Two new research papers analyze how clinicians modify AI-generated clinical notes, focusing on changes in hedging and stigmatizing language. The studies found that clinicians often introduce more hedging terms, leading to greater uncertainty in the final notes, and can also be a net source of stigmatizing language entering electronic health records. These effects varied significantly based on the AI vendor and the clinical specialty. AI
IMPACT Clinician editing of AI-generated notes may inadvertently increase uncertainty and bias in patient records.
RANK_REASON Two academic papers published on arXiv analyzing AI-generated clinical notes.
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →