. During hospitalization, the first stage was conducted offline, the training was taught collectively in the rehabilitation treatment room. After discharge, the second stage was conducted online. Participants received two reminders before each session. Missing two or more sessions was considered withdrawal. The main contents of the two stages were identical; however, the pictures and example scenes varied between the stages. To ensure patients could operate the platform correctly, researchers provided a manual before discharge, guiding them in its use to prevent delays in training. The online training was conducted through the Henan Province Mental Health Rehabilitation Cloud Platform, which recorded each participant's login time, module completion status, and exercise completion rate. The research team reviewed the data weekly to ensure that participants completed the training as scheduled. The content of the online training was consistent with the offline training, including modules such as cognitive bias correction and thought monitoring. Each module comprised interactive exercises and feedback sessions. Due to the open format of the MCT, participants can join the metacognitive group training at any time and complete the related exercises. A psychiatric nurse was assigned to assist the rehabilitation training therapist in both offline and online training to ensure the intervention effect. All participants were assessed at three-time points: baseline (T0), 4 weeks after the initiation of training (T1), and upon completion of the training (T2, week 8). The main contents and methods of MCT training are highlighted in Table 1.

Table 1.

Main contents and methods of MCT

	Theme	Main contents and methods	Turn to everyday life
Week One	Attribution	Attribution: Inferring the cause of an event. Exercise: Consider the possible causes of things, and do not be too quick to accept the only explanation	Always consider multiple factors that could have contributed to the outcome of a particular event (for example, yourself/others/circumstances)
	Jumping to Conclusions I	Jump to conclusions: You make judgments based on complex information in your environment. Exercise: 1. Try to find all the details in the picture (the task is to avoid making hasty decisions)	Learning has revealed that jumping to conclusions often leads to mistakes and only partial discovery
Week Two	Changing Beliefs	People tend to have preconceptions. Exercise: Three pictures provide scenes, giving different interpretations	Seek out as much information as possible to validate your judgment
	To Empathize I	Empathy: How to feel the emotions of others. Practice: 1. List some basic human emotion/mood. 2. Give part of a picture and discuss with group members how confident they are. 3. The picture will be presented in reverse order. Ask which is more logical. 4. Decide which outcome is most likely to end the story	Facial expressions and posture are important clues to guessing a person's feelings, but they can also be misleading. The more information you consider, the more likely you will make the right judgment
Week Three	Memory	Memory: Our ability to retain information is limited. Practice: Give a series of complex scenes and try to find forgotten details; discuss what you saw	Do not be too sure your memories are true; find additional information
	Empathize II	Empathy: An individual's perception and response to others. 1. Discuss the advantages and disadvantages of using the following characteristics to judge people. 2. Present comic strips and discuss possible ideas for the characters in the story	Many misunderstandings arise from people misguessing what others think. Try to evaluate it from different angles
Week Four	Jumping to Conclusions II	Jump to conclusions: Jumping to conclusions without knowing all the facts. Exercise: In a series of oil paintings, eliminate inappropriate topics. Discuss the pros and cons of the interpretation	When making a big decision, it is best not to rush into it and consider all the relevant information
	Self-Esteem and Mood	Self-esteem and mood: Poor thinking patterns can lead to the formation and persistence of negative emotions and low self-esteem. Discussion: Delusion affects mood (negative effect/short-term positive effect)	Use tips to reduce negative emotions and thoughts. If the symptoms are severe, contact the doctor in charge

Control group standard care program

Patients in the control group received routine pharmacological treatment and nursing care during their hospitalization. The online health education for the control group was conducted on an individual basis, with a 30-min online session provided weekly through the Henan Province Mental Health Rehabilitation Cloud Platform. The content included identifying adverse drug reactions, observing psychiatric symptoms, and practicing emotional relaxation techniques. Participants could also download relevant educational materials from the platform. After discharge, patients were followed up via telephone once a month to remind them to adhere to their medication regimen and to engage with the health education content pushed by the cloud platform.

Statistical analysis

Data were analyzed using the Statistical Package for the Social Sciences software for Windows (version 27.0; IBM Corporation). Continuous variables are expressed as mean ± standard deviation ($\overline{x}$ ± SD), and categorical data are presented as numbers and percentages. A preliminary check for normality and the mean square error of the data was performed before each analysis. To compare the demographic data of the experimental and control groups, appropriate statistical tests were used based on the data distribution characteristics, including χ² test, two-tailed t-test, and Mann–Whitney U test. The measurement data of the scale scores of the two groups before and after the intervention were measured using repeated measurement analysis of variance (ANOVA), and a paired comparison at varying time points was performed using simple effect analysis. The difference between the two groups at different time points in the same group was statistically significant, with P < 0.05.

Results

Finally, 48 participants in the control group (1 case terminated due to relapse and 1 case uncontactable) and 46 participants in the MCT group (2 cases uncontactable, 1 case terminated due to relapse, and 1 case lost to follow-up) completed the study. Refer to Fig. 1. Table 2 displays the demographic characteristics of the participants, no statistically significant difference was observed in the general information between the two groups (P > 0.05).

Fig. 1.

Recruitment and participant flow

Table 2.

Patient demographic characteristics

Variable	MCT group (n = 46) n (%)	Control group (n = 48) n (%)	t/$\chi$ 2/z	P
Age (years,$\overline{x}$± SD)	34.19 ± 8.11	34.54 ± 7.07	0.22	0.826
Sex
Male	26 (56.5)	23 (47.9)	0.69	0.404
Female	20 (43.5)	25 (52.1)
Education level
Below middle school	3 (6.5)	2 (4.2)	1.41	0.157
Middle school	15 (32.6)	14 (29.2)
Middle school	28 (60.9)	32 (66.6)
Occupation
Student	1 (2.1)	2 (4.2)	2.07	0.558
Farmer	18 (39.1)	24 (50.0)
Office clerk	16 (34.8)	15 (31.3)
Other	11 (23.9)	7 (14.6)
Marital status
Unmarried	11 (23.9)	14 (29.2)	0.87	0.642
Married	24 (52.2)	26 (54.2)
Divorced	11 (23.9)	8 (16.6)
Length of schizophrenia/year
< 2	5 (10.8)	3 (6.2)	1.34	0.512
2 ~ 10	20 (43.5)	26 (54.2)
> 10	21 (45.6)	19 (39.6)
Number of admissions
1	26 (56.5)	31 (64.5)	0.65	0.514
2 ~ 3	17 (36.9)	13 (27.1)
> 3	3 (6.5)	4 (8.3)
Drug type
Olanzapine	14 (30.4)	12 (25.0)
Risperidone	12 (26.1)	14 (29.2)
Aripiprazole	8 (17.4)	7 (14.6)
Quetiapine	12 (26.1)	15 (31.2)
Conversion to olanzapine mg/d	17.60 ± 2.73	17.08 ± 2.88	0.90	0.368

The baseline PANSS, C-CDRS, PSYRATS, and ITAQ scores of the two groups are presented in Table 3. Independent sample t-test results revealed no significant difference in PANSS positive subscale (t = 0.211), negative subscale (t = 0.994), general psychopharmacology subscale t = 0.267) and total score (t = 0.528), C-CDRS score (t = 0.522), PSYRATS auditory hallucination subscale (t = 0.302), delusion subscale (t = 1.043), total score (t = 0.436), and ITAQ score (t = –0.160) between the two groups at baseline (P > 0.05).

Table 3.

Repeated measure ANOVA results comparing outcome variables between intervention and control groups

Outcomes	Group	T0	T1	T2	F value
					Group × time effect	Time effect	Group effect
PANSS score					16.461* * *	2432.208* * *	22.110* * *
	Intervention	98.02 ± 6.46	58.47 ± 3.39	52.19 ± 4.42
	Control	97.29 ± 6.92	60.97 ± 2.62	59.06 ± 2.89
Positive Syndrome					22.727* * *	626.784* * *	41.970* * *
	Intervention	26.69 ± 4.45	14.00 ± 2.25	9.47 ± 1.74
	Control	26.50 ± 4.51	16.39 ± 1.77	15.04 ± 2.31
Negative Syndrome					0.579	881.898* * *	1.171
	Intervention	23.32 ± 3.22	12.08 ± 1.34	11.97 ± 1.29
	Control	22.62 ± 3.59	11.95 ± 1.28	11.83 ± 1.19
General psychopathology					1.988	1323.748* * *	3.582
	Intervention	48.00 ± 3.11	32.39 ± 1.69	30.73 ± 3.56
	Control	48.16 ± 2.94	32.62 ± 1.67	32.18 ± 1.74
C-CDRS					21.778* *	1079.315* * *	64.760* * *
	Intervention	65.17 ± 8.52	28.23 ± 4.39	19.71 ± 4.00
	Control	64.25 ± 8.63	33.66 ± 3.78	30.89 ± 4.55
PSYRA score					21.457* * *	1200.260* * *	20.661* * *
	Intervention	30.93 ± 3.16	16.69 ± 2.44	13.80 ± 2.07
	Control	30.64 ± 3.25	17.95 ± 2.16	17.79 ± 1.93
Auditory hallucination					0.062	356.872* * *	0.146
	Intervention	15.02 ± 2.36	9.47 ± 1.58	9.30 ± 1.56
	Control	15.16 ± 2.29	9.47 ± 1.59	9.45 ± 1.59
Delusion					47.123* * *	1019.084* * *	57.495* * *
	Intervention	15.91 ± 1.79	7.21 ± 1.59	4.50 ± 1.18
	Control	15.47 ± 2.20	8.47 ± 1.33	8.33 ± 1.13
ITAQ					110.593* * *	847.952* * *	145.973* * *
	Intervention	7.04 ± 1.24	12.86 ± 1.49	17.00 ± 1.03
	Control	7.08 ± 1.16	11.87 ± 1.45	11.85 ± 1.42

^*P < 0.05

^{* *}P < 0.01

^{* * *}P < 0.001

The results of the repeated measures ANOVA for the PANSS total score and each subscale demonstrated that the positive and negative subscales, the general psychopathology subscale, the main effect of total score time, and the main effect and group interaction of positive subscale and total PANSS score were statistically significant. The interaction between the PANSS negative and general psychopathology subscales was statistically non-significant. A post-mortem examination revealed that the general psychopathological subscale and total score of PANSS positive and negative subscales in MCT and control groups were lower than baseline scores at the first and second stages. The positive and total scores of the PANSS positive subscale and total scores at baseline, the first and second stages, and the MCT group were lower than those of the control group. A simple effect analysis of the interaction between the PANSS positive subscale and total score demonstrated that the difference in PANSS scores before and after MCT (t = –13.106) and control groups (t = 8.943) was statistically significant (P < 0.016).

The results of the repeated ANOVA of the CDRS total score revealed that the time main effect, group main effect, and group interaction were statistically significant. post hoc analysis indicated that the scores of MCT and control groups were lower than baseline scores in the first and second stages. Additionally, the scores of the MCT group were lower than those of the control group at baseline and the first and second stages. The simple effect analysis demonstrated that the C-CDRS scores in MCT and control groups were statistically significant (t = –12.612, P < 0.016).

The repeated measurement ANOVA results of the PSYRATS total score and each subscale demonstrated that the PSYRATS auditory illusion subscale, delusion subscale, the main effect of total score time, and the main effect of the PSYRATS delusion subscale, total score group, and group interaction were statistically significant. No statistically significant interaction was observed between the PSYRATS auditory hallucination subscale scores. post hoc analysis revealed that the PSYRATS auditory hallucination scale, delusion scale, and total score in MCT and control groups were lower than the baseline score in the first and second stages. The PSYRATS delusion scale and total score in the MCT group were lower than the baseline score in the first and second stages. Simple effect analysis on the interaction between the PSYRATS delusion scale and total score revealed a statistically significant difference in the PSYRATS score (P < 0.016) between MCT (t = –15.992) and control groups (t = –9.647).

Repeated ITAQ total score ANOVA revealed that the time main effect, group main effect, and group interaction were statistically significant. The post hoc analysis demonstrated that the scores of MCT and control groups were higher than the baseline scores in the first and second stages, and the scores of the MCT group were higher than the baseline scores in the first and second stages. Simple effect analysis revealed that the difference in ITAQ scores between MCT and control groups was statistically significant (t = 19.936, P < 0.016).

Discussion

This study represents the first implementation of hybrid offline-online group MCT training in patients with schizophrenia, partially confirming our hypotheses. In this study, the positive symptoms of patients in the MCT group were better than those in the control group during weeks four and eight of training. The total scores and subscales of PANSS in MCT and control groups were lower than baseline scores at weeks four and eight of training, indicating that the training effect of the MCT group was better than that of the control group. Schizophrenia treatment emphasizes a whole-course integrated intervention model, and many people with schizophrenia spectrum disorders have metacognitive deficits, such as impaired social cognition and severe deficits in reasoning, abstract thinking, and problem-solving [26]. The results of this study suggest that this training can improve patients' positive symptoms, likely because group MCT helps patients better understand their thinking and cognitive bias [15]. Through rehabilitation training, therapists can help patients better monitor their thinking processes, assisting them in considering multiple possibilities for unsubstantiated ideas when problems occur [33]. Using pictures, examples, and other methods, patients can learn cognitive transfer skills, correct inappropriate thoughts, and challenge and change negative cognitive patterns [11]. Through reflection and thinking exercises both offline and online, patients can gradually become aware of the irrationality of their paranoid thinking and improve self-insight, thereby reducing positive symptoms. Group training can provide a supportive environment for patients [5], where they can share experiences and feel understood and supported by others. Ongoing social support has been demonstrated to positively affect mental health. It can effectively reduce loneliness and the intensification of paranoid thinking [12], helping to reduce positive symptoms, such as suspicion, conceptual confusion, and delusions. Confirmed that integrating metacognitive information, such as reasoning ability and insight, reduces the symptoms of schizophrenia [4]. Research highlighted that metacognitive therapy can promote the health and recovery of schizophrenia, which paralleled the results of this study [1].

Delusions constitute a cardinal symptom of schizophrenia. This study revealed that the MCT group was superior to the control group in delusional degree at four and eight weeks of training. The total C-CDRS score, PSYRATS score, and all subscales in MCT and control groups were lower than the baseline scores at four and eight weeks of training, indicating that the training effect of the MCT group was superior to that of the control group. The analysis of the cause may involve attribution in the module, informing the patient that the same situation may have completely different reasons, rather than taking one aspect as an important factor, and then improving the degree of delusion through the positive guidance of the rehabilitation therapist on the attribution, such as the misunderstanding caused by attribution bias. In the hasty conclusion module, patients are trained to avoid making hasty judgments based on their surrounding environment, and there may be more explanations through examples; accordingly, they should use evidence to confirm the facts [33]. In the belief change module, patients are trained to avoid preconceived thoughts, and the information obtained from only one channel is not absolutely correct. Consequently, they should make judgments by verifying as much information as possible. Patients can better monitor their automatic thought patterns, thereby reducing the intensity of delusions. Both Han [14] and Chen Qi (Chen et al., 2021) demonstrated that metacognitive intervention can reduce delusions in schizophrenia symptoms. Dual learning modes are combined online and offline, especially for patients who cannot participate regularly due to distance, transportation, or physical limitations, thereby improving access and effectiveness of treatment [37].

Using Internet technology to deliver health knowledge to patients with schizophrenia can improve their treatment compliance [45]. This study revealed that the MCT group demonstrated improved self-knowledge and attitudes compared to the control group at four and eight weeks of training. Besides, the total ITAQ score was higher than the baseline score, indicating that MCT enhanced patients' understanding of their thinking and cognitive processes. It also helped them develop a more positive thinking style and attitude by identifying and changing these negative cognitive patterns, allowing patients to fully understand their strengths and weaknesses and learn to actively respond to improve their self-knowledge and attitude [24]. The MCT emphasizes improving patients' awareness of their mental state, helping them to identify and monitor their thinking process [25]. Through a combined offline and online mode, patients can experience the opinions of others in a safe social environment, improve their self-reflection ability, and thus enhance their self-knowledge. Lam revealed that MCT training can improve the self-awareness and attitude of patients with schizophrenia [22]. Roux indicated that MCT training improves medication compliance in patients with schizophrenia [39]. Lysaker revealed that metacognitive ability is significantly correlated with the need for treatment, which paralleled the results of this study [27].

The MCT group demonstrated no statistically significant differences in scores on the negative symptom subscale, general psychopathology subscale, and auditory hallucination subscale. This null finding may be attributed to inherent limitations of the online therapeutic modality [30], where restricted emotional transmission by therapists and diminished group therapeutic dynamics likely attenuated intervention efficacy for negative symptoms and general psychopathology in schizophrenia. Contrary to evidence from Kumar and Venkatasubramanian showing that combined metacognitive and mindfulness therapy reduces auditory hallucinations in schizophrenia [21], our study failed to replicate these effects. This discrepancy might stem from the complex psychopathology underlying schizophrenic auditory hallucinations. As MCT primarily targets cognitive distortion identification and modification [11], its effectiveness may be constrained in addressing hallucinatory phenomena that are frequently triggered by real-world stressors or emotional fluctuations [28]. The online format further limited therapists' capacity to promptly detect and contextually intervene in these dynamic situational triggers, potentially compromising the intervention's impact on hallucination severity.

Limitations

As a pilot study investigating the feasibility of hybrid offline-online MCT training, this research did not establish independent online or offline MCT control groups. While the hybrid group exhibited significant symptomatic improvements, the current design precludes differentiation of therapeutic effects stemming from offline interactions, online exercises, or synergistic mechanisms. Future investigations should employ factorial designs to delineate intervention components while standardizing total intervention duration to control for dose–effect relationships. The exclusive recruitment of participants from a single hospital necessitates subsequent multicenter studies with larger samples to enhance generalizability. Notably, inherent limitations of the online modality were observed, particularly constrained emotional transmission and attenuated interactive feedback between therapists and patients during group sessions. To address these challenges, future implementations could integrate wearable monitoring devices to enhance interactivity and engagement in virtual group therapy. Furthermore, critical long-term prognostic indicators—including functional recovery and quality of life metrics—were not incorporated in outcome assessments. Additionally, this study omitted evaluation of patients' experiences with the online platform. Subsequent research should systematically assess user perceptions to optimize platform functionality and improve therapeutic adherence. These underexplored dimensions warrant prioritized investigation in future studies.

Conclusions

This randomised controlled trial demonstrated that hybrid offline-online MCT intervention significantly improved positive symptoms, delusion severity, and illness insight in patients with schizophrenia, supporting its accessibility as a complementary intervention to pharmacotherapy. Future studies should establish independent control groups to enable more rigorous examination of intervention effects.

Abbreviations

MCT

Metacognitive Training

PANSS

Positive and Negative Syndrome Scale

C-CDRS

Chinese version of the delusion feature rating scale

PSYRATS

Chinese version of the psychotic symptom rating scale

ITAQ

Insight and treatment attitude questionnaire

Authors’ contributions

Y Z and S H contributed to the conception and design, and wrote the original draft of the manuscript. W H, W C, F Y, Y Q, J C, Y Z and X Z participated in the revision of the paper, literature review and analysis, and table creation. All authors contributed to the article and acknowledge the submitted version.

Funding

This study received major results from the Henan Province science and technology research and development plan joint fund (industry) project (235101610004), Open topic of Collaborative Innovation Center for Prevention and Treatment of mental illness in China Henan Province (XTkf12), Henan Province medical science and technology research project (RKX202302024) The open project of Henan Province psychology in the second Affiliated Hospital of Xinxiang Medical College (XYEFYJSSJ-2023–01). Funders played no role in the design of the study, the writing of the report, and the decision whether to submit the paper for publication.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study passed the ethical approval of the Ethics Committee of the Second Affiliated Hospital of Xinxiang Medical College, and the study followed the principle of "informed consent" to ensure that participants were informed of the purpose of the study. All patients participating in the study received written informed consent. In addition, participants were informed that their ability to opt out of the study was protected and guaranteed confidentiality and not to use their information elsewhere.

Consent for publication

Not Applicable.

Competing interests

The authors declare no competing interests.