Typed versus Voice Recognition for Data Entry in an Electronic Health Record: Emergency Department Physician Time Utilization and Interruptions

Abstract

IntroductionUse of electronic health record (EHR) systems can place a considerable data entry burden upon the emergency department (ED) physician. Voice recognition data entry has been proposed as one mechanism to mitigate some of this burden; however, no reports are available specifically comparing emergency physician (EP) time use or number of interruptions between typed and voice recognition data entry-based EHRs. We designed this study to compare physician time use and interruptions between an EHR system using typed data entry versus an EHR with voice recognition.MethodsWe collected prospective observational data at 2 academic teaching hospital EDs, one using an EHR with typed data entry and the other with voice recognition capabilities. Independent raters observed EP activities during regular shifts. Tasks each physician performed were noted and logged in 30 second intervals. We compared time allocated to charting, direct patient care, and change in tasks leading to interruptions between sites.ResultsWe logged 4,140 minutes of observation for this study. We detected no statistically significant differences in the time spent by EPs charting (29.4% typed; 27.5% voice) or the time allocated to direct patient care (30.7%; 30.8%). Significantly more interruptions per hour were seen with typed data entry versus voice recognition data entry (5.33 vs. 3.47; p=0.0165).ConclusionThe use of a voice recognition data entry system versus typed data entry did not appear to alter the amount of time physicians spend charting or performing direct patient care in an ED setting. However, we did observe a lower number of workflow interruptions with the voice recognition data entry EHR. Additional research is needed to further evaluate the data entry burden in the ED and examine alternative mechanisms for chart entry as EHR systems continue to evolve.

Rapid Electronic Provider Documentation Design and Implementation in an Academic Pediatric Emergency Department

“…One of the main complaints about EPD adoption is the increase in charting time. 20 Providers may be spending more time after clinical shifts to chart; this was not accurately captured in our metrics. We had attempted to compare subjective reported provider charting times after shifts, but the data were incomplete and could not be adequately analyzed.…”

“…The act of dictation may prevent provider interruptions in the ED and may not necessarily increase charting time. 20 Voice recognition profiles do require additional training and curation to improve accuracy and are cheaper than transcriptionist but more expensive than typing. 21 The advantage of the voice recognition program spelling correctly is offset by the need to proofread the dictated text, but accuracy has been shown to nearly as good as human transcription.…”

Emergency Medicine Residents on Electronic Medical Records: Perspectives and Advice

“…In a direct comparison of voice recognition software and transcription services, one group found voice recognition to be superior given the shorter turnaround time and decreased transcription costs [22]. Another study found a decreased number of physician interruptions when voice recognition software was used in place of traditional typing-based charting [23].…”

Physician experience with speech recognition software in psychiatry: usage and perspective

“…Cost-effectiveness of SRS over traditional transcription has also been noted [ 9 ]. Fewer interruptions of emergency room physicians occurred with SRS when compared to written data entry [ 10 ].…”