Peer-reviewed research
Speech Recognition as a Transcription Aid: A Randomized Comparison With Standard Transcription
David N Mohr et al., Journal of the American Medical Informatics Association, 2003 · DOI 10.1197/jamia.M1130
Read the originalRead on October 16, 2026
- What it is
- A randomized comparison in which clinical note dictations from physicians in two specialty divisions, and the secretaries and transcriptionists who processed them, were assigned either to standard transcription or to speech recognition. Productivity was measured as hours of secretary work per minute of dictation processed.
- How much weight it carries
- Peer-reviewed, randomized and published in a major informatics journal. It covers one organization with an existing document-processing system and an older recognizer, and it measures transcription staff, not business staff who draft their own text. It shows that gains are not automatic. It does not measure current software.
Facts taken from this source
- In the endocrinology division of a 2003 randomized comparison, secretaries were 87.3 percent as productive with speech recognition as with standard transcription.
“Secretaries in the endocrinology division were 87.3% (confidence interval, 83.3%, 92.3%) as productive with the speech-recognition technology” (Abstract, Results)
- In a 2003 randomized comparison, Mohr and colleagues concluded that speech recognition did not improve the productivity of secretaries or transcriptionists in the organization they studied.
“speech recognition did not improve the productivity of secretaries or transcriptionists” (Abstract, Conclusion)