Impact of Generative Artificial Intelligence on Diagnosing Rheumatoid Arthritis Complications

What this Trial Is About

Researchers are evaluating how generative artificial intelligence (GenAI), based on large language models, may influence the diagnosis of various complications related to rheumatoid arthritis (RA). This study includes both a retrospective examination of electronic health records (EHR) to measure physician adoption of GenAI predictions and a prospective observational phase assessing the feasibility of using GenAI as additional clinical information to improve diagnosis of RA complications and co-existing diseases, including rare or complex conditions. The study involves inpatients newly diagnosed with RA admitted to the rheumatology department between October 1, 2025, and June 2026. Physicians will use GenAI prediction reports generated from patients' EHR data to identify potential RA complications and co-existing diseases. These AI predictions, alongside confirmatory diagnostic tests, will support physicians in the differential diagnosis process. The retrospective phase will validate physician adoption of GenAI predictions from past patient records before moving to the prospective observational phase. Participants will be observed from admission through discharge, during which physicians will review and use GenAI reports to inform their diagnostic decisions. The primary outcome measures include whether physicians adopt GenAI predictions during diagnosis at admission, while secondary measures assess how much the final RA complication diagnoses at discharge are influenced by GenAI. The study aims to better understand the role of AI in improving diagnosis for RA patients over the study period ending in June 2026.

Generative AI Impact on Rheumatoid Arthritis Complications Diagnosis