Predicting Diagnostic Progression to Schizophrenia or Bipolar Disorder via Machine Learning

Abstract

Importance: The diagnosis of schizophrenia and bipolar disorder is often delayed several years despite illness typically emerging in late adolescence or early adulthood, which impedes initiation of targeted treatment.

Objective: To investigate whether machine learning models trained on routine clinical data from electronic health records (EHRs) can predict diagnostic progression to schizophrenia or bipolar disorder among patients undergoing treatment in psychiatric services for other mental illness.

Design, setting, and participants: This cohort study was based on data from EHRs from the Psychiatric Services of the Central Denmark Region. All patients aged 15 to 60 years with at least 2 contacts (at least 3 months apart) with the Psychiatric Services of the Central Denmark Region between January 1, 2013, and November 21, 2016, were included. Analysis occurred from December 2022 to November 2024.

Exposures: Predictors based on EHR data, including medications, diagnoses, and clinical notes.

Main outcomes and measures: Diagnostic transition to schizophrenia or bipolar disorder within 5 years, predicted 1 day before outpatient contacts by means of elastic net regularized logistic regression and extreme gradient boosting (XGBoost) models. The area under the receiver operating characteristic curve (AUROC) was used to determine the best performing model.

Results: The study included 24 449 patients (median [Q1-Q3] age at time of prediction, 32.2 [24.2-42.5] years; 13 843 female [56.6%]) and 398 922 outpatient contacts. Transition to the first occurrence of either schizophrenia or bipolar disorder was predicted by the XGBoost model, with an AUROC of 0.70 (95% CI, 0.70-0.70) on the training set and 0.64 (95% CI, 0.63-0.65) on the test set, which consisted of 2 held-out hospital sites. At a predicted positive rate of 4%, the XGBoost model had a sensitivity of 9.3%, a specificity of 96.3%, and a positive predictive value (PPV) of 13.0%. Predicting schizophrenia separately yielded better performance (AUROC, 0.80; 95% CI, 0.79-0.81; sensitivity, 19.4%; specificity, 96.3%; PPV, 10.8%) than was the case for bipolar disorder (AUROC, 0.62, 95% CI, 0.61-0.63; sensitivity, 9.9%; specificity, 96.2%; PPV, 8.4%). Clinical notes proved particularly informative for prediction.

Conclusions and relevance: These findings suggest that it is possible to predict diagnostic transition to schizophrenia and bipolar disorder from routine clinical data extracted from EHRs, with schizophrenia being notably easier to predict than bipolar disorder.