A Machine Learning Model Identifies Patients In Need Of Autoimmune Disease Testing Using Electronic Health Records

This study explored the potential of a machine learning model to identify patients with a clinical profile characteristic of receiving rheumatological evaluation for suspected systemic autoimmune rheumatic diseases (SARDs). The model was trained and tested using electronic health records (EHRs) from 161,584 individuals across two institutions, and targeted a core group of SARDs. The model was found to identify more individuals with autoantibodies and autoimmune disease diagnoses than current clinical standards, and accurately predicted the need for autoantibody and rheumatologist testing up to 5 years before the actual testing date.