Sauve Et Al. 2020

Clinical Psychology Review 78 (2020) 101854
Contents lists available at ScienceDirect
Clinical Psychology Review

journal homepage: www.elsevier.com/locate/clinpsychrev
Review
Efficacy of psychological interventions targeting cognitive biases in T

schizophrenia: A systematic review and meta-analysis
Geneviève Sauvéa,b, Katie M. Lavignea,c,d, Gabrielle Pochieta,e, Mathieu B. Brodeura,
⁎
Martin Lepagea,c,
a
Douglas Mental Health University Institute, Montreal, Canada
b
Department of Psychology, Université du Québec À Montréal, Montreal, Canada
c
Department of Psychiatry, McGill University, Montreal, Canada
d
Montreal Neurological Institute, Montreal, Canada
e
Integrated Program in Neuroscience, McGill University, Montreal, Canada
H I GH L IG H T S
• Psychological interventions systematically targeting cognitive biases were reviewed.

• Small to moderate effects were found for cognitive biases, positive symptoms and insight.
• Results for cognitive biases may be driven by publication bias and risk of bias.
• Future studies should examine the effects in first-episode and high-risk populations.
A R T I C LE I N FO A B S T R A C T
Keywords: Cognitive biases, which are tendencies to systematically process, select and remember certain information (e.g.,
Thinking errors jumping to conclusions), are exacerbated in schizophrenia and associated with delusions. Here we review and
Psychosis quantitatively assess psychological interventions targeting cognitive biases (e.g., metacognitive training, rea-
Therapy soning training, Maudsley review training programme) to evaluate their efficacy in improving cognitive biases,
Treatment
positive symptoms, and insight. Overall, thirty-two studies, including 15 distinct interventions and 2738 par-
Jumping to conclusions
ticipants, were identified through a comprehensive keyword database search. Meta-analytic effect sizes were
Bias against disconfirmatory evidence
calculated and heterogeneity, publication bias, and subgroup analyses (study bias, active/passive intervention)
were conducted. We observed significant small to moderate beneficial effects of cognitive interventions on
cognitive biases (Hedges' g = 0.27; 95% CI = [0.13–0.41]; z = 3.77; p < .001), positive symptoms (Hedges'
g = 0.30; 95% CI = [0.13–0.48]; z = 3.44, p < .005), and insight (Hedges' g = 0.35; 95% CI = [0.15–0.56];
z = 3.37,p < .005). Interestingly, studies with high risk of bias or passive control condition did not differ
significantly from those with low risk or active control condition, respectively. Thus, cognitive biases are mal-
leable via psychological interventions, which also exert, either directly or indirectly through reduced cognitive
biases, beneficial effects on positive symptoms and insight.
1. Introduction characteristics, individuals with SZ&RP show systematic cognitive

biases, which are not deficits per se but rather tendencies to treat in-
Schizophrenia and related psychoses (SZ&RP) significantly impacts formation differently or adopt an alternative thinking style (Moritz &
psychosocial functioning, quality of life and well-being (Yanos & Moos, Woodward, 2007b). Formally, cognitive biases are conceptualized as a
2007). SZ&RP is primarily characterized by positive (i.e., hallucinations systematic and preferential orientation toward appraising, processing,
and delusions) and negative (e.g., affective flattening, avolition, and selecting and remembering certain information (Grisham, Becker,
anhedonia) symptoms as well as pervasive cognitive impairments Williams, Whitton, & Makkar, 2014; Lester, Mathews, Davison, Burgess,
(Tandon, Nasrallah, & Keshavan, 2009). In addition to these cardinal & Yiend, 2011). On the other hand, cognitive deficits refer to reduced
⁎
Corresponding author at: Douglas Mental Health University, FBC Pavilion, 6875 Blvd. LaSalle, Verdun, Québec H4H 1R3, Canada.
E-mail address: [email protected] (M. Lepage).
https://doi.org/10.1016/j.cpr.2020.101854
Received 26 July 2019; Received in revised form 1 April 2020; Accepted 4 April 2020
Available online 24 April 2020
0272-7358/ © 2020 Elsevier Ltd. All rights reserved.
G. Sauvé, et al. Clinical Psychology Review 78 (2020) 101854
cognitive capacity for which the following seven domains have been given. BADE is defined as a decreased tendency to downrate inter-
found to be impaired and potentially malleable via treatment in SZ&RP: pretations that become implausible as the story progresses, that is, a
speed of information processing, attention/vigilance, working memory, tendency not to incorporate evidence that contradicts a belief. In con-
verbal learning and memory, visual learning and memory, reasoning trast to JTC, which may contribute to the formation of delusions, BADE
and problem solving and social (Nuechterlein et al., 2004). Although is hypothesized to underlie delusion maintenance, in that an unwill-
associations have been found between cognitive biases and cognitive ingness to integrate disconfirmatory evidence may prevent delusional
deficits, principal component analyses have shown that they are se- beliefs from being challenged (Broyd et al., 2017). A related class of
parable constructs (Eifler et al., 2015; Moritz et al., 2010). Also, cor- cognitive bias frequently observed in individuals with SZ&RP is called
relations with positive symptoms have been consistently reported for overconfidence in errors and refers to having unreasonably heigh-
cognitive biases, but are less evident for cognitive deficits (McLean, tened confidence in one's judgement, inferences and predictions
Mattiske, & Balzan, 2017; Moritz, Heeren, Andresen, & Krausz, 2001). (Balzan, 2016; Köther et al., 2017; Moritz et al., 2015).
Further, cognitive biases are common in the general population and Finally, attributional biases represent a family of cognitive biases
often addressed in psychological therapies via cognitive restructuring, wherein patients unjustly and uniquely blame others or external cir-
but tend to be exacerbated and generalized in psychosis/mental illness, cumstances for negative personal events (Salvatore et al., 2012;
and may contribute to symptoms. Research shows they are related to, Savulich et al., 2012). Attributional biases may contribute to positive
but generally distinguishable from, cognitive deficits, such as attention symptoms, especially persecutory delusions, by distorting neutral
and memory impairments (Eisenacher & Zink, 2017). While cognitive events in a negative manner (Bentall, Corcoran, Howard, Blackwood, &
biases are equally observed in non-clinical subjects and across psy- Kinderman, 2001); although this hypothesis requires further empirical
chiatric diagnoses (e.g., obsessive-compulsive disorder; Grisham et al., validation according to Garety and Freeman (2013b)
2014), some (e.g. jumping to conclusions which is defined later on)
have been specifically associated with psychotic symptoms in in- 1.2. Interventions targeting cognitive biases
dividuals with SZ&RP (McLean et al., 2017), at-risk groups (Eisenacher
et al., 2016), and healthy individuals with sub-clinical delusional As key factors in the formation and maintenance of positive symp-
ideation (Balzan, Delfabbro, Galletly, & Woodward, 2013; Menon et al., toms, cognitive biases are being increasingly targeted by novel man-
2013; Woodward, Buchy, Moritz, & Liotti, 2007). This suggests they ualized psychological interventions for SZ&RP. One of the earliest and
may be a cognitive marker of psychosis and/or psychosis proneness most influential interventions targeting cognitive biases is metacognitive
(Eisenacher & Zink, 2017; Lepage, Sergerie, Pelletier, & Harvey, 2007; training (MCT; Moritz & Woodward, 2007b), which teaches individuals
Moritz, Vitzthum, Randjbar, Veckenstedt, & Woodward, 2010; Moritz & about cognitive biases, how they contribute to the positive symptoms of
Woodward, 2007b). The ones that have been most systematically ob- psychosis, and how they can affect daily life. Several MCT variants and
served in SZ&RP patients are presented next. novel interventions drawing from the tenets of MCT have emerged since
MCT was first introduced, and these are reviewed below. Previous
1.1. Cognitive biases meta-analyses on MCT-specific interventions have demonstrated small
to moderate effects on positive symptoms (Eichner & Berna, 2016;
Among biases specifically implicated in SZ&RP, ‘jumping to con- Philipp et al., 2018); however, the first meta-analysis conducted on
clusions’ (JTC) has perhaps received the greatest amount of attention MCT (Van Oosterhout et al., 2016) did not report significant effects on
(Savulich, Shergill, & Yiend, 2012) and refers to the tendency to collect symptoms or data-gathering bias, though this may have been influenced
very little information before reaching a conclusion or making a deci- by overly conservative exclusion criteria, according to Eichner and
sion (Garety & Freeman, 2013a; Ross, McKay, Coltheart, & Langdon, Berna (2016). They notably criticize that 3 positive studies were ex-
2015). A recent meta-analysis (Dudley, Taylor, Wickham, & Hutton, cluded from the Van Oosterhout et al. (2016) study because of alleged
2016) reported JTC in approximately 60% of SZ&RP patients, but only unavailable data, which they argue could have been obtained otherwise
in 38% of individuals with other psychiatric diagnoses and 29% of by statistical calculations or via corresponding authors; given the al-
healthy controls. JTC is measured using a probabilistic reasoning task, ready small number of studies (7 for data-gathering bias and 9 for
such as the traditional “beads task” (or its variant, the “fish task”). In symptoms), this could indeed have an important impact on the study's
the beads task, a coloured bead is drawn from one of two jars, which conclusions.
have different colour ratios (e.g., 85% white, 15% black) and partici- There exist several cognitive interventions other than MCT that
pants are required to determine which of the two is being drawn from target cognitive biases, which have not been included in previous meta-
(Huq, Garety, & Hemsley, 1988). After each drawn bead, participants analyses focusing on MCT alone. Moreover, previous investigations
are asked whether they have made a decision (i.e., from which jar the have used symptoms as the major outcome variable and have not sys-
beads are being drawn from) and how confident they are in their de- tematically validated that these interventions positively affected cog-
cision. The most common outcome measure of this task is the ‘draws to nitive biases as is their intention. Interventions targeting cognitive
decision’ (DTD) index, which is simply the number of beads drawn biases may also exert positive effects on lack of clinical (unawareness of
before a decision was reached. Most often, JTC is operationalized as being ill) and cognitive (self-reflectiveness and self-certainty) insight
making a decision after drawing one or two beads (Moritz et al., 2013; (Andreou et al., 2017; Favrod et al., 2015). Poor clinical insight is
Ross, Freeman, Dunn, & Garety, 2011; So et al., 2015). JTC is hy- frequently observed in SZ&RP (50–80%) and broadly refers to the
pothesized to underlie the formation of delusions in SZ&RP, as it can failure of acknowledging the signs of one's illness because of a difficulty
influence the likelihood of adopting a belief with very little evidence to reflect on one's own thinking (Amador & Kronengold, 2004; Poyraz
(Broyd, Balzan, Woodward, & Allen, 2017). et al., 2016; Vohs, George, Leonhardt, & Lysaker, 2016). Similarly, poor
The bias against disconfirmatory evidence (BADE) refers to the cognitive insight is also widely documented in SZ&RP and is defined as
tendency to disregard evidence that contradicts one's beliefs (Moritz, the ability to reflect upon one's own thoughts and adopt a critical stance
Vitzthum, et al., 2010; Sanford, Veckenstedt, Moritz, Balzan, & toward the validity of one's beliefs (Beck, Baruch, Balter, Steer, &
Woodward, 2014; Speechley, Moritz, Ngan, & Woodward, 2012). This Warman, 2004; Nair, Palmer, Aleman, & David, 2014). The importance
bias is commonly assessed using short three-sentence vignettes where of metacognition (i.e., thoughts about thoughts) in clinical and cogni-
each sentence provides additional information about the situation tive insight suggests that it may be an important secondary target/
(Sanford et al., 2014; Speechley et al., 2012; Woodward et al., 2007). outcome variable for interventions addressing cognitive biases, the
After each sentence, participants rate and re-rate four interpretations of majority of which train metacognition. An increasing number of in-
the story, which become more or less plausible as more information is dependent studies on these interventions have shown promising results
2
in reducing cognitive biases and positive symptoms as well as im- analysis because the number of included studies would have been too
proving clinical and cognitive insight in SZ&RP; however, a systematic low otherwise and due to the high correlation between the two (Beck
evaluation of their efficacy has yet to be published. Hence, the aims of et al., 2004). Subsequent use of the term insight therefore refers to both
the present article were (1) to conduct a systematic review of literature clinical and cognitive insight, unless specified. Sample sizes, means and
on psychological interventions developed to address cognitive biases in standard deviations for pre- and post-treatment measures were ex-
SZ&RP and (2) to evaluate via meta-analysis their efficacy in reducing tracted from the published articles or obtained from the corresponding
cognitive biases and psychotic symptoms, and in improving insight authors. Hedges' g effect size was chosen, in contrast to Cohen's d, in an
(clinical and cognitive). We hope this synthesis and quantitative ex- attempt to correct for small sample sizes (Hedges & Olkin, 1985).
amination of these evidence-based techniques will provide clinicians Hedges' g effect sizes were standardized using the change score standard
and researchers alike with insight into these techniques and their effi- deviation and were calculated for each study from the reported means
cacy as well as provide directions for future research on interventions and standard deviations of both intervention and control groups. For
targeting cognitive biases in schizophrenia. studies that reported multiple follow-up time points (e.g., 3-months, 6-
months follow-up) and outcome measures (e.g., PANSS and PSYRATS as
2. Methods measures of positive symptoms), effect sizes were pooled to obtain a
composite score. When a study failed to report the correlation between
The review protocol for the current study was registered in the their pre-treatment and post-treatment scores, a conservative value of
PROSPERO database (CRD 42017065218) and the PRISMA guidelines 0.7 was adopted, as suggested by Rosenthal (Rosenthal, 1993). When
for systematic and meta-analysis studies were followed (Moher, studies reported outcomes using percentages (e.g., percentage of par-
Liberati, Tetzlaff, Altman, & Group, 2009). The literature search was ticipants showing the JTC bias), the percentage was converted into the
conducted using the MEDLINE, PsycInfo and EMBASE databases on number of participants and used the number of events to compute
May 10th, 2019 with no restriction regarding the year of publication. Hedge's g effect sizes. Hedges's g was interpreted in the following
The following keywords were used: (schizophreni* OR psychosis OR fashion: 0.2 a small effect, 0.5 a medium effect, and 0.7 or greater a
psychoses OR psychotic*) AND (cogniti* OR think* OR reason*) AND large effect. A positive g value indicates an improvement in cognitive
(bias* OR error* OR distort* OR style). The search was limited to ar- biases, a decrease in positive symptoms and increase in insight.
ticles written in English or French. Additionally, the reference lists of all The presence of a publication bias was assessed for each outcome
articles included in the review were searched for additional studies. The using the following methods: visual examination of the funnel plot
MCT developers, Steffen Moritz and Todd Woodward, were also con- (Egger, Smith, Schneider, & Minder, 1997), Egger's asymmetry test, and
sulted to obtain any unpublished data. the fail-safe N of Rosenthal (Rosenthal, 1979). If publication bias is
The flowchart of study selection is presented in Fig. 1. A total of present, it will be detected by visual inspection of the funnel plot and
7844 references were initially retrieved, another two articles were Egger's test for bias (Egger et al., 1997). In the absence of publication
identified through other sources (by reference list and unpublished bias, the studies are expected to fall symmetrically above and below the
data). Following the removal of duplicates (n = 2366), an initial se- mean effect size, suggesting that any sampling error would be random
lection by G.S. and G.P. based on articles' titles reduced the number of (Borenstein, Hedges, Higgins, & Rothstein, 2009). The fail-safe N of
relevant abstracts to 599. Abstracts of these selected articles were Rosenthal indicates the number of studies required to refute significant
screened according to the following criteria: (a) peer-reviewed (e.g., meta-analytic means (Rosenthal, 1979). The unlikelihood of publica-
books and conference abstracts were excluded); (b) included in- tion bias is suggested if Rosenthal's N exceeds the cutoff estimate, which
dividuals with a schizophrenia-spectrum diagnosis (e.g., schizoaffective represents five times the number of studies, plus 10 (Fragkos, Tsagris, &
diagnoses were included); (c) reported on interventions addressing Frangos, 2014; Rosenthal, 1991).
cognitive biases irrespective of study design (randomized controlled Considerable heterogeneity between included studies was expected
trial or naturalistic study); and (d) evaluated effects on cognitive biases, because of methodological differences between them (i.e., the studies
positive symptoms, and/or insight (clinical and/or cognitive). Notably, administered different tests to measure symptoms, insight and cognitive
because cognitive-behavioral therapies, such as the ‘Cognitive-Beha- biases). Therefore, we planned to use a random effects model to esti-
vioral Therapy for psychosis’ (CBTp; Beck, Rector, Stolar, & Grant, mate the mean distribution of intervention effects, as it accommodates
2011), address specific cognitive distortions as part of a case formula- the variation in effect sizes between studies (Lipsey & Wilson, 2001).
tion that is idiosyncratic to the patient, we decided not to include such Heterogeneity of effect sizes' was estimated using Cochran's Q-statistic
studies. These types of interventions also often include other ther- (Cochran, 1954) and the I2 index (Higgins, Thompson, Deeks, &
apeutic targets (e.g., negative symptoms) and strategies (e.g., beha- Altman, 2003). By convention, a Q-statistic p-value below 0.1 indicates
vioral activation) which could risk confounding the therapeutic source heterogeneity (Potvin, 2014), while I2 values of 25, 50 and 75 are as-
of the effects analyzed in the present study. Similarly, studies reporting sociated with low, moderate and strong heterogeneity, respectively
on the effects of the “Social Cognition and Interaction Training’ (SCIT; (Higgins et al., 2003).
Roberts, Penn, & Combs, 2015) were not included even though they Subgroup analyses were performed to determine if any hetero-
address attributional biases because they focus on broader therapeutic geneity of effect sizes was influenced by the overall risk of bias and the
targets (e.g., improving social functioning) and incorporate different use of an active versus passive control condition, as these variables are
techniques (e.g., exposure exercises). Following screening of abstracts, established moderators of meta-analytic findings in schizophrenia
88 full-text articles were assessed for eligibility with careful con- (Eichner & Berna, 2016; Jauhar et al., 2014). Two authors, GS and GP,
sideration of inclusion and exclusion criteria. A total of 32 studies were independently evaluated the 32 studies included in this review for
included in the review; 29 studies (including 2738 participants) were study quality and risk of bias using the criteria described by Eichner and
quantitatively synthesized in the meta-analysis portion. Three studies Berna (2016), which classifies studies as being at high or low risk based
(Andreou et al., 2015; Favrod et al., 2015; Moritz et al., 2014) were not on three factors: randomization to group allocation, masking of out-
included in the meta-analysis portion as they represented reanalyses of come assessments, and incompleteness of outcome data. Studies stating
already published data. Study information (e.g., sample characteristics, that participants were randomly allocated to different groups were
data outcomes) is listed in Table 1. considered to be at a low risk for bias with regard to randomized group
Three meta-analyses assessing cognitive biases, positive symptoms, allocation. Studies that used interviewers for assessing outcomes, who
and insight were separately performed using the Comprehensive Meta- were blind to group allocation of the questioned participants, were
analysis software (version 2.2.021, Biostat, Englewood, NJ). We com- considered as being at a low risk for bias. Studies with dropout rates of
bined measures of clinical and cognitive insight in a single meta- more than 20% that used no intent-to-treat approach were considered
3
Fig. 1. PRISMA flow diagram of study selection, inclusion, and exclusion. N = number of studies.
to be at a high risk for bias. Studies being at low risk of bias regarding 3. Results
randomization, masking and incomplete outcome data were considered
to be at low risk of bias, and studies being at high risk of bias regarding 3.1. Systematic review
at least one of these factors were considered to be at high risk. Ninety-
six risk of bias ratings were assigned as binary outcomes: high and low. The 32 reviewed studies covered 15 different interventions directly
Q-statistics with significance tests were used to test for subgroup dif- targeting cognitive biases. Table 1 presents details of each study as well
ferences between high and low-risk studies. as their main outcome. Twenty reported on cognitive biases, 19 on
Additional subgroup analyses were conducted to determine whether positive symptoms and 11 on insight (clinical insight = 4; cognitive
the use of an active control intervention influenced the effect sizes. insight = 5; both = 2). Eight studies reported 2 out of the 3 outcomes
Interventions including contact with treatment providers, typically measures (i.e., cognitive biases, positive symptoms, and insight) and 6
delivered in treatment-as-usual settings, were defined as an active studies reported on all of them. For each study, a list of the cognitive
control condition (e.g., CogPack, attentional control). Q-statistics with bias, symptoms, and/or insight outcome measures and participant in-
significance tests were used to test for subgroup differences. clusion/exclusion criteria regarding psychotic symptoms is provided in
Supplementary Material Table S1. As well, the quality/risk of bias as-
sessments for each study are presented in Supplementary Material
Table S2.
4
Table 1
Description of studies included in the systematic review.
Study Intervention Design* Control condition Assessments Outcome of experimental intervention Included in meta-
G. Sauvé, et al.
analysis
Aghotor, Pfueller, Moritz, MCT RCT (n = 16: 12 m + 4f, Active control Baseline, Post No improvement related to uniquely to experimental intervention Yes (cognitive biases,
Weisbrod, and Roesch-Ely 28.9yo) (newspaper discussion (MCT). symptoms)
(2010) group)
Andreou et al. (2015) MCT FUP analysis of data Active control Baseline, 6-mth FUP Improvements in data-gathering is associated with delusion No (duplicate
presented in (Moritz et al., (CogPack) decline. publication)
2013)
Andreou et al. (2017) MCT+ RCT (n = 46: 21 m + 25f, Active control (CogPack) Baseline, Post, 6mth-FUP Improvement of delusions and self-reflectiveness score at Yes (cognitive biases,
36.9yo) post-MCT+. symptoms, insight)
Balzan et al. (2014) MCT-T Intervention study TAU Baseline, Post Improvement on global delusion score, positive symptoms, Yes (symptoms,
(n = 14: 11 m + 3f, 38yo) delusional conviction and clinical insight. insight)
Balzan, Mattiske, Delfabbro, Liu, MCT+ RCT (n = 27: 15 m + 12f, Active control (cognitive Baseline, Post, 6mth-FUP Improvement in delusional and overall positive symptom Yes (cognitive biases,
and Galletly (2018) 35.4yo) remediation) severity, and clinical insight. No improvement in JTC bias at post- symptoms, insight)
intervention.
Briki et al. (2014) MCT RCT (n = 25: 16 m + 9f, Active control Baseline, Post Improvement on positive symptoms. Yes (symptoms,
41.1yo) (Supportive therapy) insight)
Buonocore et al. (2015) CACR+MCT RCT (n = 30: 17 m + 13f, Active control Baseline, Post Improvement on the BADE measure. Yes (cognitive biases)
34.4yo) (CACR + newspaper group
discussion)
Erawati, Keliat, Helena, and Hamid MCT Intervention study TAU Baseline, Post Improvement on delusional severity and metacognition. Yes (symptoms)
(2014) (n = 26: 16 m + 10f,
37.1yo)
Favrod et al. (2015) MCT RCT (n = 26: 17 m + 9f, TAU Baseline, Post, Improvement of awareness of delusional ideation at FUP. No (duplicate
36.9yo) 6-mth FUP publication)
Favrod, Maire, Bardy, Pernier, and MCT Pilot study (n = 18: None Baseline, Post Improvement in symptoms (delusions and hallucinations), Yes (symptoms,
5
Bonsack (2011) 11 m + 7f, 41.8yo) insight, and depression. insight)
Favrod et al. (2014) MCT RCT (n = 26: 17 m + 9f, TAU Baseline, Post, 6mth-FUP Improvement on delusions and auditory hallucinations. Yes, symptoms,
36.9yo) insight)
Garety, Waller et al. (2015) MRTP Intervention study Active control Baseline, Pre, Post, 1wk- Improvement of JTC and belief flexibility and reduction in state Yes (cognitive biases)
(n = 51, info on age and (cognitive tasks) FUP paranoia at FUP.
sex n/a)
Gaweda, Krezolek, Olbrys, Turska, MCT Intervention study TAU Baseline, Post Improvement on catastrophization, emotional reasoning, JTC, Yes (cognitive biases,
and Kokoszka (2015) (n = 23: 11 m + 12f, clinical insight, and paranoia. symptoms, insight)
50.4yo)
Hurley et al. (2018) MRTP and CBM-I Intervention Study None Baseline, Pre, Post, 1mth- Improvement on delusional severity and positive symptoms. Yes (symptoms)
(n = 12: 8 m + 4f, 39.4yo) FUP
Ishakawa et al. (2019) MCT RCT (n = 50: 25 m + 25f, TAU Baseline, Mid- Improvement on JTC, hallucinations, positive symptoms, and Yes (cognitive biases,
47.5yo) intervention, Post, 4wk- cognitive insight at post-intervention. Improvement maintained symptoms, insight)
FUP at FUP for JTC, hallucination and positive symptoms.
Improvement on cognitive biases (CBQp) at FUP only.
Kowalski et al. (2017) MCT-JTC Pilot study (n = 12: 1) MCT-ToM 2) Control Baseline, Post Improvement on JTC. No improvement in paranoia. Yes (cognitive biases,
9 m + 3f, 28yo) symptoms)
Lam et al. (2015) MCT RCT (n = 38: 21 m + 17f, TAU Baseline, Post Improvement of self-reflectiveness (cognitive insight). Yes (insight)
41.3yo)
Moritz et al. (2011) MCT Intervention study Wait-list Baseline, Post Improvement on JTC, memory and social relationships. Yes (cognitive biases)
(n = 18, 15 m + 3f,
33.6yo)
Moritz, Mayer-Stassfurth, et al. CBC Intervention study Wait-list Baseline, Post Fewer participants showed JTC at post-intervention. Yes (cognitive biases,
(2015) (n = 33: 14 m + 19f, symptoms)
41.8yo)
Moritz, Thoering, et al. (2015) CRT + MCT Intervention study 1) Active control Baseline, Post, 3mth-FUP Delayed decision-making (baseline-post) and fewer participants Yes (cognitive biases,
(n = 30: 12 m + 18f, (mybraintraining) showed a JTC bias at FUP. symptoms, insight)
40.8yo) 2) Wait-list
(continued on next page)
Table 1 (continued)
Study Intervention Design* Control condition Assessments Outcome of experimental intervention Included in meta-
analysis
G. Sauvé, et al.
Moritz, Veckenstedt, Randjbar, MCT Clinical trial (n = 24: Active control Baseline, Post Improvement on delusion (severity and conviction) and JTC. Yes (symptoms)
Vitzthum, and Woodward (group + ind) 17 m + 7f, 32.6yo) (CogPack)
(2011)
Moritz et al. (2013) MCT RCT (n = 76: 45 m + 31f, Active control Baseline, Post, 6mth-FUP Improvement on delusions which was partially sustained at FUP. Yes (cognitive biases,
36.8yo) (CogPack) Reduction of preoccupation, and distress (amount and intensity). symptoms)
Moritz, Veckenstedt, et al. (2014) MCT FUP analysis of Moritz Active control Baseline, Post, 6mth- Improvement on delusions at both FUPs. No advantage of MCT on No (duplicate
et al. (2013) (CogPack) FUP, 3y-FUP JTC compared to active control. publication)
Naughton et al. (2012) MCT Intervention study Wait-list Baseline, Post Improvement in mental capacity (understanding and reasoning) Yes (symptoms)
(n = 11: 11 m + 0f, and functioning.
37.5yo)
Ross et al. (2011) RT Intervention study Active control Baseline, Post Improvement in JTC and belief flexibility. Yes (cognitive biases)
(n = 34: 25 m + 9f, 39yo) (cognitive tasks)
So et al. (2015) MCTd RCT (n = 23: 12 m + 11f, TAU Baseline (Ctrl), Pre-int Improvement of positive symptoms, which is mediated by Yes (cognitive biases,
32.4yo) (exp.), Post, 4wk-FUP improvement in belief flexibility. symptoms)
Steel et al. (2010) CBM Intervention study Active control Baseline, Post No improvement related to uniquely to experimental intervention Yes (cognitive biases)
(n = 21: 15 m + 6f, 43yo) (cognitive tasks) (CBM).
Turner et al. (2011) CBM-I Pilot study None Baseline, Post Improvement on interpretive bias (change from a negative to Yes (cognitive biases)
(n = 8: 7 m + 1f, 24.8yo) positive interpretive bias).
Turner et al. (2018) MCT-JTC (single RCT (n = 19: 14 m + 5f, Active control (attention Baseline, Post Improvement in JTC (draws to decision). Yes (cognitive biases)
session) 45.3yo) control)
van Oosterhout et al. (2014) MCT RCT (n = 75: 54 m + 21f, TAU Baseline, Post, 6mth-FUP No improvement related to uniquely to experimental intervention Yes (cognitive biases,
38.3yo) (MCT). symptoms, insight)
Waller et al. (2015) TW RC feasibility study TAU Baseline, Post-MRTP, Improvement of belief flexibility at post-MRTP and post-TW. Yes (cognitive biases)
(n = 20: 15 m + 5f, Post-TW, 2wk-FUP
39.1yo)
6
Waller et al. (2011) MRTP Pilot study None. Baseline, Pre, Post, 2wk- Improvement on JTC, belief flexibility, and delusional conviction. Yes (cognitive biases)
(n = 13: 7 m + 6f, 44.6yo) FUP
Note. * N refers to number of patients in experimental group. Abbreviations: BADE: Bias against disconfirmatory evidence; CACR+MCT: Computer-assisted cognitive remediation + metacognitive training; CBC:
Cognitive bias correction; CBMeI: Cognitive bias modification – for interpretive biases; CBM: Cognitive bias modification; CRT + MCT: Cognitive remediation therapy + metacognitive training; Ctrl: control; f: Female;
FUP: Follow-up; JTC: Jumping to conclusions; m: Male; MCT (group+ind): Group and individual metacognitive training; MCT-T: Metacognitive training targeted; MCT: Metacognitive training; MCTd: Metacognitive
training for delusions; MRTP: Maudsley review training programme; mth: Month; n/a: non-available; RC: Randomized controlled; RCT: Randomized controlled trial; RT: Reasoning training; TAU: Treatment as usual; TW:
Thinking Well; wk.: week; y: Year; yo: years old. A more detailed description of the studies outcomes can be found in Supplementary Material.
3.1.1. Metacognitive training and adaptations over the course of 16 weeks as modules are completed in two sessions.
One of the most frequently used interventions was metacognitive
training (MCT), which was developed by Moritz and Woodward 3.1.3. Cognitive bias correction
(2007b). MCT combines techniques from psychoeducation, cognitive Another reviewed intervention is called ‘cognitive bias correction’
remediation and cognitive-behavioral therapies, and aims to help par- (CBC) and was developed by Moritz et al. (2015). CBC is an online
ticipants develop insight and awareness into the different cognitive psychoeducational program offering 6 modules that aim to teach par-
biases known to be related to delusions (Kumar, Menon, Moritz, & ticipants about 20 general cognitive biases not necessarily implicated in
Woodward, 2015). This intervention also includes a knowledge trans- psychosis (e.g., Cocktail party effect of selective attention, optical il-
lation component, which further helps participants realize the negative lusions, hindsight bias). Participants first complete tasks that are de-
consequences of their cognitive biases to daily life. MCT was initially signed to elicit the cognitive biases so that they can be experienced
developed for a group format (see below for individual format adap- firsthand. Afterwards, participants receive psychoeducation on these
tations) and comprises 8 modules targeting the following cognitive common thinking mistakes and how these cognitive biases emerge.
biases: JTC, BADE, attributional biases, and overconfidence in memory
errors. Two cycles with different examples are available. The training 3.1.4. Cognitive bias modification
aims to enhance participants' metacognitive abilities (i.e., being more The ‘cognitive bias modification’ (CBM) method specifically targets
aware of their cognitive biases) by (a) engaging them in numerous negative interpretive biases. It trains participants to generate positive
cognitive tasks, (b) providing feedback and corrective exercises, and (c) resolutions of ambiguous situations that can be interpreted in a nega-
explaining links between what has been learnt and daily life. In addi- tive way (Steel et al., 2010). Participants are presented with 100 audio-
tion to the aforementioned biases, aspects related to theory of mind, recordings of scenarios depicting ambiguous situations. Each scenario
mood and self-esteem are also covered (Moritz et al., 2014). The de- describes an initially ambiguous situation that is subsequently resolved
velopers of the intervention provide all the materials (including pre- in a positive way. The CBM intervention was originally developed for
sentation slides and therapist manual) necessary to conduct the inter- individuals with anxiety and depression disorders and used visual ma-
vention free of charge (http://www.uke.de/mct), which has fostered terial instead of audio recordings (Grey & Mathews, 2000; Salemink,
several adaptations and multiple language translations. van den Hout, & Kindt, 2007). A variant of the CBM intervention targets
As mentioned earlier, the MCT developers have adapted their threat-related interpretive bias (CBMeI) and uses visual training ma-
training to an individualized setting (referred to as MCT+). This flex- terial (Hurley, Hodgekins, Coker, & Fowler, 2018; Turner et al., 2011).
ible manualized individual version uses the same exercises as those Three-sentence scenarios describing emotionally ambiguous social si-
presented in the group version but addresses them in relation to pa- tuations are presented to participants on a computer screen. The final
tients' specific symptoms and challenges (Moritz, Vitzthum, et al., word of the first sentence is presented in fragments (e.g., ‘ap—gis-’ for
2010). The material is divided into 11 units in its most updated version apologise). These fragments can lead to negative or positive words, but
(2.3) and each is covered over several sessions (Moritz et al., 2016). as the remaining sentences are revealed, the scenario is always dis-
Several included studies also presented targeted adaptations of MCT+. ambiguated in a positive manner. The training involves asking parti-
For instance, some studies delivered combinations of units in a few cipants to complete the fragmented words before they are revealed. A
sessions to specifically target JTC, delusions or belief flexibility (Balzan, comprehension question follows each trial to ensure proper under-
Delfabbro, Galletly, & Woodward, 2014; Kowalski, Pankowski, Lew- standing of the described situation.
Starowicz, & Lukasz, 2017; So et al., 2015).
3.1.5. Maudsley Review Training Programme and adaptations
3.1.2. Combination of metacognitive training and cognitive remediation The ‘Maudsley Review Training Programme’ (MRTP) consists of a
therapy computerised program that introduces participants to the concept of
Other interventions included in our review combined aspects of the JTC. Participants are also invited to complete 5 training tasks accom-
MCT with cognitive remediation therapy (CRT) techniques. CRT is an panied by a therapist who provides positive feedback, reinforces insight
evidence-based intervention aimed to enhance cognitive skills in order and normalizes JTC (Hurley et al., 2018; Waller, Freeman, Jolley,
to compensate for the various neurocognitive deficits (e.g., memory, Dunn, & Garety, 2011). The first task, named ‘What's the picture’ is
attention) frequently observed in SZ&RP (Medalia & Choi, 2009). One adapted from Moritz and Woodward (2007a) and teaches participants
included intervention consists of combining MCT elements with the to look for additional evidence before making a decision. Six pictures
online cognitive remediation program, called ‘mybraintraining’, de- are revealed one piece at a time. After each revealed piece, participants
veloped by Dr. Ryuta Kawashima (Moritz et al., 2015). The original are asked if they would prefer to see another piece or immediately
online CRT program targets abilities in calculation, logic, memory and decide on what the picture was, based on a choice of 6 options. At first,
vision. In the intervention integrating MCT elements (CRT + MCT), all options seem plausible, but as the picture is incrementally revealed,
participants are asked to rate their confidence in their answers to each certain options can be ruled out. The second task teaches participants to
exercise comprised in the training. When hasty incorrect decisions are slow their decision-making process by trying to see other interpreta-
made with high confidence, participants automatically receive feedback tions of optical illusions. The third task, also addressing the JTC bias,
and are encouraged to take more time before making a decision for the shows participants series of 3 video clips. The clips are designed to
next trials. The CRT + MCT intervention is conducted online without a make participants jump to conclusions at first, while the subsequent
therapist and participants can complete the training at the location of clips show alternative interpretations. The fourth task addresses
their choice. thinking flexibility by showing participants three video clips that il-
Another reviewed intervention combines group sessions of com- lustrate scenarios with a potential paranoid interpretation. After each
puter-assisted CRT (CA-CRT) and MCT. The intervention (CA- clip, participants are invited to think about alternative interpretations
CRT + MCT) differs from the CRT + MCT one in that it consists of (neutral, positive and negative). Finally, in the fifth task, participants
three structured 1-h sessions per week of CA-CRT using the CogPack are shown 4 video clips depicting scenarios in which one character
program® (Marker, 2003) followed by a fourth session of MCT during jumps to conclusions. Participants are then asked which character
the week (Buonocore et al., 2015). Also, the sessions are conducted in jumped to conclusions and how this character could have avoided such
small groups of 4–5 participants and led by trained psychologists. The a bias. Handouts with key aspects of the training are provided to par-
CogPack program includes four sets of exercises that are tailored to the ticipants. The five tasks are completed in one session lasting about
participants' needs according to their performances on a baseline neu- 1.5 h. An MRTP adaptation, the Thinking Well (TW) intervention
ropsychological assessment. The 8 modules of MCT are administered (Waller et al., 2015), combines MRTP with four sessions of individual
7
therapy to further apply their learning to their own thinking errors. a publication bias is unlikely. However, the Rosenthal's fail-safe N was
Through these sessions, participants apply the techniques learned 80, which is lower than the cut-off of 110, indicating a potential pub-
during the MRTP to their own delusional beliefs and work toward a lication bias.
chosen goal with the therapist. Subgroup analyses were also conducted to test whether overall risk
bias (high vs. low) and inclusion of active control condition modified
3.1.6. Reasoning training the effect of interventions on cognitive biases in comparison to control
The reasoning training (RT) targeted JTC and BADE. RT was de- conditions. For studies at high risk of bias, the mean effect size was
livered in a single 45-min session and comprised three training tasks of higher (Hedge's g = 0.35; 95% CI = [0.18–0.53], z = 3.990; p < .001)
about 15 min each. RT introduces participants to cognitive biases and than the main results where all levels of risk of bias were combined.
provides strategies to avoid them (Ross et al., 2011). The first two tasks However, statistical significance was not retained in studies at low of
are adapted from the MCT material (object identification and picture risk bias (Hedges' g = 0.14; 95% CI = [−0.07–0.34]; z = 1.32;
interpretation; Moritz & Woodward, 2007a) and the third task is the p = .19). Although the effect size of studies with a high risk of bias was
optical illusion task of MRTP (Waller et al., 2011). Each task is divided larger than those with low risk of bias, there was no statistically sig-
in 3 phases. In the object identification task, participants are first pre- nificant difference between the effect sizes of studies at high and low
sented 5 pictures of incomplete objects that are incrementally revealed risk of bias, Q(1)btwn = 2.43, p = .12. This indicates that the main result
over a series of 8 slides. After each slide, participants are asked if they may be driven by studies at high risk of bias.
want to see another slide before identifying the object from a list of 6 When examining differences in control conditions, both the pre-
options. Participants are free to select their answer after any number of sence (Hedges' g = 0.27; 95%CI = [0.08–0.47]; z = 2.73; p = .006) and
revealed pieces although some options became less plausible as pieces absence of an active control condition (Hedges' g = 0.20;
are revealed. This represents the first phase (baseline). In the second 95%CI = [−0.04–0.45]; z = 1.66; p = .10) identified a small effect on
phase (training), the same pictures are reviewed with the therapist and the improvement of cognitive biases. Further, the difference between
all pieces are shown to the participant to illustrate how hasty decision- these effects was not statistically significant (Q(1)btwn = 0.91; p = .64).
making can lead to erroneous answers. In the third and final phase Overall, both risk of bias and type of control condition therefore do not
(bolster), a different set of 5 pictures are presented and participants are considerably affect the impact of interventions on the reduction of
encouraged to request as many slides as they wish before making their cognitive biases.
decision. Similarly, in the picture interpretation task, participants are
asked to identify among 4 options the correct title of 9 paintings. 3.2.2. Positive symptoms
During the baseline phase, answers are collected for 4 paintings without Our second meta-analysis investigating the effects of interventions
indicating to the participants whether they are correct or not. The on positive symptoms included 19 studies, totalling 1005 participants.
paintings with their correct answers are then reviewed with the The vast majority of studies investigated the effects of diverse forms of
therapist during the training phase and participants are encouraged to MCT (N = 16). One study examined the outcomes of the CBC program
weight the evidence supporting and refuting each possible option be- (Moritz, Mayer-Stassfurth, et al., 2015). The other two studies verified
fore making a decision. In the bolster phase, participants are shown an the impacts of the following combinations: (1) MCT + CRT (Moritz,
additional 5 paintings and encouraged to weigh the evidence before Thoering, et al., 2015), and (2) MRTP + CBM (Hurley et al., 2018).
making a choice. Finally, the optical illusion task consists of 11 images Results indicate that interventions have a moderate significant positive
that can be interpreted in 2 ways, for example the depicted woman can effect on the improvement of psychotic symptoms (Hedge's g = 0.30;
be either perceived as old or young. The baseline phase comprises 5 95% CI = [0.13–0.48]; z = 3.44, p < .005). The forest plot is pre-
pictures and participants freely describe what they see. In the training sented in Fig. 3.
phase, each picture is reviewed and the different perspectives are re- Additional analyses conducted to verify the robustness of this
vealed. An additional 6 pictures are presented after during the bolster finding suggest that characteristics of included studies are hetero-
phase. geneous (Q18 = 37.1; df = 19; p = .008; I2 = 51.5). Such heterogeneity
can stem from the differences between the interventions, outcomes
3.2. Meta-analysis results measures, samples' characteristics, etc. While both the funnel plot
(Supplementary Material Fig. S2) and results of Egger's asymmetry test,
3.2.1. Cognitive biases t(17) = 1.01, p = .33 indicate that the presence of a publication bias is
Twenty studies, comprised of 1085 participants with a schizo- unlikely, such a bias could not be entirely ruled out because Rosenthal's
phrenia spectrum diagnosis, were included in our first meta-analysis fail-safe N = 99 was slightly below the cut-off of 105.
investigating the effects of interventions on cognitive biases. About half Subgroup analyses were conducted to test whether overall risk bias
of the studies (N = 11) investigated the effects MCT or one of its (high vs. low) and inclusion of active control condition modified the
adaptations. Two studies verified the impact of combining MCT with effect of interventions on positive symptoms in comparison to control
cognitive remediation (Buonocore et al., 2015; Moritz, Thoering, et al., conditions. The mean effect size of studies at high risk of bias was
2015). Three studies used the MRTP alone (Garety et al., 2015) or in higher (Hedge's g = 0.40; 95%CI = [0.17–0.63]; z = 3.45; p = .001)
combination with the CBM-I (Hurley et al., 2018) or its adaptation, the compared to the main result including all levels of risk. In contrast, the
TW program (Waller et al., 2015). The remaining four studies verified mean effect size of studies at low risk of bias was lower (Hedges'
the effects of RT (Ross et al., 2011), CBC (Moritz, Mayer-Stassfurth, g = 0.19; 95%CI = [−0.06–0.44]; z = 1.52; p = .13) than the main
et al., 2015), and CBM (Steel et al., 2010; Turner et al., 2011). Results of result. However, there was no statistically significant difference be-
the meta-analysis suggest that interventions have a small, positive and tween the improvement of positive symptoms among studies presenting
statistically significant effect on the reduction of cognitive biases a high versus low risk of bias (Q(1)btwn = 1.45; p = .23).
(Hedge's g = 0.27; 95% CI = [0.13–0.41]; z = 3.77; p < .001). The When examining differences in control conditions, both the pre-
forest plot is presented in Fig. 2. sence (Hedges' g = 0.23; 95%CI = [−0.01–0.47]; z = 1.92; p = .06)
Additional analyses were conducted to verify the robustness of this and absence of an active control condition (Hedges' g = 0.22;
finding. Results suggest it is unlikely that the included studies' char- 95%CI = [−0.01–0.45]; z = 1.86; p = .06) identified a small effect on
acteristics are heterogeneous (Q19 = 24.649; df = 20; p = .21; the improvement of positive symptoms. The difference between these
I2 = 23.66). The funnel plot (Supplementary Material Fig. S1) and the effects was not statistically significant (Q(1)btwn = 0.01; p = .92). Thus,
results of Egger's asymmetry test, t(18) = 1.48; p = .16, which suggest both risk of bias and type of control condition do not considerably affect
no evidence of funnel plot asymmetry, also indicate that the presence of the impact of interventions on the improvement of positive symptoms.
8
Fig. 2. Forest plot of studies in the meta-analysis of cognitive biases. Effect sizes of interventions on cognitive biases. Positive effect sizes favour the effect of
treatment on cognitive biases over the effect of control.
3.2.3. Insight symptoms, characteristics of included studies assessing insight were

For this third meta-analysis verifying the effects of interventions on found to be heterogeneous (Q10 = 18.57; df = 11; p = .069; I2 = 46.1).
insight levels, 11 studies were included. This represents 648 partici- While visual inspection of the funnel plot (Supplementary Material Fig.
pants. All studies investigated the effects of MCT or its variants, and one S3) and results of Egger's asymmetry test (t(9) = 0.16, p = .88) did not
its combination with CRT (Moritz, Thoering, et al., 2015). Results of hint toward a publication bias, Rosenthal's fail-safe N of 50 was lower
this meta-analysis indicate that interventions have a moderate sig- than the cut-off of 65, suggesting the likelihood of publication bias.
nificant positive effect on the improvement of patients' insight levels Further, subgroup analyses indicate that effects sizes do not sig-
(Hedge's g = 0.35; 95% CI = [0.15–0.56]; z = 3.37, p < .005). The nificantly differ between studies of high versus low risk of bias (Q(1)
forest plot is presented in Fig. 4. As with the meta-analysis on btwn = 0.53; p = .47), nor between studies with or without an active
Fig. 3. Forest plot of studies in the meta-analysis of positive symptoms. Effect sizes of interventions on positive symptoms. Positive effect sizes favour the effect of
treatment on positive symptoms over the effect of control condition.
9
Fig. 4. Forest plot of studies in the meta-analysis of insight. Effect sizes of interventions on insight. Positive effect sizes favour the effect of treatment on insight over
the effect of control condition.
control group (Q(1)btwn = 2.58; p = .11). Therefore, both risk of bias change in cognitive biases. Several issues have been raised for the
and type of control condition do not considerably affect the impact of beads/fish task (e.g., difficulty understanding the task, lack of parallel
interventions on the improvement of insight. test-retest versions; Moritz et al., 2017), which was the most frequently
used in our included studies. On the other hand, positive symptoms and
4. Discussion insight were most often evaluated with robust instruments that con-
tained multiple items and were clinician-rated (e.g., SAPS, PSYRATS,
The present study reviewed the literature on psychological inter- SUMD). The fact that multiple cognitive biases were regularly measured
ventions systematically targeting cognitive biases in SZ&RP and eval- using a single tool (e.g., beads/fish task), as opposed to the variety of
uated their efficacy at improving cognitive biases, positive symptoms, scales used to evaluate positive symptoms and insight, could also partly
and insight (clinical and cognitive) via meta-analysis. We identified 32 explain why larger effect sizes were found. Further, interventions could
relevant studies, which included 15 different psychological interven- have more generalized effects on positive symptoms and insight, due to
tions directly targeting cognitive biases in patients with SZ&RP, and their integrative and normalizing nature, also explaining in part the
wherein the following cognitive biases were measured: JTC, BADE, larger effects sizes compared to cognitive biases. These findings have
belief inflexbility, intentionalising, catastrophizing, dichotomous important theoretical and clinical implications, which will be discussed
thinking, emotional reasoning, representativeness bias, illusion of below.
control bias, and interpretive bias. Surprisingly, no study reported re-
sults on the overconfidence in errors bias, which calls for more com-
prehensive investigations of cognitive biases in intervention studies. As 4.1. Theoretical implications
expected, the most common intervention used to target cognitive biases
was MCT. Several studies used MCT variants, developed for individual Several theoretical models of SZ&RP include cognitive biases as an
treatment (MCT+, MCT-T, MCT-JTC), delusion-specific biases (MCTd), important mechanism of the formation and maintenance of positive
or combined with other cognitive interventions (CA-/CRT + MCT). symptoms (Bell, Halligan, & Ellis, 2006; Broyd et al., 2017; Moritz
Several of the additional reviewed interventions borrow modules or et al., 2016; Sarin & Wallin, 2014). In a seminal paper, Kapur (2003)
modify tasks from MCT, but all shared the aims of improving cognitive proposed the “aberrant salience” account of positive symptoms in
biases and/or symptoms by teaching patients about cognitive biases psychosis, which posits that cognitive biases modify perceptual pro-
that have been associated with symptoms in psychosis. Due to this cessing of certain irrelevant stimuli to render them hypersalient; hal-
overlap, and our interest in investigating psychological interventions lucinations are a direct manifestation of this hypersalience, while de-
targeting cognitive biases overall, we included all relevant studies in lusions arise from a natural desire to explain these experiences. Garety,
our review and meta-analysis. Kuipers, Fowler, Freeman, and Bebbington’ (2001) cognitive model of
A total of 29 studies were included in our quantitative meta-ana- positive symptoms places greater emphasis on affective disturbances
lyses. We found that psychological interventions targeting cognitive and emotional distress interacting with cognitive biases to produce
biases have small to moderate significant effects on the improvement of hallucinations and delusions. In a similar vein, Salvatore et al. (2012)
cognitive biases, psychotic symptoms and insight. Overall, these results indicated that cognitive biases could contribute to paranoid delusions
appear to be relatively robust. While studies' characteristics do not because they arise when patients feel threatened. Bentall and Kaney
appear to be heterogeneous for cognitive biases, heterogeneity was (2005) proposed that cognitive biases arise from attempts to reduce
found for positive symptoms and insight. The presence of a publication discrepancies between actual and ideal self-representations, which in
bias seems unlikely for insight, but is possible for cognitive biases and turn may lead to persecutory delusions. More recent cognitive models
positive symptoms. Nonetheless, the risk of bias and the inclusion of an of positive symptoms in psychosis (Broyd et al., 2017; Moritz, Pfuhl,
active control group does not seem to artificially increase effect sizes. et al., 2016) build on previous accounts and distinguish between biases
Interestingly, the global effect size for cognitive biases was smaller affecting the formation and maintenance of delusional beliefs. In our
than for either positive symptoms or insight although all included in- view, these aforementioned theoretical models are further supported by
terventions were developed to specifically target cognitive biases. This our findings. Indeed, the currently reviewed interventions specifically
could be explained by several factors. First, it could represent a non- targeting cognitive biases appear to efficaciously improve positive
significant numerical difference given that the confidence intervals are symptoms without addressing them directly. This further raises im-
fairly large and overlap (Bakker et al., 2019). It could also partly stem portant clinical implications for the development and treatment of SZ&
from the psychometric properties of the instruments used to measure RP.
10
4.2. Clinical implications 4.4. Conclusions and future directions
Cognitive biases have not only been observed in multi-episode or The current study reviewed the literature on psychological inter-
enduring SZ&RP patients. Individuals at clinical high-risk (CHR) of ventions targeting cognitive biases in patients with enduring SZ&RP.
developing SZ&RP and those experiencing a first episode of psychosis Our review highlights several available interventions addressing a
(FEP) also seem to present with cognitive biases (Eisenacher & Zink, range of cognitive biases affected in SZ&RP that show good feasibility
2017; Ross et al., 2015). This suggests that cognitive biases could begin and acceptance in this population. The meta-analytic results support the
to increase in the early stages of the illness. Given evidence that cog- use of these interventions in enduring SZ&RP and indicate that they
nitive biases could be markers of psychosis (Eisenacher & Zink, 2017), have small to moderate effects on cognitive biases, symptoms, and in-
they may represent an interesting therapeutic target. Further, psycho- sight (clinical and cognitive insight combined). However, future re-
logical interventions targeting cognitive biases may also have pre- search should systematically include change in cognitive biases as a
ventative or beneficial effects in these at-risk and early illness groups. primary outcome to better understand how improvement in cognitive
Therefore, it would be worth investigating whether the interventions biases lead or be associated with better insight and reduced positive
currently reviewed could be beneficial for these populations and per- symptoms. Future studies should also use an active control condition,
haps even prevent conversion to psychosis. Promising results have been and reduce the risk of bias by using randomization, blinding/masking,
published so far. Studies offering MCT to FEP participants have shown and avoiding incomplete outcome data. A promising avenue will be to
improvements in positive symptoms and cognitive insight (Orcel et al., assess the efficacy of interventions targeting cognitive biases in CHR
2013; Ussorio et al., 2016). Future research would benefit from asses- and FEP groups to determine whether they may help mitigate pro-
sing the effects of psychological interventions targeting cognitive biases dromal or early symptoms, improve insight, or even help prevent
to determine whether they may be utilized as preventative or mitigating conversion to psychosis.
treatments for these groups.
Relatedly, our results suggest that cognitive biases are malleable in Declaration of Competing Interest
SZ&RP via psychological interventions. Such finding adds important
information to the current debate of whether cognitive-behavioral Role of funding sources: This research was not funded by a spe-
therapy for psychosis (CBTp) represents an efficient treatment for po- cific granting agency, commercial or/not-for-profit sector. Salary
sitive symptoms (McKenna, Leucht, Jauhar, Laws, & Bighelli, 2019). awards include: doctoral award from the Fonds de Recherche du Québec –
Change in cognitive biases following intervention could arguably re- Santé (FRQeS) for author GS; postdoctoral fellowship from the
present one of the mechanisms at work in CBTp. Future trials examining Canadian Institutes of Health Research (CIHR) for author KML; FRQ-S
the efficacy of CBTp could likely benefit from including outcome Research Scholar salary award for author MBB; and James McGill
measures of cognitive biases in addition to the typical evaluation of Professorship from McGill University and Research Chair from the FRQ-
positive and negative symptoms. Although it was not included in the S for author ML. The funding sources had no role in the study design,
current analyses, one study by Lincoln et al. (2014) have reported that collection, analysis, or interpretation of the data, writing the manu-
cognitive biases were significantly related to positive symptoms at 1- script, or the decision to submit the paper for publication.
year follow-up of CBTp in 80 SZ&RP patients. The authors concluded Contributors: GS, MB and ML designed the study and wrote the
that their finding supports the notion that the success of CBTp can protocol. GS, KML and GP conducted literature searches and data ex-
partly be explained by correcting cognitive biases. The fact that psy- traction. GS and GP conducted the statistical analysis. GS and KML
chological interventions systematically targeting cognitive biases also wrote the first draft of the manuscript. All authors contributed to and
improve insight could represent a further evidence of this notion. have approved the final manuscript.
Conflict of Interest Statement: Authors GS, KML, GP and MBB
declare no conflicts of interest. Author ML reports grants from Otsuka
4.3. Limitations Lundbeck Alliance, personal fees from Otsuka Canada, personal fees
from Lundbeck Canada, grants and personal fees from Janssen, and
We observed heterogeneity of effect sizes for studies assessing personal fees from MedAvante-Prophase, all outside the submitted
symptoms and insight, but not for cognitive biases as well as some work.
evidence of publication bias for and positive symptoms cognitive biases
due to the low fail-safe N (though analysis of the funnel plots suggests Appendix A. Supplementary data
no publication bias). This heterogeneity may be due to the differences
in intervention types, outcome measures, and sample characteristics Supplementary data to this article can be found online at https://
included in the meta-analysis. This limitation was, however, cir- doi.org/10.1016/j.cpr.2020.101854.
cumvented by using a random-effects model, which assumes that the
real effect size varies from one study to another (Borenstein et al., References
2009). Our risk of bias analysis indicated that low versus high quality
studies in terms of randomization, masking, and incomplete data did Aghotor, J., Pfueller, U., Moritz, S., Weisbrod, M., & Roesch-Ely, D. (2010). Metacognitive
not result in significantly different effect sizes; however, the subgroup training for patients with schizophrenia (MCT): Feasibility and preliminary evidence
analysis suggested a trend toward stronger effect sizes for interventions for its efficacy. Journal of Behavior Therapy and Experimental Psychiatry, 41(3),
207–211. https://doi.org/10.1016/j.jbtep.2010.01.004.
on symptoms when using an active control. Further, it would have been Amador, X. F., & Kronengold, H. (2004). Understanding and assessing insight. In X. F.
interesting to conduct additional subgroup analyses (e.g., different Amador, & A. David (Eds.). Insight and psychosis: awareness of illness in schizophrenia
targeted cognitive biases, different types of interventions). This was and related disorders (pp. 402). Oxford University Press.
Andreou, C., Schneider, B. C., Balzan, R., Luedecke, D., Roesch-Ely, D., & Moritz, S.
unfortunately not possible because the number of studies in each sub- (2015). Neurocognitive deficits are relevant for the jumping-to-conclusions bias, but
group would have been too small. These findings highlight the im- not for delusions: A longitudinal study. Schizophrenia Research: Cognition, 2(1), 8–11.
portance of conducting additional high quality, randomized-controlled https://doi.org/10.1016/j.scog.2015.02.001.
Andreou, C., Wittekind, C. E., Fieker, M., Heitz, U., Veckenstedt, R., Bohn, F., & Moritz, S.
trials in larger samples and taking into consideration the type of control (2017). Individualized metacognitive therapy for delusions: A randomized controlled
condition (active versus passive) used. Finally, assessment of the tol- rater-blind study. Journal of Behavior Therapy and Experimental Psychiatry, 56,
erance/feasibility of these interventions was beyond the scope of this 144–151. https://doi.org/10.1016/j.jbtep.2016.11.013.
Bakker, A., Cai, J., English, L., Kaiser, G., Mesa, V., & Van Dooren, W. (2019). Beyond
report, but has been addressed in a previous meta-analysis on MCT
small, medium, or large: Points of consideration when interpreting effect sizes.
(Eichner & Berna, 2016).
11
Educational Studies in Mathematics, 102, 1–8. https://doi.org/10.1007/s10649-019- inexplicable to the treatable. The British Journal of Psychiatry, 203(5), 327–333.
09908-4. https://doi.org/10.1192/bjp.bp.113.126953.
Balzan, R. (2016). Overconfidence in psychosis: The foundation of delusional conviction? Garety, P., Waller, H., Emsley, R., Jolley, S., Kuipers, E., Bebbington, P., ... Freeman, D.
Cogent Psychology, 3(1), 1–11. https://doi.org/10.1080/23311908.2015.1135855. (2015). Cognitive mechanisms of change in delusions: An experimental investigation
Balzan, R., Delfabbro, P., Galletly, C., & Woodward, T. (2013). Confirmation biases across targeting reasoning to effect change in Paranoia. Schizophrenia Bulletin, 41(2),
the psychosis continuum: The contribution of hypersalient evidence-hypothesis 400–410. https://doi.org/10.1093/schbul/sbu103.
matches. British Journal of Clinical Psychology, 52(1), 53–69. https://doi.org/10. Garety, P. A., & Freeman, D. (2013b). The past and future of delusions research: From the
1111/bjc.12000. inexplicable to the treatable. The British Journal of Psychiatry, 203(5), 327–333.
Balzan, R., Delfabbro, P. H., Galletly, C. A., & Woodward, T. S. (2014). Metacognitive https://doi.org/10.1192/bjp.bp.113.126953.
training for patients with schizophrenia: Preliminary evidence for a targeted, single- Garety, P. A., Kuipers, E., Fowler, D., Freeman, D., & Bebbington, P. E. (2001). A cognitive
module programme. Australian and New Zealand Journal of Psychiatry, 48(12), model of the positive symptoms of psychosis. Psychological Medicine, 31(2), 189–195.
1126–1136. https://doi.org/10.1080/17522439.2017.1300186. https://doi.org/10.1017/s0033291701003312.
Balzan, R. P., Mattiske, J. K., Delfabbro, P., Liu, D., & Galletly, C. (2018). Individualized Gaweda, L., Krezolek, M., Olbrys, J., Turska, A., & Kokoszka, A. (2015). Decreasing self-
metacognitive training (MCT+) reduces delusional symptoms in psychosis: A ran- reported cognitive biases and increasing clinical insight through meta-cognitive
domized clinical trial. Schizophrenia Bulletin, 45(1), 27–36. https://doi.org/10.1093/ training in patients with chronic schizophrenia. Journal of Behavior Therapy and
schbul/sby152. Experimental Psychiatry, 48, 98–104. https://doi.org/10.1016/j.jbtep.2015.02.002.
Beck, A. T., Baruch, E., Balter, J. M., Steer, R. A., & Warman, D. M. (2004). A new in- Grey, S., & Mathews, A. (2000). Effects of training on interpretation of emotional am-
strument for measuring insight: The Beck cognitive insight scale. Schizophrenia biguity. Quarterly Journal of Experimental Psychology, 53(4), 1143–1162. https://doi.
Research, 68(2–3), 319–329. https://doi.org/10.1016/S0920-9964(03)00189-0. org/10.1080/713755937.
Beck, A. T., Rector, N. A., Stolar, N., & Grant, P. (2011). Schizophrenia: Cognitive theory, Grisham, J. R., Becker, L., Williams, A. D., Whitton, A. E., & Makkar, S. R. (2014). Using
research, and therapy. New York, USA: Guilford Press. cognitive bias modification to deflate responsibility in compulsive checkers. Cognitive
Bell, V., Halligan, P. W., & Ellis, H. D. (2006). Explaining delusions: A cognitive per- Therapy and Research, 38, 505–517. https://doi.org/10.1007/s10608-014-9621-0.
spective. Trends in Cognitive Sciences, 10(5), 219–226. https://doi.org/10.1016/j.tics. Hedges, L. V., & Olkin, I. (1985). Statistical methods for meta-analysis. Orlando, Florida:
2006.03.004. Academic Press.
Bentall, R. P., Corcoran, R., Howard, R., Blackwood, N., & Kinderman, P. (2001). Higgins, J. P., Thompson, S. G., Deeks, J. J., & Altman, D. G. (2003). Measuring incon-
Persecutory delusions: A review and theoretical integration. Clinical Psychology sistency in meta-analyses. BMJ, 327(7414), 557–560. https://doi.org/10.1136/bmj.
Review, 21(8), 1143–1192. https://doi.org/10.1016/s0272-7358(01)00106-4. 327.7414.557.
Bentall, R. P., & Kaney, S. (2005). Attributional lability in depression and paranoia. The Huq, S. F., Garety, P. A., & Hemsley, D. R. (1988). Probabilistic judgements in deluded
British Journal of Clinical Psychology, 44(Pt 4), 475–488. https://doi.org/10.1348/ and non-deluded subjects. Quarterly Journal of Experimental Psychology, 40(4),
014466505x29602. 801–812. https://doi.org/10.1080/14640748808402300.
Borenstein, M., Hedges, L. V., Higgins, J. P. T., & Rothstein, H. R. (2009). Introduction to Hurley, J., Hodgekins, J., Coker, S., & Fowler, D. (2018). Persecutory delusions: Effects of
meta-analysis. Wiley & Sons Ltd. cognitive bias modification for interpretation and the Maudsley Review Training
Briki, M., Monnin, J., Haffen, E., Sechter, D., Favrod, J., Netillard, C., ... Vandel, P. Programme on social anxiety, jumping to conclusions, belief inflexibility and para-
(2014). Metacognitive training for schizophrenia: A multicentre randomised con- noia. Journal of Behavior Therapy and Experimental Psychiatry, 61, 14–23. https://doi.
trolled trial. Schizophrenia Research, 157(1–3), 99–106. https://doi.org/10.1016/j. org/10.1016/j.jbtep.2018.05.003.
schres.2014.06.005. Ishakawa, R., Ishagaki, T., Shimada, T., Tanoue, H., Yoshinaga, N., Oribe, N., ... Hosono,
Broyd, A., Balzan, R. P., Woodward, T. S., & Allen, P. (2017). Dopamine, cognitive biases M. (2019). The efficacy of extended metacognitive training for psychosis: A rando-
and assessment of certainty: A neurocognitive model of delusions. Clinical Psychology mized controlled trial. Schizophrenia Research. https://doi.org/10.1016/j.schres.
Review, 54, 96–106. https://doi.org/10.1016/j.cpr.2017.04.006. 2019.08.006.
Buonocore, M., Bosia, M., Riccaboni, R., Bechi, M., Spangaro, M., Piantanida, M., ... Jauhar, S., McKenna, P. J., Radua, J., Fung, E., Salvador, R., & Laws, K. R. (2014).
Cavallaro, R. (2015). Combined neurocognitive and metacognitive rehabilitation in Cognitive–behavioural therapy for the symptoms of schizophrenia: Systematic review
schizophrenia: Effects on bias against disconfirmatory evidence. European Psychiatry, and meta-analysis with examination of potential bias. British Journal of Psychiatry,
30(5), 615–621. https://doi.org/10.1016/j.eurpsy.2015.02.006. 204(1), 20–29. https://doi.org/10.1192/bjp.bp.112.116285.
Cochran, W. G. (1954). The combination of estimates from different experiments. Kapur, S. (2003). Psychosis as a state of aberrant salience: A framework linking biology,
Biometrics, 10, 101–129. https://doi.org/10.2307/3001666. phenomenology and pharmacology in schizophrenia. American Journal of Psychiatry,
Dudley, R., Taylor, P., Wickham, S., & Hutton, P. (2016). Psychosis, delusions and the 160, 13–23. https://doi.org/10.1176/appi.ajp.160.1.13.
"Jumping to Conclusions" reasoning bias: A systematic review and meta-analysis. Köther, U., Vettorazzi, E., Veckenstedt, R., Hottenrott, B., Bohn, F., Scheu, F., ... Moritz, S.
Schizophrenia Bulletin, 42(3), 652–665. https://doi.org/10.1093/schbul/sbv150. (2017). Bayesian analyses of the effect of metacognitive training on social cognition
Egger, M., Smith, G. D., Schneider, M., & Minder, C. (1997). Bias in meta-analysis de- deficits and overconfidence in errors. Experimental Psychopathology, 8(2), 158–174.
tected by a simple, graphical test. BMJ, 315(7109), 629–634. https://doi.org/10. https://doi.org/10.5127/jep.054516.
1136/bmj.315.7109.629. Kowalski, J., Pankowski, D., Lew-Starowicz, M., & Lukasz, G. (2017). Do specific meta-
Eichner, C., & Berna, F. (2016). Acceptance and efficacy of metacognitive training (mct) cognitive training modules lead to specific cognitive changes among patients diag-
on positive symptoms and delusions in patients with schizophrenia: A meta-analysis nosed with schizophrenia? A single module effectiveness pilot study. Psychosis:
taking into account important moderators. Schizophrenia Bulletin, 42(4), 952–962. Psychological, Social and Integrative Approaches, 9(3), 254–259. https://doi.org/10.
https://doi.org/10.1093/schbul/sbv225. 1080/17522439.2017.1300186.
Eifler, S., Rausch, F., Schirmbeck, F., Veckenstedt, R., Mier, D., Esslinger, C., ... Zink, M. Kumar, D., Menon, M., Moritz, S., & Woodward, T. S. (2015). Using the back door:
(2015). Metamemory in schizophrenia: Retrospective confidence ratings interact Metacognitive training for psychosis. Psychosis: Psychological, Social and Integrative
with neurocognitive deficits. Psychiatry Research, 225(3), 596–603. https://doi.org/ Approaches, 7(2), 166–178. https://doi.org/10.1080/17522439.2014.913073.
10.1016/j.psychres.2014.11.040. Lam, K. C. K., Ho, C. P. S., Wa, J. C., Chan, S. M. Y., Yam, K. K. N., Yeung, O. S. F., ...
Eisenacher, S., Rausch, F., Mier, D., Fenske, S., Veckenstedt, R., Englisch, S., ... Zink, M. Balzan, R. P. (2015). Metacognitive training (MCT) for schizophrenia improves
(2016). Bias against disconfirmatory evidence in the ‘at-risk mental state’ and during cognitive insight: A randomized controlled trial in a Chinese sample with schizo-
psychosis. Psychiatry Research, 238, 242–250. https://doi.org/10.1016/j.psychres. phrenia spectrum disorders. Behaviour Research and Therapy, 64, 38–42. https://doi.
2016.02.028. org/10.1016/j.brat.2014.11.008.
Eisenacher, S., & Zink, M. (2017). Holding on to false beliefs: The bias against dis- Lepage, M., Sergerie, K., Pelletier, M., & Harvey, P. O. (2007). Episodic memory bias and
confirmatory evidence over the course of psychosis. Journal of Behavior Therapy and the symptoms of schizophrenia. Canadian Journal of Psychiatry, 52(11), 702–709.
Experimental Psychiatry, 56, 79–89. https://doi.org/10.1016/j.jbtep.2016.08.015. https://doi.org/10.1177/070674370705201104.
Erawati, E., Keliat, B. A., Helena, N., & Hamid, A. (2014). The influence of metacognitive Lester, K. J., Mathews, A., Davison, P. S., Burgess, J. L., & Yiend, J. (2011). Modifying
training on delusion severity and metacognitive ability in schizophrenia. Journal of cognitive errors promotes cognitive well being: A new approach to bias modification.
Psychiatric and Mental Health Nursing, 21(9), 841–847. https://doi.org/10.1111/jpm. Journal of Behavior Therapy and Experimental Psychiatry, 42(3), 298–308. https://doi.
12130. org/10.1016/j.jbtep.2011.01.001.
Favrod, J., Brana, A., Nguyen, A., Nicolier, A., Perret, M., & Rexhaj, S. (2015). Effects of Lincoln, T. M., Rief, W., Westrmann, S., Ziegler, M., Kesting, M. L., Heibach, E., & Mehl, S.
metacognitive training on the awareness of delusions. Journal de Therapie (2014). Who stays, who benefits? Predicting dropout and change in cognitive be-
Comportementale et Cognitive, 25(3), 117–124. https://doi.org/10.1016/j.jtcc.2015. haviour therapy for psychosis. Psychiatry Research, 216(2), 198–205. https://doi.org/
06.002. 10.1016/j.psychres.2014.02.012.
Favrod, J., Maire, A., Bardy, S., Pernier, S., & Bonsack, C. (2011). Improving insight into Lipsey, M. W., & Wilson, D. B. (2001). Practical meta-analysis. Thousand Oaks, CA, US:
delusions: A pilot study of metacognitive training for patients with schizophrenia. Sage Publications, Inc.
Journal of Advanced Nursing, 67(2), 401–407. https://doi.org/10.1111/j.1365-2648. Marker, K. (2003). COGPACK manual version 5.9. Ladenburg: Marker Software.
2010.05470.x. McKenna, P. J., Leucht, S., Jauhar, S., Laws, K. R., & Bighelli, I. (2019). The controversy
Favrod, J., Rexhaj, S., Bardy, S., Ferrari, P., Hayoz, C., Moritz, S., ... Bonsack, C. (2014). about cognitive behavioural therapy for schizophrenia. World Psychiatry, 18(2),
Sustained antipsychotic effect of metacognitive training in psychosis: A randomized- 235–236. https://doi.org/10.1002/wps.20636.
controlled study. European Psychiatry, 29(5), 275–281. https://doi.org/10.1016/j. McLean, B. F., Mattiske, J. K., & Balzan, R. P. (2017). Association of the jumping to
eurpsy.2013.08.003. conclusions and evidence integration biases with delusions in psychosis: A detailed
Fragkos, K. C., Tsagris, M., & Frangos, C. C. (2014). Publication Bias in meta-analysis: meta-analysis. Schizophrenia Bulletin, 43(2), 344–354. https://doi.org/10.1093/
Confidence intervals for Rosenthal's fail-safe number. International Scholarly Research schbul/sbw056.
Notices, 2014, 1–17. https://doi.org/10.1155/2014/825383. Medalia, A., & Choi, J. (2009). Cognitive remediation in schizophrenia. Neuropsychology
Garety, P., & Freeman, D. (2013a). The past and future of delusions research: From the Review, 19(3), 353–364. https://doi.org/10.1007/s11065-009-9097-y.
12
Menon, M., Quilty, L. C., Zawadzki, J. A., Woodward, T. S., Sokolowski, H. M., Boon, H. Francophone de Clinique Comportementale et Cognitive, 18(4), 5–19.
S., & Wong, A. H. C. (2013). The role of cognitive biases and personality variables in Philipp, R., Kriston, L., Lanio, J., Kuhne, F., Harter, M., Moritz, S., & Meister, R. (2018).
subclinical delusional ideation. Cognitive Neuropsychiatry, 18(3), 208–218. https:// Effectiveness of metacognitive interventions for mental disorders in adults-a sys-
doi.org/10.1080/13546805.2012.692873. tematic review and meta-analysis (METACOG). Clinical Psychology & Psychotherapy,
Moher, D., Liberati, A., Tetzlaff, J., Altman, D. G., & Group, P (2009). Preferred reporting 26(2), 227–240. https://doi.org/10.1002/cpp.2345.
items for systematic reviews and meta-analyses: The PRISMA statement. BMJ, 339, Potvin, S. (2014). La méta-analyse: Illustration pour déterminer si la toxicomanie aggrave
b2535. https://doi.org/10.1136/bmj.b2535. les déficits cognitifs chez les personnes avec une schizophrénie. In M. Corbière, & N.
Moritz, S., Andreou, C., Schneider, B. C., Wittekind, C. E., Menon, M., Balzan, R. P., & Larivière (Eds.). Méthodes qualitatives, quantitatives et mixtes dans la recherche en sci-
Woodward, T. S. (2014). Sowing the seeds of doubt: A narrative review on meta- ences humaines, sociales et de la santé (pp. 166–188). Québec, Canada: Presses de
cognitive training in schizophrenia. Clinical Psychology Review, 34(4), 358–366. l'Université du Québec.
https://doi.org/10.1016/j.cpr.2014.04.004. Poyraz, B. C., Arikan, M. K., Poyraz, C. A., Turan, S., Kani, A. S., Aydin, E., & Ince, E.
Moritz, S., Bohn, F., Veckenstedt, R., Leighton, L., Woodward, T. S., & Hauschildt, M. (2016). Clinical and cognitive insight in patients with acute-phase psychosis:
(2016). Manual for individualized metacognitive therapy program for psychosis (MCT+) Association with treatment and neuropsychological functioning. Nordic Journal of
(2 ed.). Hamburg, Germany: VanHam Campus Press. Psychiatry, 70(7), 528–535. https://doi.org/10.1080/08039488.2016.1178328.
Moritz, S., Goritz, A. S., Balzan, R. P., Gaweda, L., Kulagin, S. C., & Andreou, C. (2017). A Roberts, D. L., Penn, D. L., & Combs, D. R. (2015). Social cognition and interaction training
new paradigm to measure probabilistic reasoning and a possible answer to the (SCIT): Group psychotherapy for schizophrenia and other psychotic disorders, clinician
question why psychosis-prone individuals jump to conclusions. Journal of Abnormal guide. New York, USA: Oxford University Press.
Psychology, 126(4), 406–415. https://doi.org/10.1037/abn0000262. Rosenthal, R. (1979). The file drawer problem and tolerance for null results. Psychological
Moritz, S., Goritz, A. S., Gallinat, J., Schafschetzy, M., Van Quaquebeke, N., Peters, M. J., Bulletin, 86(3), 638–641. https://doi.org/10.1037//0033-2909.86.3.638.
& Andreou, C. (2015). Subjective competence breeds overconfidence in errors in Rosenthal, R. (1991). Meta-analysis: A review. Psychosomatic Medicine, 53(3).
psychosis. A hubris account of paranoia. Journal of Behaviora Therapy Experimental Rosenthal, R. (1993). Meta-analytic procedures for social research. Newbury Park, CA: Sage
Psychiatry, 48, 118–124. https://doi.org/10.1016/j.jbtep.2015.02.011. Publications.
Moritz, S., Heeren, D., Andresen, B., & Krausz, M. (2001). An analysis of the specificity Ross, K., Freeman, D., Dunn, G., & Garety, P. (2011). A randomized experimental in-
and the syndromal correlates of verbal memory impairments in schizophrenia. vestigation of reasoning training for people with delusions. Schizophrenia Bulletin,
Psychiatry Research, 101(1), 23–31. https://doi.org/10.1016/s0165-1781(00) 37(2), 324–333. https://doi.org/10.1093/schbul/sbn165.
00241-9. Ross, R. M., McKay, R., Coltheart, M., & Langdon, R. (2015). Jumping to conclusions
Moritz, S., Kerstan, A., Veckenstedt, R., Randjbar, S., Vitzthum, F., Schmidt, C., ... about the beads task? A meta-analysis of delusional ideation and data-gathering.
Woodward, T. S. (2011). Further evidence for the efficacy of a metacognitive group Schizophrenia Bulletin, 41(5), 1183–1191. https://doi.org/10.1093/schbul/sbu187.
training in schizophrenia. Behaviour Research and Therapy, 49(3), 151–157. https:// Salemink, E., van den Hout, M., & Kindt, M. (2007). Trained interpretive bias: Validity
doi.org/10.1016/j.brat.2010.11.010. and effects on anxiety. Journal of Behavior Therapy and Experimental Psychiatry, 38(2),
Moritz, S., Mayer-Stassfurth, H., Endlich, L., Andreou, C., Ramdani, N., Petermann, F., & 212–224. https://doi.org/10.1016/j.jbtep.2006.10.010.
Balzan, R. P. (2015). The benefits of doubt: Cognitive Bias correction reduces hasty Salvatore, G., Lysaker, P. H., Popolo, R., Procacci, M., Carcione, A., & Dimaggio, G.
decision-making in schizophrenia. Cognitive Therapy and Research, 39(5), 627–635. (2012). Vulnerable self, poor understanding of others' minds, threat anticipation and
https://doi.org/10.1007/s10608-015-9690-8. cognitive biases as triggers for delusional experience in schizophrenia: A theoretical
Moritz, S., Pfuhl, G., Ludtke, T., Menon, M., Balzan, R. P., & Andreou, C. (2016). A two- model. Clinical Psychology & Psychotherapy, 19(3), 247–259. https://doi.org/10.
stage cognitive theory of the positive symptoms of psychosis. Highlighting the role of 1002/cpp.746.
lowered decision thresholds. Journal of Behavior Therapy and Experimental Psychiatry. Sanford, N., Veckenstedt, R., Moritz, S., Balzan, R., & Woodward, T. S. (2014). Impaired
https://doi.org/10.1016/j.jbtep.2016.07.004. integration of disambiguating evidence in delusional schizophrenia patients.
Moritz, S., Thoering, T., Kuhn, S., Willenborg, B., Westermann, S., & Nagel, M. (2015). Psychological Medicine, 44, 2729–2738. https://doi.org/10.1017/
Metacognition-augmented cognitive remediation training reduces jumping to con- S0033291714000397.
clusions and overconfidence but not neurocognitive deficits in psychosis. Frontiers in Sarin, F., & Wallin, L. (2014). Cognitive model and cognitive behavior therapy for schi-
Psychology, 6. https://doi.org/10.3389/fpsyg.2015.01048. zophrenia: An overview. Nordic Journal of Psychiatry, 68(3), 145–153. https://doi.
Moritz, S., Veckenstedt, R., Andreou, C., Bohn, F., Hottenrott, B., Leighton, L., ... Roesch- org/10.3109/08039488.2013.789074.
Ely, D. (2014). Sustained and "sleeper" effects of group metacognitive training for Savulich, G., Shergill, S., & Yiend, J. (2012). Biased cognition in psychosis. Journal of
schizophrenia a randomized clinical trial. JAMA Psychiatry, 71(10), 1103–1111. Experimental Psychopathology, 3(4), 514–536. https://doi.org/10.5127/jep.016711.
https://doi.org/10.1001/jamapsychiatry.2014.1038. So, S. H.-W., Chan, A. P., Chong, C. S.-Y., Wong, M. H.-M., Lo, W. T.-L., Chung, D. W.-S., &
Moritz, S., Veckenstedt, R., Bohn, F., Hottenrott, B., Scheu, F., Randjbar, S., ... Roesch-Ely, Chan, S. S. (2015). Metacognitive training for delusions (MCTd): Effectiveness on
D. (2013). Complementary group Metacognitive Training (MCT) reduces delusional data-gathering and belief flexibility in a Chinese sample. Frontiers in Psychology, 6.
ideation in schizophrenia. Schizophrenia Research, 151(1–3), 61–69. https://doi.org/ https://doi.org/10.3389/fpsyg.2015.00730.
10.1016/j.schres.2013.10.007. Speechley, W. J., Moritz, S., Ngan, E. T. C., & Woodward, T. S. (2012). Impaired evidence
Moritz, S., Veckenstedt, R., Hottenrott, B., Woodward, T. S., Randjbar, S., & Lincoln, T. M. integration and delusions in schizophrenia. Journal of Experimental Psychopathology,
(2010). Different sides of the same coin? Intercorrelations of cognitive biases in 3(4), 688–701. https://doi.org/10.5127/jep.018411.
schizophrenia. Cognitive Neuropsychiatry, 15(4), 406–421. https://doi.org/10.1080/ Steel, C., Wykes, T., Ruddle, A., Smith, G., Shah, D. M., & Holmes, E. A. (2010). Can we
13546800903399993. harness computerised cognitive bias modification to treat anxiety in schizophrenia? A
Moritz, S., Veckenstedt, R., Randjbar, S., Vitzthum, F., & Woodward, T. S. (2011). first step highlighting the role of mental imagery. Psychiatry Research, 178(3),
Antipsychotic treatment beyond antipsychotics: Metacognitive intervention for 451–455. https://doi.org/10.1016/j.psychres.2010.04.042.
schizophrenia patients improves delusional symptoms. Psychological Medicine, 41(9), Tandon, R., Nasrallah, H. A., & Keshavan, M. S. (2009). Schizophrenia, "just the facts" 4.
1823–1832. https://doi.org/10.1017/s0033291710002618. Clinical features and conceptualization. Schizophrenia Research, 110(1–3), 1–23.
Moritz, S., Vitzthum, F., Randjbar, S., Veckenstedt, R., & Woodward, T. S. (2010). https://doi.org/10.1016/j.schres.2009.03.005.
Detecting and defusing cognitive traps: Metacognitive intervention in schizophrenia. Turner, D. T., MacBeth, A., Larkin, A., Moritz, S., Campbell, A., Livingstone, K., & Hutton,
Current Opinion in Psychiatry, 23(6), 561–569. https://doi.org/10.1097/yco. P. (2018). The effect of reducing the “jumping to conclusions” Bias on treatment
0b013e32833d16a8. decision-making capacity in psychosis: A randomized controlled trial with mediation
Moritz, S., & Woodward, T. S. (2007a). Metacognitive training for schizophrenia patients analysis. Schizophrenia Bulletin, 136, 1–10. https://doi.org/10.1093/schbul/sby136.
(MCT): A pilot study on feasibility, treatment adherence, and subjective efficacy. Turner, R., Hoppitt, L., Hodgekins, J., Wilkinson, J., Mackintosh, B., & Fowler, D. (2011).
German Journal of Psychiatry, 10(3), 69–78. Cognitive bias modification in the treatment of social anxiety in early psychosis: A
Moritz, S., & Woodward, T. S. (2007b). Metacognitive training in schizophrenia: From single case series. Behavioural and Cognitive Psychotherapy, 39(3), 341–347. https://
basic research to knowledge translation and intervention. Current Opinion in doi.org/10.1017/s1352465810000706.
Psychiatry, 20(6), 619–625. https://doi.org/10.1097/yco.0b013e3282f0b8ed. Ussorio, D., Giusti, L., Wittekind, C. E., Bianchini, V., Malavolta, M., Pollice, R., ...
Nair, A., Palmer, E. C., Aleman, A., & David, A. S. (2014). Relationship between cognition, Roncone, R. (2016). Metacognitive training for young subjects (MCT young version)
clinical and cognitive insight in psychotic disorders: A review and meta-analysis. in the early stages of psychosis: Is the duration of untreated psychosis a limiting
Schizophrenia Research, 152(1), 191–200. https://doi.org/10.1016/j.schres.2013.11. factor? Psychology and Psychotherapy, 89(1), 50–65. https://doi.org/10.1111/papt.
033. 12059.
Naughton, M., Nulty, A., Abidin, Z., Davoren, M., O'Dwyer, S., & Kennedy, H. G. (2012). Van Oosterhout, B., Smit, F., Krabbendam, L., Castelein, S., Staring, A., & Van der Gaag,
Effects of group metacognitive training (MCT) on mental capacity and functioning in M. (2016). Metacognitive training for schizophrenia spectrum patients: A meta-
patients with psychosis in a secure forensic psychiatric hospital: A prospective-cohort analysis on outcome studies. Psychological Medicine, 46(1), 47–57. https://doi.org/
waiting list controlled study. BMC Research Notes, 5(1), 302. https://doi.org/10. 10.1017/s0033291715001105.
1186/1756-0500-5-302. Vohs, J. L., George, S., Leonhardt, B. L., & Lysaker, P. H. (2016). An integrative model of
Nuechterlein, K. H., Barch, D. M., Gold, J. M., Goldberg, T. E., Green, M. F., & Heaton, R. the impairments in insight in schizophrenia: Emerging research on causal factors and
K. (2004). Identification of separable cognitive factors in schizophrenia. Schizophrenia treatments. Expert Review of Neurotherapeutics, 16(10), 1193–1204. https://doi.org/
Research, 72(1), 29–39. https://doi.org/10.1016/j.schres.2004.09.007. 10.1080/14737175.2016.1199275.
van Oosterhout, B., Krabbendam, L., de Boer, K., Ferwerda, J., van der Helm, M., Stant, A. Waller, H., Emsley, R., Freeman, D., Bebbington, P., Dunn, G., Fowler, D., ... Garety, P.
D., & van der Gaag, M. (2014). Metacognitive group training for schizophrenia (2015). Thinking well: A randomised controlled feasibility study of a new CBT
spectrum patients with delusions: A randomized controlled trial. Psychological therapy targeting reasoning biases in people with distressing persecutory delusional
Medicine, 44(14), 3025–3035. https://doi.org/10.1017/s0033291714000555. beliefs. Journal of Behavior Therapy & Experimental Psychiatry, 48, 82–89. https://doi.
Orcel, S., Urben, S., Lambelet, F., Hafil, M., Jaugey, L., & Holzer, L. (2013). Group me- org/10.1016/j.jbtep.2015.02.007.
tacognitive training for adolescents with psychosis: Multiple case study. Revue Waller, H., Freeman, D., Jolley, S., Dunn, G., & Garety, P. (2011). Targeting reasoning
13
biases in delusions: A pilot study of the Maudsley Review Training Programme for Katie M. Lavigne is a post-doctoral fellow at McGill University. Her research focuses on
individuals with persistent, high conviction delusions. Journal of Behavior Therapy neuroimaging, cognitive biases and cognition in schizophrenia.
and Experimental Psychiatry, 42(3), 414–421. https://doi.org/10.1016/j.jbtep.2011.
03.001.
Woodward, T. S., Buchy, L., Moritz, S., & Liotti, M. (2007). A bias against disconfirmatory Gabrielle Pochiet is a graduate student at McGill University. Her research examines
evidence is associated with delusion proneness in a nonclinical sample. Schizophrenia factors affecting cognitive performance in first-episode psychosis patients.
Bulletin, 33(4), 1023–1028. https://doi.org/10.1093/schbul/sbm013.
Yanos, P. T., & Moos, R. H. (2007). Determinants of functioning and well-being among Mathieu B. Brodeur is a research agent at the Douglas Mental Health University
individuals with schizophrenia: An integrated model. Clinical Psychology Review, Institute. He has worked on contextualized cognition in psychosis and healthy popula-
27(1), 58–77. https://doi.org/10.1016/j.cpr.2005.12.008. tions.
Geneviève Sauvé is a doctoral student at Université du Québec À Montréal completing a Martin Lepage is a professor of psychiatry at McGill University and the deputy scientific
second PhD in clinical psychology. Her research focuses on cognitive remediation and director at the Douglas Mental Health University Institute. His research explores psy-
occupational functioning in schizophrenia. chological interventions, neurocognition and neuroimaging in psychosis.
14

Sauve Et Al. 2020

Uploaded by

Copyright:

Available Formats

Sauve Et Al. 2020

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Sauve Et Al. 2020

Uploaded by

Copyright:

Available Formats

Clinical Psychology Review 78 (2020) 101854

Contents lists available at ScienceDirect

Clinical Psychology Review

Eﬃcacy of psychological interventions targeting cognitive biases in T

• Psychological interventions systematically targeting cognitive biases were reviewed.

1. Introduction characteristics, individuals with SZ&RP show systematic cognitive

3.2.3. Insight symptoms, characteristics of included studies assessing insight were

4.2. Clinical implications 4.4. Conclusions and future directions

You might also like