Background: Clinicians spend nearly half of their time at work on clinical documentation, impacting time spent with patients and trainees. Studies have demonstrated a correlation with electronic health record (EHR) documentation burden and clinician burnout. Finding innovative ways to balance increasing EHR demands and mitigate this burden is essential.
Purpose: We identified that time spent on EHR documentation contributed to work compression and fatigue. Voice recognition (VR) software can decrease time spent on clinical documentation, improve documentation quality, and increase clinician satisfaction. Though VR software has been utilized in many ambulatory practices, there are few studies evaluating its use for hospitalists. Our goal was to assess the adoption of VR technology for clinical documentation and its impact on hospitalist well-being and EHR burden.
Description: We utilized the M*Modal Fluency Direct system for our VR pilot because it was previously validated by our health system’s IT and successfully used across our ambulatory practices. Thirty-four hospitalists across 3 hospitals within our health system participated over the course of this 18-month pilot. In total, VR was used for 4845 notes during the pilot. Roughly half of those clinicians (18) who initially agreed to participate did not use VR at all and 5 were rare users (0.1%-5% of notes). Eight hospitalists used VR in at least 10-30% of their notes, 3 utilized VR over 30% of notes, and the highest utilizer employed VR in 49% of notes. Among regular users of VR (more than 5% of notes), we looked at what types of notes VR was used to write. VR was used most frequently for consult notes (31.3%), progress notes (25.6%) and H&Ps (20.0%). Event notes (0.7%) and discharge summaries (5.3%) were least likely to involve VR. Based on focus groups, we identified barriers to VR utilization included time to learn the system, dictation in shared workspaces, and availability of the technology only in designated workstations.While turnover of participants during the pilot confounded our assessment, a qualitative survey completed by 16 participants at the pilot’s conclusion found that 88% moderately or strongly agreed that VR improved productivity, enhanced note writing efficiency, and made their job easier. Most (81%) agreed that VR reduced the amount of documentation after work hours. Users also found that VR increased note comprehensiveness. Approximately 30% of respondents reported prior experience with VR technology. Not surprisingly, those with prior experience were more likely to perceive enhanced note-writing efficiency using VR. We measured feelings of burnout using the abbreviated Maslach inventory in October 2022 and again three months later. In the category of personal accomplishment, average scores were 13.0 (moderate burnout) in the pre-intervention group and 16.7 (low burnout) in the post-intervention group, indicating a significant improvement (p = 0.02).
Conclusions: Voice Recognition technology use in the inpatient setting is feasible and has potential to reduce clerical burden. Our pilot determined that VR can be transformative in the domains of documentation ease, efficiency and comprehensiveness for hospitalists who are able to successfully integrate the technology into their workflow and for hospitalist groups who are able to establish a work environment that can facilitate the widespread use of dictation software. Use of the technology by regular adopters may also improve feelings of burnout in the domain of personal accomplishment.
