Prediction of estimated risk for bipolar disorder using machine learning and structural MRI features.

BACKGROUND: Individuals with bipolar disorder are commonly correctly diagnosed a decade after symptom onset. Machine learning techniques may aid in early recognition and reduce the disease burden. As both individuals at risk and those with a manifest disease display structural brain markers, structural magnetic resonance imaging may provide relevant classification features. METHODS: Following a pre-registered protocol, we trained linear support vector machine (SVM) to classify individuals according to their estimated risk for bipolar disorder using regional cortical thickness of help-seeking individuals from seven study sites (N = 276). We estimated the risk using three state-of-the-art assessment instruments (BPSS-P, BARS, EPIbipolar). RESULTS: For BPSS-P, SVM achieved a fair performance of Cohen's κ of 0.235 (95% CI 0.11-0.361) and a balanced accuracy of 63.1% (95% CI 55.9-70.3) in the 10-fold cross-validation. In the leave-one-site-out cross-validation, the model performed with a Cohen's κ of 0.128 (95% CI -0.069 to 0.325) and a balanced accuracy of 56.2% (95% CI 44.6-67.8). BARS and EPIbipolar could not be predicted. In post hoc analyses, regional surface area, subcortical volumes as well as hyperparameter optimization did not improve the performance. CONCLUSIONS: Individuals at risk for bipolar disorder, as assessed by BPSS-P, display brain structural alterations that can be detected using machine learning. The achieved performance is comparable to previous studies which attempted to classify patients with manifest disease and healthy controls. Unlike previous studies of bipolar risk, our multicenter design permitted a leave-one-site-out cross-validation. Whole-brain cortical thickness seems to be superior to other structural brain features.

Machine Learning Prediction of Estimated Risk for Bipolar Disorders Using Hippocampal Subfield and Amygdala Nuclei Volumes.

The pathophysiology of bipolar disorder (BD) remains mostly unclear. Yet, a valid biomarker is necessary to improve upon the early detection of this serious disorder. Patients with manifest BD display reduced volumes of the hippocampal subfields and amygdala nuclei. In this pre-registered analysis, we used structural MRI (n = 271, 7 sites) to compare volumes of hippocampus, amygdala and their subfields/nuclei between help-seeking subjects divided into risk groups for BD as estimated by BPSS-P, BARS and EPIbipolar. We performed between-group comparisons using linear mixed effects models for all three risk assessment tools. Additionally, we aimed to differentiate the risk groups using a linear support vector machine. We found no significant volume differences between the risk groups for all limbic structures during the main analysis. However, the SVM could still classify subjects at risk according to BPSS-P criteria with a balanced accuracy of 66.90% (95% CI 59.2-74.6) for 10-fold cross-validation and 61.9% (95% CI 52.0-71.9) for leave-one-site-out. Structural alterations of the hippocampus and amygdala may not be as pronounced in young people at risk; nonetheless, machine learning can predict the estimated risk for BD above chance. This suggests that neural changes may not merely be a consequence of BD and may have prognostic clinical value.

Remission with or without comorbid substance use disorders in early psychosis: long-term outcome in integrated care (ACCESS III study).

Introduction: Schizophrenia-Spectrum-Disorders are associated with poor long-term outcome as well as disability and often severely affect the lives of patients and their families often from symptom onset. Up to 70% of first episode psychosis (FEP) patients suffer from comorbid substance use disorders (SUD). We aimed at studying the course of illness in FEP patients within evidence-based care, with and without comorbid SUD, to examine how decreased, remitted or persistent substance use impacted rates of a combined symptomatic and functional long-term recovery compared with patients without SUD. Methods: ACCESS III is an integrated care model for FEP or patients in the early phase of non-affective and affective psychotic disorders. Treatment trajectories of patients, who had been in ACCESS care for 1 year, with and without SUD were compared with regard to the course of illness and quality of life using Mixed Model Repeated Measures (MMRM) and recovery rates were compared using binary logistic regression. Change in substance use was coded as either persistent, decreased/remitted or no use. Results: ACCESS III was a prospective 1-year study (N = 120) in patients aged 12-29 years. Of these, 74 (61.6%) had a comorbid SUD at admission. There were no group differences regarding the course of illness between patients with or without comorbid SUD or between patients with a substance abuse or substance dependence. The only outcome parameter that was affected by SUD was quality of life, with larger improvement found in the group without substance use (p = 0.05) compared to persistent and remitted users. Using LOCF, 44 patients (48.9%) fulfilled recovery criteria at the endpoint; recovery did not differ based on substance use status. Discussion: SUD and especially substance dependence are common in psychotic disorders even in FEP patients. Evidence-based integrated care led to long-term improvement in patients with comorbid SUD and rate of recovery did not differ for patients with substance use.

Mediators of quality of life change in people with severe psychotic disorders treated in integrated care: ACCESS II study.

BACKGROUND: Patients with severe psychotic disorders exhibit a severely reduced quality of life (QoL) at all stages of the disease. Integrated care often led to an improvement in QoL. However, the specific mediators of QoL change are not yet well understood. METHODS: The ACCESS II study is a prospective, long-term study investigating the effectiveness of an integrated care program for people with severe psychotic disorders (IC-TACT) that includes Therapeutic Assertive Community Treatment within a care network of in- and outpatient services at the University Medical Center Hamburg-Eppendorf, Germany. We examined longitudinal associations between QoL and the hypothesized mediators of change (i.e., negative symptoms, depression, and anxiety), using cross-lagged panel models. RESULTS: The sample includes 418 severely ill patients treated in IC-TACT for at least 1 year. QoL increased, whereas symptom severity decreased significantly from baseline to 6-month follow-up (p-values ≤ 0.001), and remained stable until 12-month follow-up. QoL and symptom severity demonstrated significant auto-correlated effects and significant cross-lagged effects from QoL at baseline to negative symptoms (6 months, ß = -0.20, p < 0.001) to QoL (12 months, ß = -0.19, p < 0.01) resulting in a significant indirect, mediated effect. Additionally, negative symptoms after 6 months had a significant effect on the severity of depression after 12 months (ß = 0.13, p < 0.05). CONCLUSIONS: Negative symptoms appear to represent an important mechanism of change in IC-TACT indicating that improvement of QoL could potentially be achieved through optimized intervention on negative symptoms. Moreover, this may lead to a reduction in the severity of depression after 12 months.

Health-related quality of life in severe psychotic disorders during integrated care: 5-year course, prediction and treatment implications (ACCESS II).

PURPOSE: Studies on outcomes mapping Quality of Life (QoL) as patient-reported outcome over a longer period in severe psychotic disorders are scarce. However, such data would be particularly important for structuring, implementing and operating effective and efficient care models and for promoting satisfaction with care, service engagement and adherence. METHODS: The ACCESS II study is a prospective long-term study of an integrated care model for people with severe psychotic disorders. The model includes Therapeutic Assertive Community Treatment within a cross-sectoral and interdisciplinary network. This publication analyses the course of QoL assessed with the Q-LES-Q-18 using a mixed model for repeated measures. RESULTS: Mapping the course of QoL in N = 329 participants, there is a significant increase in the first 6 weeks of treatment (early course). Comparison to a published norm show significant lower QoL for severe psychotic disorders. The variable having a traumatic event before the age of 18 was significantly negatively associated with QoL. A decrease in the severity of depressive as well as in positive symptomatology in the first six weeks after admission was associated with increase of QoL. CONCLUSION: Results indicate that the overall symptom burden at time of inclusion is not decisive for the perceived QoL in the long-term course while the reduction in the severity of depressive and positive symptoms is important. This means focusing even more on the treatment of depressive symptoms and include traumatherapeutic aspects in the long-term treatment of severe psychotic disorders if needed. TRAIL REGISTRATION: ClinicalTrials.gov (identifier: NCT01888627).

Individuals at increased risk for development of bipolar disorder display structural alterations similar to people with manifest disease.

In psychiatry, there has been a growing focus on identifying at-risk populations. For schizophrenia, these efforts have led to the development of early recognition and intervention measures. Despite a similar disease burden, the populations at risk of bipolar disorder have not been sufficiently characterized. Within the BipoLife consortium, we used magnetic resonance imaging (MRI) data from a multicenter study to assess structural gray matter alterations in N = 263 help-seeking individuals from seven study sites. We defined the risk using the EPIbipolar assessment tool as no-risk, low-risk, and high-risk and used a region-of-interest approach (ROI) based on the results of two large-scale multicenter studies of bipolar disorder by the ENIGMA working group. We detected significant differences in the thickness of the left pars opercularis (Cohen's d = 0.47, p = 0.024) between groups. The cortex was significantly thinner in high-risk individuals compared to those in the no-risk group (p = 0.011). We detected no differences in the hippocampal volume. Exploratory analyses revealed no significant differences in other cortical or subcortical regions. The thinner cortex in help-seeking individuals at risk of bipolar disorder is in line with previous findings in patients with the established disorder and corresponds to the region of the highest effect size in the ENIGMA study of cortical alterations. Structural alterations in prefrontal cortex might be a trait marker of bipolar risk. This is the largest structural MRI study of help-seeking individuals at increased risk of bipolar disorder.

Course, remission and recovery in patients with severe psychotic disorders with or without comorbid substance use disorders: Long-term outcome in evidence-based integrated care (ACCESS II study).

OBJECTIVES: People with psychotic disorders have a high prevalence of comorbid mental disorders, especially if severe mental illness (SMI) criteria are fulfilled. Substance Use Disorders (SUD) are the most common comorbidity. The aim of the study is to investigate whether SMI patients with and without comorbid SUD have a comparable course, remission and recovery rates within evidence-based care. METHODS: ACCESS is an integrated care model for patients with severe nonaffective and affective psychotic disorders. Treatment trajectories of patients, who have been in ACCESS care for at least 4 years, with and without SUD were compared with regard to the course of illness using Mixed Model Repeated Measures (MMRM) as well as recovery rates and its predictors. RESULTS: 187 of 312 patients (60%) were at least 4 years in ACCESS. Of these, 126 (67.4%) had a comorbid SUD at admission. Patients had on average 2.96 SUD, 87 (69%) had a dependence. Both groups improved significantly over 4 years in all outcome parameters. However, patients with substance dependence showed significantly worse outcomes in psychopathology (p < 0.001), functioning (p = 0.006) and quality of life (p = 0.026). Using LOCF, 44 patients (23.5%) fulfilled recovery criteria at endpoint. Comorbid substance use dependence was the only significant predictor for non-recovery (OR = 0.462, p = 0.048). CONCLUSION: SUD and especially substance dependence are common in psychotic disorders with SMI. Evidence-based integrated care also leads to long-term improvement in these patients, but to a lesser extent than in patients without SUD. In particular, the "optimal" outcome recovery is made more difficult by SUD dependence.

Study protocol for a randomised controlled trial evaluating an evidence-based, stepped and coordinated care service model for mental disorders (RECOVER).

INTRODUCTION: Healthcare systems around the world are looking for solutions to the growing problem of mental disorders. RECOVER is the synonym for an evidence-based, stepped and cross-sectoral coordinated care service model for mental disorders. RECOVER implements a cross-sectoral network with managed care, comprehensive psychological, somatic and social diagnostics, crisis resolution and a general structure of four severity levels, each with assigned evidence-based therapy models (eg, assertive community treatment) and therapies (eg, psychotherapy). The study rationale is the investigation of the effectiveness and efficiency of stepped and integrated care in comparison to standard care. METHODS AND ANALYSIS: The trial is conducted in accordance to the Standard Protocol Items: Recommendations for Interventional Trials Statement. The study aims to compare the RECOVER model with treatment as usual (TAU). The following questions are examined: Does RECOVER reduce healthcare costs compared with TAU? Does RECOVER improve patient-relevant outcomes? Is RECOVER cost-effective compared with TAU? A total sample of 890 patients with mental disorders will be assessed at baseline and individually randomised into RECOVER or TAU. Follow-up assessments are conducted after 6 and 12 months. As primary outcomes, cost reduction, improvement in symptoms, daily functioning and quality of life as well as cost-effectiveness ratios will be measured. In addition, several secondary outcomes will be assessed. Primary and secondary outcomes are evaluated according to the intention-to-treat principle. Mixed linear or logistic regression models are used with the direct maximum likelihood estimation procedure which results in unbiassed estimators under the missing-at-random assumption. Costs due to healthcare utilisation and productivity losses are evaluated using difference-in-difference regressions. ETHICS AND DISSEMINATION: Ethical approval from the ethics committee of the Hamburg Medical Association has been obtained (PV5672). The results will be disseminated to service users and their families via the media, to healthcare professionals via professional training and meetings and to researchers via conferences and publications. TRIAL REGISTRATION NUMBER AND REGISTRY NAME: ClinicalTrials.gov (NCT03459664), RECOVER PROTOCOL VERSION: 19 March 2020 (V.3.0).

[Innovative care models for people with schizophrenia]. / Innovative Versorgungsmodelle für Menschen mit schizophrenen Erkrankungen.

Schizophrenia is a complex mental disorder. Hence, the mental health system faces enormous structural, therapeutic, and health-economic challenges. Innovative models of healthcare can facilitate making treatment more efficient and effective.This paper aims to give an overview of evidence-based and innovative models of care and treatment for schizophrenia and severe mental illnesses. For this purpose, we performed a literature search covering the last 10 years of publications regarding "care or treatment models" and "schizophrenia," "psychosis," or "severe mental illness."Many publications do not provide details about the tested care model. Innovative care models for schizophrenia comprise early psychosis services (EPSs) or models for severe mental illnesses (SMI) integrating, combining and/or developing the treatment models crisis resolution team (CRT), assertive community treatment (ACT), and/or (intensive) case management (ICM). For illustration, the innovative models Early Psychosis Prevention and Intervention Centre (EPPIC) in Australia, the Flexible Assertive Community Treatment model in the Netherlands, and the Therapeutic Assertive Community Treatment model in Hamburg are explained. All three models regularly offer integrated, specialized, and assertive care.In Germany, innovative models of care have not been implemented sufficiently: neither EPS nor ACT are part of standard care; CRT can now be funded by health insurances via "ward-equivalent treatment" as defined by a new German law (§§ 39 und 115d SGB V). Regarding the implementation of CRT and ACT in Germany, a clinical study evaluating the RECOVER model is underway. This is a stepped-care, trans-sectorally coordinated and evidence-based treatment model, the evaluation of which is supported by the innovation fund of the Joint Federal Committee (Gemeinsamen Bundesausschuss, G­BA) from 2017 to 2020.

Comorbidity of chronic somatic diseases in patients with psychotic disorders and their influence on 4-year outcomes of integrated care treatment (ACCESS II study).

BACKGROUND: People with psychotic disorders fulfilling criteria of a severe and persistent mental illness (SPMI) display a high risk of somatic comorbidity (SC). METHODS: ACCESS II is a prospective, long-term study examining the effectiveness of Integrated Care for people with psychotic disorders fulfilling SPMI criteria. Chronic comorbid somatic disorders were systematically assessed according to ICD-10-GM criteria. Patients treated for ≥4years in ACCESS were categorized as early psychosis (treatment: ≤2years) or non-early psychosis (treatment: >2years) patients. RESULTS: Of 187 patients treated in ACCESS for ≥4years (mean age=41.8years, males=44.4%), 145 (77.5%) had SC, (mean=2.1±2.1). Overall, 55 different diseases from 15 different ICD-10-GM disease areas were identified. Prevalence of ≥1 SC (p=0.09) and specific types of SC (p=0.08-1.00) did not differ between early and non-early psychosis patients, but non-early psychosis patients had a higher mean number of SC (2.3±2.2 vs. 1.3±1.3, p=0.002). SC patients had higher rates of comorbid mental disorders (93% vs. 81%, p=0.002), specifically posttraumatic stress disorder (23% vs. 7%, p=0.002), and suicide attempts (43% vs. 19%, p<0.001). At the 4-year endpoint, both patients with and without comorbidity displayed major improvements in psychopathology, severity of illness, functioning, quality of life and satisfaction with care. CONCLUSIONS: SC is frequent in patients with severe psychotic disorders, even in the early psychosis phase. The magnitude of the problem underlines the need for regular screening, comprehensive assessment, preventive pharmacotherapy, and targeted SC management.