Cardiometabolic prediction models for young people with psychosis spectrum disorders in the UK (PsyMetRiC 2.0): a retrospective, multicohort clinical prediction model study

Background Young people with psychosis spectrum disorders are at a high risk of cardiometabolic morbidity and subsequent premature mortality, but there are no accurate clinic-ready prediction models for this group. We aimed to collaboratively refine, extend, and validate the Psychosis Metabolic Risk Calculator (PsyMetRiC) prediction models for accuracy, clinical usefulness, and acceptability, and to translate the models into a regulated, clinically available medical device. Methods In this retrospective, multicohort clinical prediction model study, we used primary care (Clinical Practice Research Datalink and QResearch) and secondary care (South London and Maudsley NHS Foundation Trust) datasets. Individuals from primary care sources were aged 16-35 years when they received a first recorded diagnosis of a psychosis-spectrum disorder between Jan 1, 2005, and Dec 31, 2015, with follow-up to Dec 31, 2020. Individuals from the secondary care source were enrolled in the psychosis early intervention service between Jan 1, 2012, and Dec 31, 2024. We developed models for a binary outcome of metabolic syndrome within 1-6 years using logistic regression; a time-to-event outcome of type 2 diabetes within 10 years using Weibull regression; and a binary outcome of clinically significant weight gain within 1 year using logistic regression. We revised existing predictors (hereafter referred to as the PsyMetRiC1 models) for finer detail and added new predictors: a family history of cardiometabolic disorder, antidepressant prescription, systolic blood pressure, and HbA_1C (hereafter PsyMetRiC2 models). Refinement and external validation were performed for metabolic syndrome models (PsyMetRiC1-MetS and PsyMetRiC2-MetS), and development and external validation were performed for the type 2 diabetes model (PsyMetRiC2-T2D). Development and internal validation were performed for the clinically significant weight gain model (PsyMetRiC2-WG), but external validation was not possible due to data availability. Partial versions without biochemical results were also developed for weight gain and metabolic syndrome models. We involved stakeholders including people with lived experience; and implemented the models in a web application compliant with regulatory standards in Great Britain. Findings In total, we included 25 850 individuals (male, n=13 614 [52·7%]; female, n=12 236 [47·3%]; White European, 16 445 [63·6%]; Black African or Caribbean, south Asian, mixed, and east Asian or other n=9405 [36·3%]; and mean age 26·7 years [SD=5·4]). For primary care, we included 3989 individuals for development and 4347 individuals for external validation of metabolic syndrome outcomes; and 9181 individuals for development and 7487 individuals for external validation of type 2 diabetes outcomes. For secondary care, we included 846 individuals for development and internal validation of weight gain outcomes. For metabolic syndrome, the performance of PsyMetRiC2-MetS at external validation was C=0·81 (95% CI 0·77-0·84) for the full model (with biochemical predictors) and C=0·79 (0·76-0·83) for the partial model (without biochemical predictors). For type 2 diabetes, discriminative performance at internal validation of PsyMetRiC2-T2D was C=0·86 (0·76-0·95) for the full model, and at external validation it was C=0·81 (0·71-0·88). For weight gain, discriminative performance at internal validation of PsyMetRiC2-WG was C=0·78 (0·73-0·82) for the full model and C=0·77 (0·72-0·80) for the partial model. Calibration plots were acceptable for all models. All models displayed evidence of clinical usefulness at all plausible thresholds. The PsyMetRiC web application is available at https://psymetric.app. Interpretation We developed prediction models for incident cardiometabolic disorders in young people with psychosis. The PsyMetRiC models are among the first in psychiatry to be available for routine clinical use. PsyMetRiC can support a shift toward collaborative, preventive physical health care for young people with psychosis. Funding National Institute for Health and Care Research.