Economic evaluation of voice recognition for the clinicians' desktop at the Naval Hospital Roosevelt Roads

• HIT Description: Voice recognition technology as an input device to the electronic medical record. More info...
  
• Purpose of Study: To determine the viability and economic benefits of using a commercial off-the-shelf voice recognition application as a clinician input device for transcribing clinical encounter notes.
  
• Years of study: August 1996-April 1997.
  
• Study Design: Survey of users, standardized scenario analysis, and preliminary cost benefit analysis.
  
• Outcomes: Perceptions of the use and benefit of voice recognition technology, effects on patient-clinician relationships, and potential cost of widespread implementation.
  

• Settings: The family practice clinic at the family medical center affiliated with the Naval Hospital Roosevelt Roads in Puerto Rico.
• Intervention: Pilot study of the use of DragonDictate as a voice recognition system.
• Evaluation Method: Surveys, standardized scenarios and preliminary cost benefit analysis comparing implementing a voice recognition system to hiring additional transcriptionists.
• Description: The use of DragonDictate voice recognition software for use as an input device into an existing electronic medical record. No other details of the electronic medical record are provided
• Interoperability: DragonDictate allowed direct dictation into the EMR.
• Financial Context: There was no money to hire additional medical transcriptionists and the decision to assess this HIT was driven by an attempt to find a cheaper method for managing clinical input.
• Facilitators: A series of interviews were conducted about how the implementation strategy worked and selected statements are selected. Most statements were positive but an important feedback was the need to have one or two dedicated people from the information system department to support the voice recognition department on a daily basis in order to ensure its success.
• Costs: The article estimated that the first year cost per clinician for implementing the voice recognition input including purchase of the software, additional memory, and soundboard was $1,400. Maintenance costs were not included in the first year start-up cost and were judged to be ÒmodestÓ in later years. The alternative, expanding transcription services, would cost at least $6000 per clinician as startup cost which includes hiring transcriptionist and installation of Dictaphone equipment for each clinician. Cost of transcription services for subsequent years was not estimated.
• Clinical Outcomes: The use of the voice recognition technology during clinical encounters has negligible effect on patient-clinician interactions
• Changes in efficiency and productivity: Compared to transcription service, the voice recognition system is perceived to increase productivity in terms of more thorough clinical encounter notes at comparable speed, elimination of the need to hire medical transcriptionists, and reduction of graphic user interface overload on the Military Health Service Systems.�