LLM Judges Verify Presence, Not Absence: Omission Blindness in AI Clinical Notes and What Recovers It

A new study reveals that LLM-based clinical note audits frequently fail to detect missing information, a phenomenon labeled omission blindness. Researchers found that while these judges excel at verifying existing text, they struggle to identify what was said in a patient encounter but left out of the final record. This limitation suggests that current automated quality control methods may provide an overly optimistic view of AI-generated medical documentation, potentially masking significant documentation gaps in clinical workflows.