On Call Central uses a 3rd party automated transcription engine to produce an approximate transcript of patient messages; this service is provided at no additional cost to customers, and officially is listed as a beta feature. This particular transcription engine was selected because, of the engines tested, it produced the highest accuracy when challenged with a standardized set of audio files. Automated transcription of audio is imperfect, and will vary greatly in quality based on factors including:
- Audio quality
- Clarity and pace of speech
- Accents & regional vocabularies
In general, expect most transcripts to help you quickly get an idea as to the general nature of the message. In other instances, the transcript may be insufficient and you may have to use the audio. This simply is the state of computer science in the field of speech recognition at this time. We will continue to review additional compatible transcription engines as they become commercially available, and may at some point offer human reviewed transcription as an add-on service.