• E-mail
• Print
• RSS

Consider switching to speech-recognition technology

HIM Connection, October 26, 2004

Want to receive articles like this one in your inbox? Subscribe to HIM Connection!

Transcription is plagued by the same condition as coding-a shortage of qualified personnel. Even with some transcription performed remotely, the shortage continues. Moreover, the demand for dictated documents is increasing. Speech-recognition technology is a solution.

Manufacturers offering this technology readily admit it is not for everyone. It has been most successful in clinical settings with confined or specialized vocabularies, such as radiology and pathology. In these settings, physicians often edit their own reports, known as foreground or front-end speech recognition.

Although doing so trains the software to recognize the nuances of each dictator's speech, some physicians resist editing their dictation. First, they may not have the time to do so or say it slows them down so they cannot treat as many patients. Second, physicians are expensive editors. Teaching the technology to do what it should allow physicians to control costs in their office settings and expedite the creation of legible documents.

To deal with the system-training issue, the use of back-end editing has risen. This means speech-recognition software is loaded into the dictation system but is transparent to the dictating physician. Back-end editing occurs when a transcriptionist or other specialist listens to the dictated word simultaneously corrects the system-generated transcription. By managing speech recognition in this way, the physician incurs no additional effort and the organization may gain from reduced transcription efforts. Back-end editing is a role change for transcriptionists and can be difficult for some.

Seven steps to make the switch to speech-recognition software

1. Evaluate whether your current dictation system will accommodate speech-recognition software.
2. Initially assign the speech-recognition application to a small group of physicians.
3. Prepare the transcriptionists for their new role by allocating a portion of the dictation for each of the physicians in the initial program.
4. If the transcriptionists have been on a productivity plan, initially allocate credits based on transcribed lines to these individuals so they do not lose income while learning to be editors.
5. Once the system's accuracy reaches high 80s/low 90s percentile, advise physicians of its benefits. Collaborate with administration and information technology to determine whether the dictation system can be expanded to the physician's office for his or her use in their office transcription. This could be a chargeable service for HIM.
6. Expand the number of physicians assigned to speech recognition gradually so your transcriptionists' skills grow gradually.
7. Maintain statistics before and throughout the transition to reflect the lines, characters, or words transcribed per hour to determine production increases and return on investment in this software.

Want to receive articles like this one in your inbox? Subscribe to HIM Connection!

• E-mail to a Colleague
• Print
• RSS
• Archive