Free-Text Data Entry by Speech Recognition Software and its Impact on Clinical Routine

Abstract

We conducted a study to evaluate speech recog nition soft ware in an otorhino laryngo logy unit and to assess its imp act on p roductivity pri or to gene ral impl ementation. Current speech recognition sof tware (IBM ViaVoice, version 10) was impl emented on a personal computer with a 2-GHz central pro cessing unit , 256 MB of RAM, and a 30-GB hard disk drive, with and without add-on professional vocabularyforotorhino laryngo logy. This vocabulary was added by the automated analysis ofan additional 12,25 7 docum ents from our department. We compared the word recog nition error rates f or three different text typ es and determin ed their impact on the amount ofsurgeon : time that was invested in the p roduction ofan error-f ree docum ent. Although error rates without any professional vocabulary dat abase we re rather high (operation report s: 38.72 %; consultation notes: 2 7.77%), the patient inf ormation was edited with a satisf actory result (10.65 %). Best results were obtained with the specialty-related vocab ulary database added by the analysis of our ow n documents (operation reports: 5.45 %; consultation notes: 5.21%). An increase in p roductivity compa red with that of conventional tran scription was fo und at an error rate of less than 16%.

Experience With Voice Recognition in Surgical Pathology at a Large Academic Multi-Institutional Center

“…9 There have also been reports of implementation in other clinical specialties, such as orthopedics, pediatrics, and emergency medicine. 1,4,[10][11][12] The largest benefit that most of these studies have described is a decrease in turnaround time, 3,13-15 resulting in increased satisfaction from clinicians and administrators, as well as less time wasted fielding telephone calls for preliminary reports. Another advantage is that the reports are edited in real time (as opposed to a few days later when There have been several benefits of using this technology.…”

“…Numerous studies, mainly in radiology, [1][2][3][4] have compared the accuracy to human transcription. Despite the initial overhead setup cost, there is potential for considerable savings with transcription services and shorter turnaround time.…”

Enhancing Nursing Practice by Utilizing Voice Recognition for Direct Documentation

“…Developed as a means to better capture healthcare information and facilitate communication, voice recognition (VR) technology has been directed more toward medical staff practitioners. 2,3 Physicians in radiology and pathology have been reported as early adopters of VR software. 2,4 Studies in long-term-care settings have demonstrated benefits for use among nursing staff including decreased documentation burden, improved financial performance due to increased reimbursement, and decreased overtime.…”

Essential Elements of Natural Language Processing: What the Radiologist Should Know

“…For instance, linguistic analysis can be used to improve the quality of speech recognition by restricting the vocabulary of the ASR system. One 2006 study found that a general vocabulary yielded 39% error when used for medical dictation, whereas a specialized dictionary produced 9% error (43). In 2008, researchers found that the exploitation of contextual cues yields significant performance improvements in speech recognition independent of the vocabulary (44).…”