Study: Physician Task Switching Increases with Commercial EHRs
Switching from a homegrown electronic health record (EHR) to a commercial one increased physician task switching in the emergency department, which can have serious patient safety implications, according to recent research published in the Annals of Emergency Medicine.
The single-organization study—conducted by Raj Ratwani, Ph.D., scientific director at the National Center for Human Factors in Healthcare at the Washington D.C.-based MedStar Health, and others—used direct observations of physician activities before, during, and three to four months after the transition from a homegrown electronic medical record to a widely used commercial electronic medical record with provider order entry. The study was completed in an urban, tertiary care, academic ED with 90,000 annual patient visits. Participants included second- and third-year emergency medicine resident physicians and attending physicians.
During the commercial EHR go-live period, time engaged in computer tasks increased, whereas that for other tasks decreased, but this returned to baseline in the post-implementation period. However, the number of tasks engaged in per minute (a measure of task switching) increased significantly from pre- to post-implementation (the equivalent of 91 additional tasks per 8-hour shift).
Solely analyzing the allocation of physician time on various tasks might lead to the conclusion that transitioning from a homegrown electronic health record to a commercial one does not significantly affect emergency physician work activities in the long term. However, the analysis examining the number of tasks physicians engaged in minute by minute provides critical insight into how work activities change as a result of a commercial electronic health record, the researchers concluded.
The researchers said, “The increase in task-switching observed post-implementation imposes a cognitive burden on the physician and may provide quantifiable support for the perceived increase in stress and workload that many physicians have described in other qualitative studies. The increase in task switching was present three to four months after use of the electronic health record began, well beyond the period of its being novel technology for physicians. In addition to increased stress and frustration, rapid task switching can have serious patient safety implications. It results in workflow fragmentation and is well recognized in the human factors literature as increasing the likelihood of errors. Therefore, the connection found between commercial electronic health record use and increased task switching may indicate an increased potential for patient safety hazards.”
What’s more, they said, design shortcomings of the electronic health record system may be one factor that contributed to the increased task switching. For example, when writing an order for patients to receive their home medications in the hospital, physicians cannot view the home medication list while writing new electronic orders. As a result, they need to either remember the home medications or take notes on paper and then switch back and forth between the computerized physician order entry system and paper. This workflow forces physicians to perform multiple tasks in service of the specific goal they are attempting to accomplish. Given that many electronic health records are used to complete critical tasks, such as ordering medications and diagnostic tests, any errors during these tasks can have severe consequences, the researchers noted.
They concluded, “The results of this study highlight important differences between homegrown electronic health records and commercial ones. Homegrown electronic health records, like the one replaced in the ED that we studied, tend to be customized to the workflow processes of the specific provider environment. This customization occurs over time as clinicians’ needs are identified and integrated with the electronic health record. Commercial electronic health records are designed to fit the needs of a large base of users from multiple provider sites and therefore are not customized. The increased task switching observed during our study may be a result of a lack of customization.”
