Using Natural Language Understanding to Manage Incidental Radiologic Findings in an EHR
When unexpected findings are discovered in medical imaging tests, care follow-up is critical to successful patient outcomes. After Owensboro Health experienced several patient safety events, an investigation revealed gaps in care coordination. Each event had begun with an incidental finding documented in the radiology report but missed by referring physicians. Without a mechanism for tracking these findings, patients were not informed about key health issues, potentially risking increased morbidity and mortality. This case study of a patient safety initiative at Owensboro Health describes an innovative, coordinated approach that uses natural language understanding (NLU) technology fully integrated with the organization’s EHR to alert, track, audit and report incidental findings. The automated workflow has resulted in a nine-fold increase in identification and communication of incidental findings for improved patient safety.
