healthcare Article Heterogeneity in Response to MCT and Psychoeducation: A Feasibility Study Using Latent Class Mixed Models in First-Episode Psychosis Marta Ferrer-Quintero 1,2,3 , Daniel Fernández 3,4,5 , Raquel López-Carrilero 1,3,6 , Luciana Díaz-Cutraro 1,6,7 , Marina Verdaguer-Rodríguez 1,6,8 , Helena García-Mieres 9,10,11 , Elena Huerta-Ramos 1,3 , Juana Gómez-Benito 2,12 , Trini Peláez 1,3 , Irene Birulés 1,2 , Ana Barajas 13,14 , Esther Pousa 3,15 , Marisa Barrigón 16 , Alfonso Gutiérrez-Zotes 3,17 , Eva Grasa 3,15 , Isabel Ruiz-Delgado 18 , Esther Lorente-Rovira 3,19 , Jordi Cid 20 , on behalf of the Spanish Metacognition Study Group † and Susana Ochoa 1,3, * 1 Parc Sanitari Sant Joan de Déu, Sant Boi de Llobregat, 08830 Barcelona, Spain 2 Department of Social and Quantitative Psychology, Faculty of Psychology, University of Barcelona, 08035 Barcelona, Spain 3 Centro de Investigación Biomédica en Red de Salud Mental, Instituto de Salud Carlos III, 28029 Madrid, Spain 4 Serra Húnter Fellow, Department of Statistics and Operations Research (DEIO), Universitat Politècnica de Catalunya, 08028 Barcelona, Spain 5 Institute of Mathematics of UPC—BarcelonaTech (IMTech), 08028 Barcelona, Spain Citation: Ferrer-Quintero, M.; 6 Sant Joan de Déu Research Foundation, 08950 Esplugues de Llobregat, Spain Fernández, D.; López-Carrilero, R.; 7 COMSAL Research Group, Psychology Department, FPCEE Blanquerna, Ramon Llull University, Díaz-Cutraro, L.; 08022 Barcelona, Spain Verdaguer-Rodríguez, M.; 8 Departament de Psicologia Clínica i de la Salut, Facultat de Psicologia, Universitat Autònoma de Barcelona, García-Mieres, H.; Huerta-Ramos, E.; 08193 Bellaterra, Spain Gómez-Benito, J.; Peláez, T.; Birulés, 9 Health Services Research Group, Instituto Hospital del Mar de Investigaciones Médicas, I.; et al. Heterogeneity in Response to 08003 Barcelona, Spain 10 Centro de Investigación Biomédica en Red de Epidemiología y Salud Pública, Instituto de Salud Carlos III, MCT and Psychoeducation: A Feasibility Study Using Latent Class 28029 Madrid, Spain 11 Departamento de Medicina y Ciencias de la Vida, Universitat Pompeu Fabra, 08003 Barcelona, Spain Mixed Models in First-Episode 12 Grupo de Estudios de Invariancia de la Medida y Análisis del Cambio (GEIMAC), Institut de Neurociències, Psychosis. Healthcare 2022, 10, 2155. Universitat de Barcelona, 08035 Barcelona, Spain https://doi.org/10.3390/ 13 Serra Húnter Fellow Departament de Psicologia Clínica i de la Salut, Facultat de Psicologia, Universitat healthcare10112155 Autònoma de Barcelona, 08193 Bellaterra, Spain 14 Department of Research, Centre d’Higiene Mental Les Corts, 08029 Barcelona, Spain Academic Editors: Jose 15 Department of Psychiatry, Hospital de la Santa Creu i Sant Pau, Institut d’Investigació Biomèdica-Sant Antonio Monreal and Pau (IIB-Sant Pau), 08025, Barcelona, Spain Abraham Rudnick 16 Departament of Psychiatry, University Hospital Virgen del Rocio, 41013 Seville, Spain 17 Hospital Universitari Institut Pere Mata, Institut d’Investigació Sanitària Pere Virgili, CERCA, Universitat Received: 9 September 2022 Rovira i Virgili, Reus, Spain, 43206 Reus, Spain Accepted: 22 October 2022 18 Unidad de Salud Mental Comunitaria Malaga Norte, 29014 Málaga, Spain Published: 28 October 2022 19 Psychiatry Service, Hospital Clínico Universitario de Valencia, 46010 Valencia, Spain 20 Mental Health & Addiction Research Group, IdiBGi, Institut d’Assistencia Sanitària, 17190 Girona, Spain Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in * Correspondence: susana.ochoa@sjd.es † Collaborators of the Spanish Metacognition Study Group (SMSG) are indicated published maps and institutional affil- in the Acknowledgment section. iations. Abstract: Metacognitive training (MCT) is an effective treatment for psychosis. Longitudinal trajecto- ries of treatment response are unknown but could point to strategies to maximize treatment efficacy Copyright: © 2022 by the authors. during the first episodes. This work aims to explore the possible benefit of using latent class mixed Licensee MDPI, Basel, Switzerland. models (LCMMs) to understand how treatment response differs between metacognitive training and This article is an open access article psychoeducation. We conducted LCMMs in 28 patients that received MCT and 34 patients that re- distributed under the terms and ceived psychoeducation. We found that MCT is effective in improving cognitive insight in all patients conditions of the Creative Commons but that these effects wane at follow-up. In contrast, psychoeducation does not improve cognitive Attribution (CC BY) license (https:// insight, and may increase self-certainty in a group of patients. These results suggest that LCMMs are creativecommons.org/licenses/by/ valuable tools that can aid in treatment prescription and in predicting response to specific treatments. 4.0/). Healthcare 2022, 10, 2155. https://doi.org/10.3390/healthcare10112155 https://www.mdpi.com/journal/healthcare Healthcare 2022, 10, 2155 2 of 13 Keywords: first-episode psychosis; social cognition; metacognition; latent class mixed models; metacognitive training; psychoeducation 1. Introduction People that experience a first episode of psychosis (FEP) have highly variable outcomes, which range from sustained remission to treatment resistance from onset [1]. Despite efforts in identifying early treatment strategies [2], it is still a clinical challenge to deliver optimized treatment to prevent relapse and functional decline [3]. Cognitive behavioral models of psychosis have fostered the development of psycholog- ical interventions that target the cognitive biases involved in the genesis and maintenance of psychosis [4]. These have been studied as promising treatments for psychosis because in the past decades, despite considerable advances in pharmacological treatment, the out- comes of psychosis have not improved significantly [5]. Metacognitive training (MCT) [6] has emerged as one of the most effective psychologi- cal treatments for positive symptoms of psychosis [7]. MCT combines psychoeducation, cognitive bias modification, and strategy teaching to correct data-gathering biases. This ap- proach aims to sow the seeds of doubt [8,9] by considering cognitive biases as a deviation from normality. Systematic-review findings have supported its efficacy in improving positive and negative symptoms, self-esteem, and functioning in schizophrenia [7,10], and recent studies have proven that it is also a valid intervention for people with FEP [11]. These broad beneficial effects may be rooted in that MCT intervenes over most of the cognitive biases and cognitive constructs that have established evidence of their impor- tance in the genesis of psychosis, its maintenance, and outcomes. MCT includes sessions that work on domains of social cognition such as facial emotion recognition and theory of mind, both of which are strong predictors of outcome [12]. Likewise, it targets the jumping- to-conclusions bias (JTC), which has been repeatedly associated with delusions, poorer neurocognition, and measures of outcome [13–17]. A consistent finding in the literature is that MCT improves cognitive insight by increas- ing self-reflectivity and reducing self-certainty [18], the two domains that compose cognitive insight. Cognitive insight refers to the set of cognitive processes that permit questioning one’s beliefs and appraisals and reevaluating anomalous experiences or misinterpreta- tions [19]. Self-reflectivity refers to a person’s ability for introspection and willingness to admit fallibility. Conversely, self-certainty refers to the confidence a person has in their beliefs and judgments [19]. It is suggested that the formula for good cognitive insight is high self-reflectivity and low self-certainty [20]. This is because self-reflectivity has usually been associated with better outcomes and treatment response [21], while self-certainty is associated with more delusions and worse cognitive function [22,23]. Although meta-analytic findings have supported the efficacy of MCT in psychosis [7,24], this is usually at a medium effect size [24]. However, this seems to be the case for most effective psychological interventions for psychosis [25]. One reason for it may be that most clinical trials report averaged results, which blurs the vast heterogeneity in psychosis [26] and does not permit detecting the patients that indeed benefit from an intervention and their clinical characteristics. Furthermore, some of the MCT effects are apparent after a sleeper effect [27], but other patients need extended treatment to consolidate the effects [28]. Understanding the heterogeneity in response to MCT could point to better treatment strategies addressed to the specific characteristics of each patient. In particular, this is a vexing issue in first-episode psychosis because early targeted treatment may help promote recovery and prevent relapse [29]. Unsupervised learning methods such as latent class mixed models (LCMMs, also known as growth mixture models) are a useful approach for studying longitudinal trajecto- ries of latent variables. LCMMs have the advantage of being able to capture inter-individual differences in intra-individual change over time while preserving the heterogeneity of Healthcare 2022, 10, 2155 3 of 13 the population [30]. Another characteristic of the LCMM method is that it is a data-driven method, which allows studying longitudinal changes of a variable as it occurs naturally. Previous studies have used LCMMs to understand trajectories of illness or out- come [31,32], but [31,32] this method has not been used to explore trajectories of cognitive biases in response to different psychological interventions. We sought to study the feasibility of using LCMMs to understand patterns of change in cognitive insight in individuals with FEP who have received MCT. To test whether this is a specific effect of MCT, we compared latent trajectories between patients that received MCT and a group of patients who received psychoeducation. Psychoeducation is a psychological intervention that has proven to be effective in promoting better social and functional outcomes [33], but according to the literature, psychoeducation does not seem to improve cognitive insight [34,35]. Thus, the psychoeducation group will help compare the effects of MCT as opposed to how cognitive insight changes naturally. 2. Materials and Methods This study aims to provide proof-of-concept evidence of using LCMMs to understand patterns of change in response to psychotherapy. We conducted a secondary analysis of data from a blind, multicentric clinical trial that has been published elsewhere [11]. Briefly, the original study recruited 126 patients with FEP from nine participating mental health centers: Servicio Andaluz de Salud of Jaén, Málaga and Motril (Granada), Salut Mental Parc Taulí (Sabadell), Hospital de Santa Creu i Sant Pau (Barcelona), Centro de Higiene Mental Les Corts (Barcelona), Institut d’Assistència Sanitària Girona, Hospital Clínico Universitario de Valencia, and Parc Sanitari Sant Joan de Déu (coordinating center). Inclu- sion criteria were a diagnosis of schizophrenia, psychotic disorder not otherwise specified, delusional disorder, schizoaffective disorder, brief psychotic disorder, or schizophreniform disorder (according to DSM-IV-TR); (2) <5 years from the onset of symptoms; (3) a score of ≥3 in item delusions, grandiosity, or suspicions of PANSS in the previous year. Exclu- sion criteria were: traumatic brain injury, dementia, or intellectual disability (premorbid IQ ≤ 70); (2) substance dependence; and (3) PANSS ≥5 in hostile and uncooperativeness, or ≥6 in suspiciousness. This was chosen to avoid altering the dynamics of the group interventions. Participants were randomized to receive either MCT or psychoeducation. Psychoeducation included modules on healthy habits, risk behaviors, prevention of relapse, video forums, resources for work, leisure activities, and community resources. The orig- inal study randomized participants using blocks of four from a list of random numbers, and 55 patients were allocated to psychoeducation while 67 received MCT. The remaining participants were either excluded or declined to participate in the study. All participants were assessed at three points: baseline, post-treatment, and six-month follow-up. The assessment included the following: Sociodemographic questionnaire: Data on sociodemographic variables were collected on-site. Diagnosis and treatment were collected from the clinical history of the participants. We transformed the antipsychotic treatment to olanzapine defined daily dose (DDD) [36]. Clinical measures: The Positive and Negative Syndrome Scale (PANSS) [37,38] was used to measure symptom severity. The Spanish version of the Scale Unawareness of Mental Disorders (SUMD) [39,40] was used to measure unawareness of the mental disorder. Higher scores represent more unawareness of the mental disorder. We used the Rosenberg Self-Esteem Scale [41], where higher scores indicate better self-esteem. Metacognition: The Beck Cognitive Insight Scale (BCIS) [20,42] was used to measure cognitive insight. The BCIS is composed of two subscales: self-certainty and self-reflectivity, which are analyzed separately. Higher scores in self-reflectivity represent more ability to question one’s beliefs. Higher scores in self-certainty represent more certainty in one’s interpretations and misinterpretations. The beads task [43] was used to measure the JTC. Participants were shown a picture of two containers filled with 100 colored beads in re- ciprocal proportions. We used three trials with different conditions: a probabilistic trial with an 85/15 ratio, a second probabilistic trial with a 60/40 ratio, and a final trial with Healthcare 2022, 10, 2155 4 of 13 an affective condition in a 60/40 ratio. Participants were told that the computer had se- lected a container and that the goal of the task was to determine which container. To this aim, participants were shown one bead at a time. The participant was instructed to see as many beads as they needed to guess what container the beads came from. Our outcome variable was the draws to decision in the three probabilistic conditions. Less than 3 draws to decision is considered indicative of presenting the JTC bias. Social cognition: The Internal, Personal and Situational Attributions Questionnaire (IPSAQ) [44] was used to assess attributional style. We used two indexes: personalizing bias and externalizing bias. Personalizing bias refers to a tendency to blame others rather than circumstances for negative events. Externalizing bias refers to a tendency to attribute the causes of negative events to others or circumstances rather than to oneself [45]. The faces test [46,47] was used to measure emotion recognition. A reduced version of the hinting task [48,49] was used to measure theory of mind. Global functioning: The Global Assessment of Functioning (GAF) [50] was used to measure global functioning on a scale of 0–100. Higher scores represent better functioning. The Ethics Committee of each participating center approved this project. LCMMs were conducted using the R package lcmm [51] from the statistical soft- ware R version 4.0.2 [52]. This unsupervised learning technique classifies individuals into groupings with similar trajectory patterns, called latent classes. Following the strategy of Nagin et al. [53], we fit the respective models for the outcomes of interest (self-reflectivity and self-certainty) at the three points of assessment, for which the time metric was the time at the assessment (baseline, post-treatment, and follow-up). The number of group tra- jectories was determined by analyzing 2–6 group models without covariates. Model selection to determine the optimal number of latent trajectories was performed according to the Bayesian information criterion (BIC), where a lower value indicates a better fit [54]. Average posterior probabilities above 70% were checked as well [55]. Individuals were allo- cated to classes according to maximum a posteriori criterium (MAP). To ensure adequate results, in this work we only included patients with complete data on self-reflectivity and self-certainty at the three points of assessment. At baseline, the MCT and the psychoeducation groups were compared with chi-square tests for categorical variables and Student t-tests for continuous variables. To compare trajectories, we used Mann–Whitney U tests, for which the effect size was reported with ranked biserial correlations. 3. Results The final sample included 62 participants with complete data on self-reflectivity and self-certainty at the three points of assessment. At baseline, the two groups only differed in that the MCT group had higher baseline global functioning than the psychoeducation group (t(60) = −2.857, p = 0.006). Table 1 presents baseline data of the two groups. Table 1. Baseline differences in sociodemographic, clinical, social cognitive, and metacognitive variables between the psychoeducation and the MCT groups. MCT (n = 28) Psychoeducation (n = 34) Mean/% SD Mean/% SD t/χ2 p Age Sex (% males) 75.00 73.50 0.17 0.90 Number of hospital admissions Education (%) 4.665 0.46 Incomplete primary education 21.40 11.80 Complete primary education 28.60 17.60 Incomplete secondary education 10.70 20.60 Complete secondary education 25.00 26.50 Incomplete superior education 3.60 14.70 Healthcare 2022, 10, 2155 5 of 13 Table 1. Cont. MCT (n = 28) Psychoeducation (n = 34) Mean/% SD Mean/% SD t/χ2 p Complete superior education 10.70 8.80 Diagnosis (%) 4.929 0.43 Schizophrenia 46.40 55.90 Non-specified psychotic disorder 21.40 11.80 Schizoaffective disorder 3.60 11.80 Delusional disorder 7.10 11.80 Brief psychotic episode 10.70 11.80 Schizophreniform disorder 7.10 8.80 PANSS Positive symptoms 12.25 3.874 12.53 4.487 0.26 0.80 Negative symptoms 14.54 7.162 15.15 4.794 0.40 0.69 General symptoms 27.36 7.790 27.26 5.920 −0.05 0.96 Total 54.14 15.717 54.94 12.110 0.23 0.82 GAF 66.61 13.331 57.50 11.758 −2.86 0.01 SUMD global score 5.82 2.695 6.03 3.639 0.25 0.80 Beads task 85–15 3.6296 3.85455 5.3529 4.27737 1.632 0.11 60–40 6.4444 3.99358 8.5294 5.16536 1.726 0.09 Affective 6.3704 4.50767 7.7941 4.11781 1.286 1.286 IPSAQ Externalizing bias 0.1481 3.44968 1.5588 3.85488 1.486 0.14 Personalizing bias 1.2569 0.72667 1.2183 0.50730 −0.24 0.81 Hinting task 4.7143 1.01314 4.6765 0.97610 −0.15 0.88 Rosenberg self-esteem 27.5714 6.42004 27.3824 5.03933 −0.13 0.90 Faces test 17.6786 1.82683 17.4706 1.39773 −0.51 0.61 Estimated premorbid IQ 99.4231 14.85442 96.7188 13.71421 −0.72 0.48 3.1. Cognitive Insight Trajectories in the MCT Group We found two trajectories for self-reflectivity and two trajectories for self-certainty. The four trajectories are summarized in Figures 1 and 2. Differences between each pair of trajectories at baseline, post-test, and follow-up can be found in Supplementary Table S1. Figure 1. Trajectories of self-reflectivity in the MCT group. The red trajectory depicts the “High SR” trajectory (n = 20). The second trajectory (in blue), refers to “Improving SR” (n = 8). Healthcare 2022, 10, 2155 6 of 13 Figure 1. Trajectories of self-reflectivity in the MCT group. The red trajectory depicts the “High SR” trajectory (n = 20). The second trajectory (in blue), refers to “Improving SR” (n = 8). Figure 2. Trajectories Figure Trajectoriesofofself-certainty self-certaintyinin thethe MCT group: MCT “Initial group: Decline” “Initial (in red, Decline” n = 22) (in red, n =and 22)“Re- and sponsive SC” (in blue, n = 6). “Responsive SC” (in blue, n = 6). 3.1.1. 3.1.1. Self-Reflectivity Self-Reflectivity We We found found aa trajectory trajectory (in (in red), red),henceforth henceforth “High “High SR”SR” (n (n == 20), 20), that that grouped grouped patients patients with high baseline with high baselineself-reflectivity. self-reflectivity.TheThe “High “High SR”SR” trajectory trajectory included included patients patients whosewhose self- self-reflectivity reflectivity was stable during the baseline and post-treatment assessments but declinedde- was stable during the baseline and post-treatment assessments but at clined at follow-up. follow-up. The second The second trajectory trajectory (in blue), (in blue),“Improving named named “Improving SR” (n = SR” (n = 8), 8), grouped grouped patients with low baseline self-reflectivity. The “Improving SR” trajectory pre- sented an improvement in self-reflectivity at follow-up, which declined to a level below baseline at the 6-month follow-up. Figure 1 shows the graphical representations of the two trajectories of self-reflectivity in the MCT group. At baseline, the two self-reflectivity trajectories differed in negative (U = 20.00, p = 0.002) and general (U = 22.00, p = 0.003) symptoms, global functioning (U = 23.00, p = 0.004), and self-esteem (U = 29.50, p = 0.011). Patients in the “High SR” trajectory presented better scores in all the variables, indicating a better clinical state at the moment of the assessment. At post-test, the two trajectories still differed in negative symptoms (U = 12.00, p = 0.004), general symptoms (U = 21.50, p = 0.026), global functioning (U = 20.50, p = 0.021), and self-esteem (U = 27.00, p = 0.05). However, at this point of assessment, the “Improving SR” trajectory had significantly worse clinical insight (U = 24.00, p = 0.040). Finally, at the 6-month follow-up, the two trajectories only differed in negative (U = 22.00, p = 0.03) and general symptoms (U = 10.00, p = 0.010) and clinical insight (U = 20.00, p = 0.020). 3.1.2. Self-Certainty For self-certainty, we also found two trajectories in the MCT group. The “Initial Decline” (in red, n = 22) trajectory grouped patients with moderate levels of self-certainty at baseline. At post-test, the scores in self-certainty declined, but they increased again to baseline levels at the 6-month follow-up. The second trajectory (in blue, n = 6), henceforth “Responsive SC”, comprised patients with very high self-certainty at baseline. The scores in self-certainty presented a steep decline at post-test but increased again at follow-up. However, at follow-up, the scores in self-reflectivity did not reach baseline levels. The two trajectories of self-certainty in the MCT group are depicted in Figure 2. Healthcare 2022, 10, 2155 7 of 13 At baseline, the two trajectories of self-certainty did not differ in any variable. However, at post-test, the “Responsive SC” trajectory had worse positive symptoms (U = 18.50, p = 0.015). The difference in positive symptoms was maintained at follow-up (U = 20.50, p = 0.025), but patients in the “Initial Decline” trajectory also displayed worse general symptoms (U = 23.50, p = 0.05) at this point of assessment. 3.2. Cognitive Insight Trajectories in the Psychoeducation Group 3.2.1. Self-Reflectivity Healthcare 2022, 10, x FOR PEER REVIEW 8 of 13 In the psychoeducation group, we also found two trajectories for self-reflectivity. These are depicted in Figure 3. Figure3.3.Trajectories Figure Trajectoriesofofself-reflectivity self-reflectivityinin the the psychoeducation psychoeducation group. group. TheThe redred lineline refers refers to the to the “Low- “Low-SR” trajectory (n = 18). The second trajectory (in blue), refers to the “Worsening-SR” SR” trajectory (n = 18). The second trajectory (in blue), refers to the “Worsening-SR” (n = 16) trajectory.(n = 16) trajectory. Concerning self-reflectivity, the first trajectory (in red), henceforth “Low-SR” (n = 18), 3.2.2. Self-Certainty included patients that had a low baseline level of self-reflectivity. In this trajectory, self- reflectivity consistently Two trajectories of declined over were self-certainty time.apparent The second trajectory in the (in blue), henceforth psychoeducation group. The “Worsening-SR” first trajectory (in(nred), = 16), comprised named “Low andpatients Stableexhibiting self-reflectivity SC” (n = 23), that was included patients withhigh low at baseline but declined steeply over time. baseline self-certainty, which remained stable throughout the three assessments. The sec- ond At baseline, trajectory (in the two blue), trajectories named of self-reflectivity “Worsening differed in SC” (n = 11), included diagnosis patients with(χ2 = base- high 9.86, pline = 0.04). Participants in the “Low-SR” trajectory had a bigger proportion of self-certainty, which increased significantly after intervention and then decreased to patients diagnosed with schizophrenia baseline levels at follow-up. as Theopposed to other graphical diagnoses in representation ofthe spectrum these of psychosis. trajectories can be At post-test, found in Figure 4. participants in the “Low-SR” trajectory reported more personalizing bias (U = 47.00, p = 0.05). However, this difference was not maintained at follow-up. At the At baseline, the two trajectories of self-certainty differed in academic background follow-up point of assessment, the two trajectories of self-reflectivity in the psychoeducation (χ2= 11.98, p = 0.035), positive symptoms (U = 70.50, p = 0.040), clinical insight (U = 31.50, group did not differ in any variable. p < 0.001), the 60–40 condition of the beads task (U= 57.50, p = 0.011), and facial emotion recognition (U = 64.00, p = 0.019). 3.2.2. Self-Certainty At post-test, the Worsening SC trajectory presented more positive (U = 27.50, p = Two trajectories of self-certainty were apparent in the psychoeducation group. The first 0.014) and negative (U = 34.50, p = 0.029) symptoms, worse global functioning (U = 37.50, trajectory (in red), named “Low and Stable SC” (n = 23), included patients with low base- p = 0.042), worse clinical insight (U = 33.00, p = 0.003), and less draws to decision in the line self-certainty, which remained stable throughout the three assessments. The second three conditions of the beads task (U = 34.00, p = 0.023; U = 32.50, p = 0.021; U = 38.50, p = trajectory (in blue), named “Worsening SC” (n = 11), included patients with high baseline 0.043). self-certainty, which increased significantly after intervention and then decreased to base- However, at follow up, the two trajectories only differed in positive (U = 25.00, p = line levels at follow-up. The graphical representation of these trajectories can be found 0.009) and negative symptoms (U = 30.00, p = 0.020), clinical insight (U = 30.50, p = 0.005), in Figure 4. and theory of mind (U = 38.50, p = 0.05). Healthcare 2022, Healthcare 2022, 10, 10, 2155 x FOR PEER REVIEW 98 of of 13 Figure 4. Trajectories of of self-certainty self-certaintyin inthe thepsychoeducation psychoeducationgroup. group.InInred is is red depicted thethe depicted “Low “Low and Stable SC” SC”(n(n==23) 23)trajectory. The trajectory. second The trajectory, second in blue, trajectory, refers in blue, to the refers to“Worsening SC” (n the “Worsening = SC” 11) trajectory. (n = 11) trajectory. 4. Discussion At baseline, the two trajectories of self-certainty differed in academic background (χ2 =As 11.98, p = 0.035), predicted positive symptoms by previous (U = we literature [18], 70.50, p = 0.040), found clinicalthat that patients insight (U = 31.50, received MCT pimproved < 0.001), cognitive the 60–40 insight condition by of the beads task maintaining good(U= p = 0.011), 57.50,levels baseline and facial emotion of self-reflectivity or im- recognition proving low(U p = 0.019). = 64.00,levels baseline and reducing self-certainty. This was an expected result since At post-test, the Worsening the original study [11] found solid SC trajectory evidencepresented more positive of the improvement in (U p = 0.014) = 27.50, insight cognitive in and negative (U = 34.50, p = 0.029) symptoms, worse global patients who received MCT compared to those who received psychoeducation. Byfunctioning (U = 37.50, p = 0.042), using worse LCMMs, clinical insightobtain we could (U = 33.00, a deeperp = 0.003), view of and howlesspatients draws to decision respond toineach the three conditions intervention. of theInbeads task (U = 34.00, p = 0.023; U = 32.50, p = 0.021; U = 38.50, p = absence of a specific intervention on cognitive insight, it seems that self-reflectivity 0.043). tendsHowever, to decrease at steadily follow up, the two trajectories independent of its baseline only differed level, whichinwas positive (U = by evidenced 25.00, the ptwo = 0.009) and negative symptoms (U = 30.00, p = 0.020), clinical insight trajectories in the psychoeducation group. This finding is consistent with a previous (U = 30.50, p = 0.005), and theory of mind (U = 38.50, p = 0.05). study comparing psychoeducation and metacognitive training that found that psy- choeducation 4. Discussion is not an effective treatment to improve self-reflectivity [35]. Our results showed that metacognitive training is useful in maintaining high baseline As predicted by previous literature [18], we found that patients that received MCT self-reflectivity or improving low baseline self-reflectivity. However, trajectory analysis improved cognitive insight by maintaining good baseline levels of self-reflectivity or suggests that self-reflectivity tends to decrease after the intervention. In this sense, improving low baseline levels and reducing self-certainty. This was an expected result LCMMs allowed us to detect that MCT may be beneficial in improving cognitive insight since the original study [11] found solid evidence of the improvement in cognitive insight for all patients, regardless of their baseline levels. Furthermore, we found that both self- in patients who received MCT compared to those who received psychoeducation. By using reflectivity and self-certainty experienced a steep decline at follow-up, which suggests the LCMMs, we could obtain a deeper view of how patients respond to each intervention. need for maintained booster sessions of MCT to stabilize the effects. These results are con- In absence of a specific intervention on cognitive insight, it seems that self-reflectivity sistent with recent evidence showing that extended therapy could strengthen its positive tends to decrease steadily independent of its baseline level, which was evidenced by the two effects [9,28]. trajectories in the psychoeducation group. This finding is consistent with a previous study A previous study has suggested sustained and sleeper effects of MCT in people with comparing psychoeducation and metacognitive training that found that psychoeducation psychosis [27]. In treatment is not an effective this work,to the authorsself-reflectivity improve found that improvements [35]. in delusions that were not significant Our resultsatshowed the post-test became significant that metacognitive trainingatisa useful three-year follow-up.high in maintaining Because our baseline follow-up was at six months, we were not able to detect whether self-reflectivity or improving low baseline self-reflectivity. However, trajectory analysisany trajectory experi- enced more suggests that sustained effects, self-reflectivity tendsor whether to decreasetheafter trajectories may change the intervention. In this over time sense, as the LCMMs sleeper effect becomes apparent. allowed us to detect that MCT may be beneficial in improving cognitive insight for all There patients, was a surprising regardless finding. of their baseline Psychoeducation levels. Furthermore, we seemed foundto thatbeboth detrimental to pa- self-reflectivity tients and with high self-certainty self-certainty experiencedata baseline. At baseline, steep decline participants at follow-up, whichin this trajectory suggests the needexpe- for rienced more positive symptoms, less clinical insight, a bigger tendency to make hasty decisions (jumping to conclusions), and worse facial emotion recognition than their Healthcare 2022, 10, 2155 9 of 13 maintained booster sessions of MCT to stabilize the effects. These results are consistent with recent evidence showing that extended therapy could strengthen its positive effects [9,28]. A previous study has suggested sustained and sleeper effects of MCT in people with psychosis [27]. In this work, the authors found that improvements in delusions that were not significant at the post-test became significant at a three-year follow-up. Because our follow-up was at six months, we were not able to detect whether any trajectory experienced more sustained effects, or whether the trajectories may change over time as the sleeper effect becomes apparent. There was a surprising finding. Psychoeducation seemed to be detrimental to patients with high self-certainty at baseline. At baseline, participants in this trajectory experienced more positive symptoms, less clinical insight, a bigger tendency to make hasty decisions (jumping to conclusions), and worse facial emotion recognition than their counterparts. These differences were not apparent in the self-certainty trajectories of the MCT group. Moreover, at post-test, patients in the Worsening SC group presented fewer draws to decision in the three conditions of the beads task, which suggests a more pronounced tendency to present the jumping-to-conclusions bias. Furthermore, both self-certainty and the jumping-to-conclusions bias are strongly related to delusions [14,23], and clinical insight and facial emotion recognition are strong predictors of outcome and functioning in patients with psychosis [12,56]. Because this trajectory had worse scores in these domains, we speculate that it represents a group with a higher risk of relapse, although this should be tested in future studies. Although the scope of this study precludes us from drawing conclusions on the pos- sible mechanisms of worsening self-certainty in response to psychoeducation, our inter- pretation is that there may be an interaction between unawareness of the disease, positive symptoms, and poor facial emotion recognition. These patients also had poor clinical insight and thus may interpret information on the disease as threatening and react by jump- ing to the conclusion that their thoughts and experiences are certain. These results suggest that patients that have social cognitive and metacognitive difficulties may benefit from starting psychological treatment with a normalizing approach that reduces data-gathering biases and improves insight, such as MCT. The findings of this study must be interpreted considering several limitations: First, the sample size in the two groups was small, rendering the study underpowered and limiting our ability to compare longitudinal outcomes. Similarly, the sample size of each trajectory precluded us from obtaining predictors for each trajectory. Finally, our follow- up data only extended to six months after the intervention, and we could not detect sleeper effects. These limitations notwithstanding, our findings support the use of LCMMs to study cognitive biases in response to psychotherapy and highlight the heterogeneous nature of psychosis. Future studies including broader samples comparing more interventions will help detect the chances of responding to a specific intervention and detect precision treat- ment strategies to prescribe psychological treatment based on the individual characteristics of each person. This study was not designed to identify predictors of outcomes after MCT. Rather, the present study aimed to offer proof-of-concept evidence of the added value of using LCMMs to improve our current knowledge of psychological interventions for psychosis. However, using LCMMs with larger samples may help detect what specific trajectories have better chances of responding to an intervention and what their predictors are and may identify potential moderators. Furthermore, this approach is not only valid for people with psychosis, as most current psychological interventions for mental illness have sound theoretical foundations and strong evidence supporting their efficacy. Healthcare 2022, 10, 2155 10 of 13 5. Conclusions This work supports the added value of using LCMMs to understand how specific psychological interventions exert differential effects on cognitive biases while considering the variability of the patients’ responses. Specifically, we found that MCT improves cogni- tive insight in all participants with first-episode psychosis regardless of their baseline level. Conversely, psychoeducation does not affect cognitive insight and may trigger adverse effects in some patients. Finally, LCMMs could be a useful approach to detect predictors of response to different psychological treatments and to develop early targeted treatment for people with psychosis. Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/healthcare10112155/s1, Table S1: Baseline, post-test, and follow- up characteristics of each trajectory in the MCT group; Table S2: Baseline, post-test, and follow-up characteristics of each trajectory in the psychoeducation group. Author Contributions: M.F.-Q., S.O. were involved in the conception and drafting of the manuscript. D.F. conducted statistical analysis. R.L.-C., I.B., E.L.-R., E.G., T.P., M.B., H.G.-M., J.G.-B., A.B., E.P., A.G.-Z. and I.R.-D. collected data and revised the manuscript. L.D.-C., M.V.-R., E.H.-R. and J.C. reviewed and revised the manuscript. Members of the Spanish Metacognition Study Group provided resources and referred participants to the study. S.O. obtained funding for the project. All authors have read and agreed to the published version of the manuscript. Funding: This study was funded by the Instituto de Salud Carlos III, Spanish Government, (PI11/01347, PI14/00044 and PI18/00212); the Fondo Europeo de Desarrollo Regional (FEDER), Health Department of Catalonia, PERIS call (SLT006/17/00231); Progress and Health Foundation of the Andalusian Regional Ministry of Health (PI-0634/2011 and PI-0193/2014); Obra Social La Caixa (RecerCaixa call 2013), Obra Social Sant Joan de Déu, BML (RTI2018-100927-J-I00) administrated by Ministerio de Ciencia e Innovación (MCI, Spain), by the Agencia Estatal de Investigación (AEI, Spain), and by the European Regional Development Fund (FEDER, UE); Daniel Fernández has been supported by grant 2017 SGR 622 (GRBIO) administrated by the Departament d’Economia i Coneixement de la Generalitat de Catalunya (Spain) and by the Ministerio de Ciencia e Innovación (Spain) [PID2019- 104830RB-I00/ DOI (AEI): 10.13039/501100011033]; and CIBER, Consorcio Centro de Investigación Biomédica en Red, Instituto de Salud Carlos III, Ministerio de Ciencia e Innovación. Institutional Review Board Statement: The study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of Parc Sanitari Sant Joan de Déu. Informed Consent Statement: Informed consent was obtained from all subjects involved in the study. Written informed consent has been obtained from the patient(s) to publish this paper. Data Availability Statement: The data presented in this study are available on request from the cor- responding author. Acknowledgments: We thank all the clinicians that referred patients to this study. Likewise, we thank all the participants for their time, effort, and altruism. We thank all the members of the Spanish Metacognition Study Group for their valuable insights and comments. # Spanish Metacognition Study Group (SMSG): Acevedo A., Anglès J., Argany M.A., Barajas A., Barrigón M.L., Beltrán M., Birulés I., Bogas J.L., Cabezas A., Camprubí N., Carbonero M., Carrasco E., Casañas R., Cid J., Conesa E., Corripio I., Cortes P., Crosas J.M., de Apraiz A., Delgado M., Domínguez L., Escartí M.J., Escudero A., Esteban Pinos I., Franco C., García C., Gil V., Gonzalez-Casares R., González Higueras F., González- Montoro Mª.L., González E., Grasa E., Guasp A., Gutierrez-Zotes A., Huerta-Ramos Mª.E., Huertas P., Jiménez-Díaz A., Lalucat L.L., LLacer B., López-Carrilero R., Lorente E., Luengo A., Mantecón N., Mas-Expósito L., Montes M., Moritz S., Murgui E., Nuñez M., Ochoa S., Palomer E., Paniego E., Peláez T., Pérez V., Planell K., Planellas C., Pleguezuelo-Garrote P., Pousa E., Rabella M., Renovell M., Rubio R., Ruiz-Delgado I., San Emeterio M., Sánchez E., Sanjuán J., Sans B., Schilling L., Sió H., Teixidó M., Torres P., Vila M.A., Vila-Badia R., Villegas F., Villellas R. Conflicts of Interest: The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results. Healthcare 2022, 10, 2155 11 of 13 References 1. Hall, M.-H.; Holton, K.M.; Öngür, D.; Montrose, D.; Keshavan, M.S. Longitudinal Trajectory of Early Functional Recovery in Patients with First Episode Psychosis. Schizophr. Res. 2019, 209, 234–244. [CrossRef] [PubMed] 2. Liemburg, E.J.; Castelein, S.; van Es, F.; Scholte-Stalenhoef, A.N.; van de Willige, G.; Smid, H.; Visser, E.; Knegtering, H.; Bruggeman, R. The Psychosis Recent Onset GRoningen Survey (PROGR-S): Defining Dimensions and Improving Outcomes in Early Psychosis. PLoS ONE 2014, 9, e113521. [CrossRef] [PubMed] 3. Suvisaari, J.; Mantere, O.; Keinänen, J.; Mäntylä, T.; Rikandi, E.; Lindgren, M.; Kieseppä, T.; Raij, T.T. Is It Possible to Predict the Future in First-Episode Psychosis? Front. Psychiatry 2018, 9, 580. [CrossRef] 4. Gaw˛eda, Ł.; Moritz, S.; Ochoa, S.; So, S.H. Editorial: The Relationship Between Cognitive Biases and Psychosis: Searching for Mechanisms. Front. Psychiatry 2021, 12, 753317. [CrossRef] 5. Saha, S.; Chant, D.; Welham, J.; McGrath, J. A Systematic Review of the Prevalence of Schizophrenia. PLoS Med. 2005, 2, e141. [CrossRef] 6. Moritz, S.; Woodward, T.S. Metacognitive Training in Schizophrenia: From Basic Research to Knowledge Translation and Intervention. Curr. Opin. Psychiatry 2007, 20, 619–625. [CrossRef] [PubMed] 7. Penney, D.; Sauvé, G.; Mendelson, D.; Thibaudeau, É.; Moritz, S.; Lepage, M.; Mental, D. Immediate and Sustained Outcomes and Moderators Associated With Metacognitive Training for Psychosis: A Systematic Review and Meta-Analysis. JAMA 2022, 79, 417–429. [CrossRef] [PubMed] 8. Moritz, S.; Andreou, C.; Schneider, B.C.; Wittekind, C.E.; Menon, M.; Balzan, R.P.; Woodward, T.S. Sowing the Seeds of Doubt: A Narrative Review on Metacognitive Training in Schizophrenia. Clin. Psychol. Rev. 2014, 34, 358–366. [CrossRef] 9. Moritz, S.; Klein, J.P.; Lysaker, P.H.; Mehl, S. Metacognitive and Cognitive-Behavioral Interventions for Psychosis: New develop- ments. Dialogues Clin. Neurosci. 2019, 21, 309. [CrossRef] [PubMed] 10. Jiang, J.; Zhang, L.; Zhu, Z.; Li, W.; Li, C. Metacognitive Training for Schizophrenia: A Systematic Review. Shanghai Arch. Psychiatry 2015, 27, 149–157. [CrossRef] 11. Ochoa, S.; López-Carrilero, R.; Barrigón, M.L.; Pousa, E.; Barajas, A.; Lorente-Rovira, E.; González-Higueras, F.; Grasa, E.; Ruiz-Delgado, I.; Cid, J.; et al. Randomized Control Trial to Assess the Efficacy of Metacognitive Training Compared with a Psycho-Educational Group in People with a Recent-Onset Psychosis. Psychol. Med. 2017, 47, 1573–1584. [CrossRef] [PubMed] 12. Fett, A.K.J.; Viechtbauer, W.; Dominguez, M.d.G.; Penn, D.L.; van Os, J.; Krabbendam, L. The Relationship between Neurocog- nition and Social Cognition with Functional Outcomes in Schizophrenia: A Meta-Analysis. Neurosci. Biobehav. Rev. 2011, 35, 573–588. [CrossRef] [PubMed] 13. Falcone, M.A.; Murray, R.M.; O’Connor, J.A.; Hockey, L.N.; Gardner-Sood, P.; Di Forti, M.; Freeman, D.; Jolley, S. Jumping to Conclusions and the Persistence of Delusional Beliefs in First Episode Psychosis. Schizophr. Res. 2015, 165, 243–246. [CrossRef] 14. Falcone, M.A.; Murray, R.M.; Wiffen, B.D.R.; O’Connor, J.A.; Russo, M.; Kolliakou, A.; Stilo, S.; Taylor, H.; Gardner-Sood, P.; Paparelli, A.; et al. Jumping to Conclusions, Neuropsychological Functioning, and Delusional Beliefs in First Episode Psychosis. Schizophr. Bull. 2015, 41, 411–418. [CrossRef] 15. Tripoli, G.; Quattrone, D.; Ferraro, L.; Gayer-Anderson, C.; Rodriguez, V.; La Cascia, C.; La Barbera, D.; Sartorio, C.; Seminerio, F.; Tarricone, I.; et al. Jumping to Conclusions, General Intelligence, and Psychosis Liability: Findings from the Multi-Centre EU-GEI Case-Control Study. Psychol. Med. 2021, 51, 623–633. [CrossRef] 16. Díaz-Cutraro, L.; López-Carrilero, R.; García-Mieres, H.; Ferrer-Quintero, M.; Verdaguer-Rodriguez, M.; Barajas, A.; Grasa, E.; Pousa, E.; Lorente, E.; Barrigón, M.L.; et al. The Relationship between Jumping to Conclusions and Social Cognition in First-Episode Psychosis. Schizophrenia 2022, 8, 1–7. [CrossRef] 17. Díaz-Cutraro, L.; García-Mieres, H.; López-Carrilero, R.; Ferrer, M.; Verdaguer-Rodriguez, M.; Barrigón, M.L.; Barajas, A.; Grasa, E.; Pousa, E.; Lorente, E.; et al. Jumping to Conclusions Is Differently Associated with Specific Subtypes of Delusional Experiences: An Exploratory Study in First-Episode Psychosis. Schizophr. Res. 2021, 228, 357–359. [CrossRef] 18. Lopez-Morinigo, J.-D.; Ajnakina, O.; Martínez, A.S.-E.; Escobedo-Aedo, P.-J.; Ruiz-Ruano, V.G.; Sánchez-Alonso, S.; Mata-Iturralde, L.; Muñoz-Lorenzo, L.; Ochoa, S.; Baca-García, E.; et al. Can Metacognitive Interventions Improve Insight in Schizophrenia Spectrum Disorders? A Systematic Review and Meta-Analysis. Psychol. Med. 2020, 50, 2289. [CrossRef] 19. Lysaker, P.H.; Pattison, M.L.; Leonhardt, B.L.; Phelps, S.; Vohs, J.L. Insight in Schizophrenia Spectrum Disorders: Relationship with Behavior, Mood and Perceived Quality of Life, Underlying Causes and Emerging Treatments. World Psychiatry 2018, 17, 12–23. [CrossRef] 20. Beck, A.T.; Baruch, E.; Balter, J.M.; Steer, R.A.; Warman, D.M. A New Instrument for Measuring Insight: The Beck Cognitive Insight Scale. Schizophr. Res. 2004, 68, 319–329. [CrossRef] 21. O’Connor, J.A.; Ellett, L.; Ajnakina, O.; Schoeler, T.; Kollliakou, A.; Trotta, A.; Wiffen, B.D.; Falcone, A.M.; Di Forti, M.; Murray, R.M.; et al. Can Cognitive Insight Predict Symptom Remission in a First Episode Psychosis Cohort? BMC Psychiatry 2017, 17, 54. [CrossRef] [PubMed] 22. Cooke, M.A.; Peters, E.R.; Fannon, D.; Aasen, I.; Kuipers, E.; Kumari, V. Cognitive Insight in Psychosis: The Relationship between Self-Certainty and Self-Reflection Dimensions and Neuropsychological Measures. Psychiatry Res. 2010, 178, 284. [CrossRef] [PubMed] Healthcare 2022, 10, 2155 12 of 13 23. Engh, J.A.; Friis, S.; Birkenaes, A.B.; Jónsdóttir, H.; Klungsøyr, O.; Ringen, P.A.; Simonsen, C.; Vaskinn, A.; Opjordsmoen, S.; Andreassen, O.A. Delusions Are Associated with Poor Cognitive Insight in Schizophrenia. Schizophr. Bull. 2010, 36, 830–835. [CrossRef] 24. Liu, Y.C.; Tang, C.C.; Hung, T.T.; Tsai, P.C.; Lin, M.F. The Efficacy of Metacognitive Training for Delusions in Patients With Schizophrenia: A Meta-Analysis of Randomized Controlled Trials Informs Evidence-Based Practice. Worldviews Evidence-Based Nurs. 2018, 15, 130–139. [CrossRef] 25. Sauvé, G.; Lavigne, K.M.; Pochiet, G.; Brodeur, M.B.; Lepage, M. Efficacy of Psychological Interventions Targeting Cognitive Biases in Schizophrenia: A Systematic Review and Meta-Analysis. Clin. Psychol. Rev. 2020, 78, 101854. [CrossRef] 26. Moritz, S.; Mahlke, C.I.; Westermann, S.; Ruppelt, F.; Lysaker, P.H.; Bock, T.; Andreou, C. Embracing Psychosis: A Cognitive Insight Intervention Improves Personal Narratives and Meaning-Making in Patients With Schizophrenia. Schizophr. Bull. 2018, 44, 307–316. [CrossRef] 27. Moritz, S.; Veckenstedt, R.; Andreou, C.; Bohn, F.; Hottenrott, B.; Leighton, L.; Köther, U.; Woodward, T.S.; Treszl, A.; Menon, M.; et al. Sustained and “Sleeper” Effects of Group Metacognitive Training for Schizophrenia: A Randomized Clinical Trial. JAMA Psychiatry 2014, 71, 1103–1111. [CrossRef] 28. Ishikawa, R.; Ishigaki, T.; Shimada, T.; Tanoue, H.; Yoshinaga, N.; Oribe, N.; Morimoto, T.; Matsumoto, T.; Hosono, M. The Efficacy of Extended Metacognitive Training for Psychosis: A Randomized Controlled Trial. Schizophr. Res. 2020, 215, 399–407. [CrossRef] 29. Fusar-Poli, P.; McGorry, P.D.; Kane, J.M. Improving Outcomes of First-Episode Psychosis: An Overview. World Psychiatry 2017, 16, 251–265. [CrossRef] 30. Jung, T.; Wickrama, K.A.S. An Introduction to Latent Class Growth Analysis and Growth Mixture Modeling. Soc. Personal. Psychol. Compass 2008, 2, 302–317. [CrossRef] 31. Chang, W.C.; Ho, R.W.H.; Tang, J.Y.M.; Wong, C.S.M.; Hui, C.L.M.; Chan, S.K.W.; Lee, E.M.H.; Suen, Y.N.; Chen, E.Y.H. Early-Stage Negative Symptom Trajectories and Relationships With 13-Year Outcomes in First-Episode Nonaffective Psychosis. Schizophr. Bull. 2019, 45, 610–619. [CrossRef] [PubMed] 32. Salagre, E.; Grande, I.; Solé, B.; Mezquida, G.; Cuesta, M.J.; Díaz-Caneja, C.M.; Amoretti, S.; Lobo, A.; González-Pinto, A.; Moreno, C.; et al. Exploring Risk and Resilient Profiles for Functional Impairment and Baseline Predictors in a 2-Year Follow-Up First-Episode Psychosis Cohort Using Latent Class Growth Analysis. J. Clin. Med. 2020, 10, 1–20. [CrossRef] [PubMed] 33. Xia, J.; Merinder, L.B.; Belgamwar, M.R. Psychoeducation for Schizophrenia. Cochrane Database Syst. Rev. 2011, 6, 21–22. [CrossRef] 34. Dondé, C.; Senn, M.; Eche, J.; Kallel, L.; Saoud, M.; Brunelin, J. Well-Informed but Not Aware: The P.A.C.T.® Psychoeducation Program for Schizophrenia Improves Knowledge about, but Not Insight into, the Illness. Asian J. Psychiatr. 2019, 46, 15–18. [CrossRef] 35. Ahuir, M.; Cabezas, Á.; Miñano, M.J.; Algora, M.J.; Estrada, F.; Solé, M.; Gutiérrez-Zotes, A.; Tost, M.; Barbero, J.D.; Montalvo, I.; et al. Improvement in Cognitive Biases after Group Psychoeducation and Metacognitive Training in Recent-Onset Psychosis: A Randomized Crossover Clinical Trial. Psychiatry Res. 2018, 270, 720–723. [CrossRef] 36. Leucht, S.; Samara, M.; Heres, S.; Davis, J.M. Dose Equivalents for Antipsychotic Drugs: The DDD Method. Schizophr. Bull. 2016, 42, 90–94. [CrossRef] 37. Kay, S.R.; Fiszbein, A.; Opler, L.; Kay, S.R.; Fiszbein, A.; Opler, L.A.; Kay, S.R.; Fiszbein, A.; Opler, L. The Positive and Negative Syndrome Scale (PANSS) for Schizophrenia. Schizophr. Bull. 1987, 13, 261–276. [CrossRef] 38. Peralta, V.; Cuesta, M.J. Psychometric Properties of the Positive and Negative Syndrome Scale (PANSS) in Schizophrenia. Psychiatry Res. 1994, 53, 31–40. [CrossRef] 39. Amador, X.F.; Strauss, D.H.; Yale, S.A.; Flaum, M.M.; Endicott, J.; Gorman, J.M. Assessment of Insight in Psychosis. Am. J. Psychiatry 1993, 150, 873–879. [CrossRef] 40. Ruiz Ripoll, A.I.; Pousa, E.; Duñó, R.; Crosas, J.M.; Cuppa, S.; García-Ribera, C. Adaptación Al Español de La Escala de Valoración de La No Conciencia de Trastorno Mental (SUMD). Actas Esp. Psiquiatr. 2008, 36, 111–119. 41. Rosenberg, M. Society and the Adolescent Self-Image; Princeton University Press: Princeton, NJ, USA, 1965. 42. Gutiérrez-Zotes, J.A.; Valero, J.; Cortés, M.J.; Labad, A.; Ochoa, S.; Ahuir, M.; Carlson, J.; Bernardo, M.; Cañizares, S.; Escartin, G.; et al. Spanish Adaptation of the Beck Cognitive Insight Scale (BCIS) for Schizophrenia. Actas Esp. Psiquiatr. 2012, 40, 2–9. [PubMed] 43. Brett-Jones, J.; Garety, P.; Hemsley, D. Measuring Delusional Experiences: A Method and Its Application. Br. J. Clin. Psychol. 1987, 26, 257–265. [CrossRef] [PubMed] 44. Kinderman, P.; Bentall, R.P. The Development of a Novel Measure of Causal Attributions: The Internal Personal and Situational Attributions Questionnaire. Pers. Individ. Dif. 1996, 20, 261–264. [CrossRef] 45. Langdon, R.; Corner, T.; McLaren, J.; Ward, P.B.; Coltheart, M. Externalizing and Personalizing Biases in Persecutory Delusions: The Relationship with Poor Insight and Theory-of-Mind. Behav. Res. Ther. 2006, 44, 699–713. [CrossRef] 46. Baron-Cohen, S.; Wheelwright, S.; Jolliffe, T. Is There a “Language of the Eyes”? Evidence from Normal Adults, and Adults with Autism or Asperger Syndrome. Vis. Cogn. 1997, 4, 311–331. [CrossRef] 47. Huerta-Ramos, E.; Ferrer-Quintero, M.; Gómez-Benito, J.; González-Higueras, F.; Cuadras, D.; del Rey-Mejías, Á.L.; Usall, J.; Ochoa, S. Translation and Validation of Baron Cohen’s Face Test in a General Population from Spain. Actas Esp. Psiquiatr. 2020, in press. 48. Corcoran, R.; Mercer, G.; Frith, C.D. Schizophrenia, Symptomatology and Social Inference: Investigating “Theory of Mind” in People with Schizophrenia. Schizophr. Res. 1995, 17, 5–13. [CrossRef] Healthcare 2022, 10, 2155 13 of 13 49. Gil, D.; Fernández-Modamio, M.; Bengochea, R.; Arrieta, M. Adaptation of the Hinting Task Theory of the Mind Test to Spanish. Rev. Psiquiatr. Salud Ment. (Engl. Ed.) 2012, 5, 79–88. [CrossRef] 50. Endicott, J. The Global Assessment Scale: A Procedure for Measuring Overall Severity of Psychiatric Disturbance. Arch. Gen. Psychiatry 1976, 33, 766–771. [CrossRef] 51. Proust-Lima, C.; Philipps, V.; Liquet, B. Estimation of Extended Mixed Models Using Latent Classes and Latent Processes: The R Package Lcmm. J. Stat. Softw. 2017, 78, 1–56. [CrossRef] 52. R Core Team. R: A Language and Environment for Statistical Computing; R Foundation for Statistical Computing: Vienna, Austria, 2020. 53. Nagin, D.S.; Tremblay, R.E.; Sampson, R.J.; Laub, J.H. Developmental Trajectory Groups: Fact or a Useful Statistical Fiction? Criminology 2005, 43, 873–904. [CrossRef] 54. Tein, J.Y.; Coxe, S.; Cham, H. Statistical Power to Detect the Correct Number of Classes in Latent Profile Analysis. Struct. Equ. Model. 2013, 20, 640. [CrossRef] [PubMed] 55. Therneau, T.M.; Atkinson, E.J. An Introduction to Recursive Partitioning Using the RPART Routines. 2019. Available online: https://cran.r-project.org/web/packages/rpart/vignettes/longintro.pdf (accessed on 10 October 2022). 56. Wright, A.C.; Lysaker, P.H.; Fowler, D.; Greenwood, K. Clinical Insight in First Episode Psychosis: The Role of Metacognition. J. Ment. Health 2021, 17, 1–9. [CrossRef] [PubMed]