Metacognitive Training (MCT) to improve insight and work outcome in schizophrenia

Abstract

Individuals with psychotic disorders have deficits in metacognition. Thirty-four adults with schizophrenia were randomized to 2 months of Metacognitive Training (MCT) or a Healthy Living Skills control group. All participants were enrolled in a work therapy program, followed by a supported employment (SE) program. Assessments were conducted at baseline, at the end of the 2-month active intervention, and at four and twelve month follow-ups. At the end of active intervention, the MCT group demonstrated greater improvement and better work behavior relative to controls. At follow-up, the MCT group demonstrated significantly greater insight, fewer positive symptoms and a greater percentage were employed in the community. We speculate that being better able to think about one’s thoughts, recognize biases in thinking, and correct those thoughts, may aid in responding to workplace challenges and hence improve work outcomes.

Introduction

Researchers have suggested individuals with schizophrenia have deficits in metacognition. Metacognition, sometimes also called mentalizing, refers to an individual’s awareness and understanding of their own thoughts and the thoughts of others as well as awareness about their own mental deficits and biases (for a review on conceptualizations of metacognition, see Mortiz & Lysaker, 2018). Impairments in metacognitive capacity among individuals with psychoses were suggested by Bleuler (1950) and Freud (1940), with more recent research providing greater clarity to the underlying construct and provided methods of measurement (Frese, 2000; Lysaker, Wickett, Wilke, & Lysaker, 2003; Frith, 1992). Neither symptoms nor neurocognition fully account for deficits in mentalizing, supporting the view that metacognitive deficits represent a distinct domain (Mortiz & Lysaker, 2018).

One line of research ties metacognition to the awareness of cognitive biases (see Moritz, Pfuh, Ludtke, Menon, Balzan & Andreou, 2017). Herein, research suggests individuals with psychosis, as compared to healthy controls, demonstrate an elevated tendency to look for cues or facts that support their beliefs (i.e., confirmation bias), and are less likely to consider cues or facts that counter their beliefs (i.e., bias against discriminatory evidence). These individuals, in turn, demonstrate overconfidence in their thinking and are less likely to notice errors (e.g., jumping to conclusions, making false judgments, paranoid ideation).

Additional researchers exploring different components of metacognition have highlighted the association between limited metacognitive capacity and impaired psychosocial functioning (Bell, Tsang, Greig & Bryson, 2009; Brune, Abdel-Hamid, LehmKamper & Sonntag, 2007; Gagen, Zalzala, Hochheiser, Schnakenberg, & Lysaker, 2019; Lysaker et al., 2009). In one study, Lysaker and colleagues (2009) explored the relationship between self-reflectivity, “the ability to be aware of and connect one’s own thoughts and feelings with one another and their antecedents (p.125),” and work performance. They found that those with limited self-reflectivity had poorer work performance outcomes when compared to those with greater capacity to self-reflect. Consequently, they theorized that metacognitive impairments may serve as a barrier to recovery for individuals with psychosis. With a limited ability to form and scrutinize thoughts about oneself and others, it can be difficult for individuals with schizophrenia to make meaning of their experiences and be active agents in their life (Corcoran & Frith, 2003; Lysaker, Buck & Roe, 2007).

In light of this literature, there has been a need to develop and validate interventions aimed at improving metacognitive skills. An emerging approach developed by Moritz and colleagues (2007) targets insight into cognitive biases, such as overconfidence in errors and jumping to conclusions, among individuals with psychosis. Specifically, Metacognitive Training (MCT; Moritz et al., 2007) is a group intervention that provides individuals with knowledge regarding cognitive distortions and raises their awareness around the negative implications these biases may have in their daily functioning. Several empirical studies found MCT has contributed to improvements in symptoms (for most recent meta-analyses see: Lui, Tang, Hung, Tsai & Lin, 2018; Philipp, Kriston, Lanio, Kuhne, Harter, Moritz, Meister. 2019; Eichner & Berna, 2016), insight into delusional thought processes (Favrod, Marie, Bardy, Pernier & Bonsack, 2011), cognitive biases (Aghotor, Pfueller, Moritz, Weisbrod, & Roesch-Ely, 2010; Ross et al., 2011; Moritz, et al., 2011), and quality of life (Moritz, et al., 2014). To date, no studies of this training have evaluated its impact on vocational function for individuals with schizophrenia. In this exploratory pilot, we sought to examine the relationship between metacognitive training and illness insight, symptoms, work therapy, and supported employment outcomes.

Methods

Participants

Participants were 34 adults, ages 18–55, with a chart diagnosis of schizophrenia or schizoaffective disorder enrolled in an intensive outpatient rehabilitation program in the New England region of the United States of America. Prior to enrollment in the study, participants had to be psychiatrically stable for at least 30 days on their psychotropic medication. Diagnosis of schizophrenia or schizoaffective disorder was confirmed by the Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-IV; First et al., 1997). Any participants with significant auditory/visual impairment, lack of fluency in English or severe neurological disorders, were excluded. Participants who met criteria for current substance abuse/dependence were also excluded, as were those who had active substance abuse within the 30 days prior to joining the study.

As part of their standard outpatient treatment, all participants were enrolled in a work therapy program (Bell et al., 1996). Work therapy consisted of 2–6 supervised and structured hours per week for twelve weeks at sites around the psychiatric campus (e.g., gift shop, cafeteria, greenhouse, or environmental services). Following work therapy, participants enrolled in a supported employment (SE) program.

At the start of work therapy, participants were randomized to 2 months of concurrent group therapy that met once a week for 50 minutes: Metacognitive Training (MCT; Moritz & Woodward, 2007), a group-based therapy aimed at enhancing insight into cognitive biases underlying paranoia and delusions, or a comparator group focused on learning healthy living skills such as better nutrition and eating habits. Assessments of insight, work readiness, and symptoms were administered by a research assistant blind to group assignment. Assessments were administered at baseline (prior to the start of work therapy), at the conclusion of the 2-month group therapy program, and at 4-month follow-up after entry into the SE program. Employment outcomes were collected 12 months after entry into the SE program.

Interventions

Metacognitive Training (MCT) for psychosis is a dialectic group-based program that combines CBT, compensatory cognitive therapy, and psychoeducation (Moritz & Woodward, 2007). The intervention is delivered across eight sessions, each lasting approximately 45–60 minutes. The eight modules target insight and understanding of attribution bias, jumping to conclusions, empathy, memory strategies, mood, and self-esteem. Each module begins with an introduction to the target domain, reviewing current scientific findings and connecting these findings to psychosis and overall experiences of daily living. Then, participants are presented with exercises, highlighting the relevance for psychosis, before concluding with a case example of delusional thinking. The focus of the group is to explore common cognitive biases, understand the implications of holding these beliefs, and explore ways in which one might challenge or alter these thoughts. The comparator group was a group-based program designed to teach the value of nutrition and better eating habits. MCT was administered by a psychologist while the healthy living group was conducted by a licensed social worker.

Measures

Symptom insight was measured using the Scale to Assess Unawareness of Illness (SUMD; Amador & Strauss, 1993), while work readiness was measured using the Work Behavior Inventory total score from all its five scales: cooperativeness, work habits, work quality, social skills and personal presentation (WBI; Bryson et al., 1997). Symptom ratings were collected using the Brief Psychiatric Rating Scale (BPRS; Overall & Gorham, 1988; Ventura et al, 2000). A trained research assistant calculated ratings across these measures after conducting a thorough clinical interview and interview with the participant’s work supervisor. At 12 months, we gathered data regarding current employment status (i.e., employed or unemployed) for each participant through the SE counselor and/or through the Connecticut Bureau of Rehabilitation Services.

Data Analysis

To ensure that the two experimental groups were similar on baseline demographic and clinical variables, we compared these baseline variables across the groups via one-way analyses of variance (ANOVA).

To assess the impact of MCT on insight, work behavior, and symptoms, we computed the change in scores from baseline, conclusion of the group intervention at 2 months, and at follow-up 4 months from baseline for each participant on the SUMD, the WBI and BPRS. We then entered these change scores into one-way ANOVAs, one for SUMD, another for WBI, and one for BPRS, with experimental condition as the between-subjects variable. At 12 months from baseline, we tallied and compared the frequency of participants who were enrolled in community employment at that time. All statistical tests were two-tailed and alpha was set at .05.

Results

Demographic information and clinical characteristics of each group are summarized in Table 1. There were no significant group differences in demographic or clinical characteristics. There were also no significant group differences in baseline insight, work readiness (WBI total), and symptoms (BPRS total).

Table 1.

Demographic and clinical characteristics of the two groups along with baseline to post to follow-up scores.

	Metacognitive Training n=16	Healthy Living Group n=17	F-test	ANOVA (p value)	Effect size (Cohen’s d)
Age (years)	31.50 (6.06)	32.27 (6.28)	0.205	0.654	--
Education (years)	12.56 (2.23)	12.04 (2.28)	0.438	0.513	--
Male (%)	56	53	--	--	--
Race   Black-African American   Hispanic-Latinx   Non-Hispanic White	4 4 8	3 6 10	-- -- --
Duration of Illness (yrs)	10.85 (5.71)	9.13 (7.80)	0.517	0.478	--
Percentage on atypicals	69	75	--	--	--
Schizoaffective (%)	50	53	--	--	--
Insight (SUMD)  Baseline  Post  Follow-up 1	9.94 (3.69) 10.69 (2.70) 11.75 (2.13)*	9.75 (3.78) 10.36 (2.96) 10.04 (2.55)	0.021 0.111 4.342	0.885 0.740 0.046	-- 0.116 0.728
Work behavior (WBI)  Baseline  Post  Follow-up 1	100.83 (15.85) 116.10 (7.81)* --	103.98 (13.50) 108.97 (10.48) --	0.379 4.861 --	0.543 0.035 --	-- 0.771 --
Symptoms (BPRS)  Positive Factor   Baseline   Post   Follow-up 1  Negative Factor   Baseline   Post   Follow-up 1	33.36 (7.20) 31.30 (6.18) 29.27 (6.08)* 18.70 (5.15) 16.36 (5.47) 16.83 (6.80)	33.85 (8.07) 32.61 (8.72) 32.49 (8.86) 17.93 (6.94) 15.06 (4.39) 16.20 (5.17)	0.033 0.245 4.380 0.130 0.569 0.090	0.856 0.624 0.044 0.721 0.456 0.766	-- 0.173 0.423 -- 0.261 0.104
Community employment   Follow-up 2	7/16	2/17	Chi-sq=4.251	0.039	--

^*p <0.05

As shown in Table 1, at post, while both groups improved significantly in work readiness, the MCT group demonstrated greater improvement in work behavior than the healthy living group. Although there were no significant changes in insight or positive symptoms in either group at the conclusion of the group intervention, the MCT group demonstrated greater improvement in insight and fewer positive symptoms at follow-up. Regarding employment outcome, a significantly greater percentage of those who received MCT were employed in the community at least on a part-time basis relative to controls.

Discussion

Regarding psychosocial outcomes, results highlighted a significant improvement in work behavior for those in the MCT group when compared to the healthy living skills condition. Our preliminary results align with existing literature that suggests an individual’s ability to engage in self-reflection is positively related to work performance (Brune et al., 2007; Lysaker et al., 2009). MCT training may have increased the participants’ ability to think about and correct their own thinking and that this ability generalized to work-related matters. They may have been able to recognize biases in their thinking and found new ways to think about and respond to their workplace challenges. Conversely, limited capacity to identify these biases may have served as a barrier to work performance improvement.

Additionally, findings indicated 12 months after completing treatment, a greater percentage of those in the MCT group were employed in the community. These findings align with existing literature that suggests improvements in metacognition contribute to an individual’s capacity to act as an agent in their life (Lysaker, Buck & Roe, 2007). We speculate that the difference in employment outcomes may be explained by an improved ability to make meaning of and incorporate experiences into a more cohesive, integrated, sense of self and use this understanding to guide future decision-making. As such, with improved self-reflection, an individual may develop a more complex understanding of the self, including the self as an employee. In turn, they may be better able to use this new information about the self to seek out employment in the community. In other words, as an individual’s metacognition increases as does their ability to self-reflect. Those in the MCT group may have developed a more refined sense of self, including a greater understanding of their workplace strengths and weaknesses. With such information, these individuals may have been better able to seek out and select jobs that are a better fit for them.

Finally, findings indicated the MCT group demonstrated greater improvements in insight and symptomatology, when baseline scores were compared to follow-up 1 (4 months). Although these improvements were not present immediately after the conclusion of treatment (post), these findings highlight that skills learned in cognitive and work therapy may need time to consolidate before an individual is able to identify benefits. This finding aligns with previous studies that have also found positive impacts of MCT on psychosocial functioning sometime after the end of the active intervention (Moritz et al., 2014). Perhaps improvements such as those found in this study (i.e., improved insight, reduced symptoms) are only realized with a greater awareness of one’s cognitive biases and increased opportunity to practice what is learned in MCT.

There are several limitations to this study. First, the study sample size was small and all participants were enrolled in a work therapy program, intensive outpatient treatment, and medication management, limiting the generalizability of our findings. Future studies may consider replicating this study with a larger sample, representing diversity in social identities, treatment experience, and duration of illness (i.e., prodromal, first episode, chronic psychosis). Second, we did not measure metacognition in this study. Though literature has provided evidence that MCT improves metacognitive capacity (see Köther et al., 2017), future studies may consider measuring changes in metacognition both immediately after, as well as some time after completing the training. Mediation analyses could then be conducted to determine to what extent any improvements are tied to changes in metacognition. Third, we did not assess participants’ performance at their jobs in the community. Future studies may consider addressing long-term implications of this training on competitive employment. Fourth, we did not incorporate a measure that addresses potential changes in metacognition. Future studies may consider incorporating measures such as the MAS-A (Semerari et al., 2003) into their study in order to shed light on target engagement (i.e., is MCT improving the target domain). MCT addresses one area of metacognition; therefore, future studies may consider exploring interventions that target similar constructs such as mentalizing, theory of mind, or decentration.

Conclusions

This study provided additional evidence that improvements in metacognitive capacity may contribute to increased insight into illness, reduction in symptomatology, and improvements in vocational outcomes among individuals with schizophrenia.