Evaluating a Doctor-in-the-Loop AI Agent for Extracting Lines of Therapy, Stem Cell Transplantation, and Immunotherapies from Real-World EHR Data in Multiple Myeloma

Key Points

• AI effectively identifies lines of therapy and treatments in electronic health records, enhancing clinical decision support.
• The study reports an accuracy rate of 85% for the AI in therapy line extraction across various data sets.
• An observational analysis of electronic health records from multiple academic centers assessed AI performance.
• This approach highlights the potential for AI to streamline data extraction, but further validation in diverse settings is needed.