Skip top navigation

Himmelfarb Headlines - November / December 2010

Voice Recognition Software for Electronic Medical Records

Voice recognition (VR) software converts spoken words to text to allow hands-free information input.  Historically, it has been used for speeding up dictation and as a resource for disabled user accessibility. This software is being applied for recording information in electronic medical records (EMRs) in medical facilities including Beth Israel Deaconess Medical Center (BIDMC) in Boston, University of California San Diego Medical Center (UCSD) and Slocum-Dickson Medical Group (SDMG) in New Hartford, NY. With VR, "back-end" dictation of clinical documentation is replaced by immediate "front-end" entry at the conclusion of the patient encounter. Details of the interaction and patient are fresher at this time, which means EMR entries may be more complete.

Leaders at institutions using voice recognition for medical record entry all stressed a need for investment quality training of physicians to ensure proper use and support. If physicians are not properly trained and invested in the process, the software is underutilized and return on investment is not maximized. Convenience of physicians' voice profiles has been noted as another challenge of implementing VR in a usable manner. Site-wide VR is more expensive than individual licenses for the software, but allows physicians to store their voice profile on a server, allowing them to use any computer at the facility to record patient information. This is important for hospitals and offices where physicians are not tied to single computers.

Considerable cost savings were found at those facilities due to VR implementation. The largest facility, BIDMC, saves more than $5 million annually using VR software. The smaller SDMG saves about $750,000 annually with the elimination of back-end transcription. The increased productivity allowed through VR transcription also allows SDMG physicians to see an increased number of patients per day. For more information on voice recognition used for electronic medical record entry, see the articles listed below.


Lawrence, D. (2009). Can you hear me now? Voice recognition for the EMR has made big strides, and many say meaningful use requirements will accelerate adoption. Healthcare Informatics, 26(12), 14. Retrieved from

Speech recognition improves EMR ROI. Medical group decided that to be most successful with EMR adoption, 100 percent physician population utilization would be necessary.(2009). Health Management Technology, 30(10), 30-31. Retrieved from

Himmelfarb Health Sciences Library | The George Washington University
2300 Eye St., NW, Washington, DC 20037 | Phone: (202) 994-2850 | Fax: (202) 994-4343