Skip to main content
Study finds errors in speech-recognition-assisted documentation

Speech-recognition-assisted notes have an error rate of 7.4%, which dropped to 0.3% after the physician's signed note and to 0.4% after the medical transcriptionist-edited document, according to a study published in JAMA Network Open. Researchers analyzed 217 notes dictated by 144 physicians and found that deletions and insertions were the most common errors, while higher mean error rates were seen in discharge summaries.

Full Story: