Editor,Voice recognition software (VRS) has increasingly been utilised to document clinical care, typically within an electronic health care record and often with the use of ‘templates’. This software has been purported to enhance doctor efficiency, reduce costs and improve patient care. The aim of the current pilot was the mandatory adoption of VRS embedded into electronic clinical documentation within a new patient endocrine clinic.
METHODS
Dragon Medical Practice Edition 2 speech recognition software manufactured by Nuance was installed onto a single office computer; a run in period of two months was required to optimize user dictation. Prior to clinic attendance, each patient had a voice activated clinic template note constructed online within Patient Center. The online medical note was then re-opened and typed in real time during the patient consultation. Once constructed, the outpatient note was reviewed, formatted (by typing and/or VRS) then authorised with an electronic signature.
RESULTS
Data from 24 consecutive medical notes were collected before and after the implementation of VRS. The use of VRS resulted in all of the outpatient medical notes transferring to an electronic/online version. The setup time for VRS was one minute per clinic letter, the existing process did not require any set up time. The total time allocation per clinic visit was similar (n=25 minutes) per patient for both processes (included obtaining a history, examination, medical note documentation and discussion with the patient). VRS improved the number of clinic letters appearing on NIECR on the day of clinic attendance (24 v 2, p=0.01) in comparison to the existing process. There was an improved mean turnaround time with VRS from day of clinic to the completion of clinic letter (7 v 25 days, p=0.01) appearing on NIECR in comparison to the existing process. Total clinician online medical note typing time was 7 minutes per patient in comparison to the existing process which did not require any time for clinician typing. The mean dictation time for the existing process per clinic letter was 1.5 minutes in comparison to 3 minutes using VRS. One new patient clinic (n=6 patients) resulted in savings in secretarial transcription time on average 30 minutes per clinic session.
DISCUSSION
Current upgraded versions of VRS have allowed the transcription of speech into written text with speed and accuracy. The use of VRS enabled the process of construction of the electronic outpatient clinical note into a single step and resulted in ‘same day letters’, improved turnaround time and subsequent accessibility of clinic letters. The online letters could be accessed remotely and out of typical working hours if required. Advantages of the use of VRS include reducing errors in dictation and in illegible handwritten notes. Disadvantages commonly encountered are lack of accuracy and misinterpretation. The use of the software can be time consuming initially and prone to errors with background noise. VRS has the potential for additional roll out in other outpatient settings and in streamlining and easing the burden of the written outpatient clinic note.