Individualized Diagnostic and Prognostic Models for Patients With Psychosis Risk Syndromes: A Meta-analytic View on the State of the Art

Abstract

BACKGROUND: The clinical high risk (CHR) paradigm has facilitated research into the underpinnings of help-seeking individuals at risk for developing psychosis, aiming at predicting and possibly preventing transition to the overt disorder. Statistical methods such as machine learning and Cox regression have provided the methodological basis for this research by enabling the construction of diagnostic models (i.e., distinguishing CHR individuals from healthy individuals) and prognostic models (i.e., predicting a future outcome) based on different data modalities, including clinical, neurocognitive, and neurobiological data. However, their translation to clinical practice is still hindered by the high heterogeneity of both CHR populations and methodologies applied. METHODS: We systematically reviewed the literature on diagnostic and prognostic models built on Cox regression and machine learning. Furthermore, we conducted a meta-analysis on prediction performances investigating heterogeneity of methodological approaches and data modality. RESULTS: A total of 44 articles were included, covering 3707 individuals for prognostic studies and 1052 individuals for diagnostic studies (572 CHR patients and 480 healthy control subjects). CHR patients could be classified against healthy control subjects with 78% sensitivity and 77% specificity. Across prognostic models, sensitivity reached 67% and specificity reached 78%. Machine learning models outperformed those applying Cox regression by 10% sensitivity. There was a publication bias for prognostic studies yet no other moderator effects. CONCLUSIONS: Our results may be driven by substantial clinical and methodological heterogeneity currently affecting several aspects of the CHR field and limiting the clinical implementability of the proposed models. We discuss conceptual and methodological harmonization strategies to facilitate more reliable and generalizable models for future clinical practice.

Meta-analyzing the prevalence and prognostic effect of antipsychotic exposure in clinical high-risk (CHR): when things are not what they seem

“…While this conflation is widespread in the literature and it is probably an unintended consequence of the neglect of the original notion of functional equivalent of transition to psychosis (Yung et al, 2003), it certainly has important consequences that should be duly noted and amended. First, current prediction models and prognostic algorithms based on CHR populations (Sanfelici, Dwyer, Antonucci, & Koutsouleris, 2020), if combining baseline AP-exposed and AP-naïve CHR, are likely to be substantially flawed (e.g. Ciarleglio et al, 2019; Zhang et al, 2018).…”

Preventive psychiatry: a blueprint for improving the mental health of young people

“…Given their complexity, the weighting of the predictors may vary considerably with context (e.g., adolescent vs. young adult, geographic contexts). For those models that may reach higher levels of proof for clinical utility, the implementation pathway is a perilous journey undermined by several obstacles, related to individuals involved (e.g., making their data available or accepting the outputs of the risk calculators), clinicians (e.g., adherence to the recommendations made by prediction models, communicating risks), providers (e.g., confidentiality of data, interpretability of outputs) and funders/ organizations (implementing standard prediction procedures) 238 .…”

“…Modern advancements in the field of individualized prediction modelling aim to consolidate stratified (tailored to subgroups) or precision (tailored to the individual subject) preventive psychiatry in young people 237 . Several individualized risk prediction models for forecasting the onset of psychosis, bipolar and depression/anxiety in young people 238 (see Table 5) have been externally validated in terms of prognostic accuracy, which is an essential step to address the extent to which predictions can be generalized to the data from plausibly related settings.…”

Association between formal thought disorders, neurocognition and functioning in the early stages of psychosis: a systematic review of the last half-century studies

“…[60] Furthermore, Mota et al [61] showed that speech disorganization measured by graph connectedness could correctly predict schizophrenia diagnosis at 6-month follow-up with 91.67% accuracy in patients undergoing first clinical contact for recent-onset psychosis and 21 wellmatched healthy subjects. Despite promising, these findings need to be externally replicated in greater samples to fulfil the challenging state-of-art criterion of machine learning applications in computational psychiatry [62].…”