Providers using an EMR system may encounter more than 100 alerts every day, and the volume of alerts continues to grow over time. Conventional wisdom suggests that these alerts are part of the reason physicians experience burnout in steadily increasing levels.
In theory, it’s possible that physicians would find the alerts to be less distracting and unpleasant if they were more relevant and customized to their practice. However, EMRs generally don’t offer such customization options.
Despite these concerns, the problem of EMR alert fatigue is little studied. Not only that, research on this topic is undermined by the fact that no clear definition or agreed-upon metrics exist for measuring the alert-fatigue problem.
However, what research does exist on this subject suggests that the barrage of alerts is a problem. One retrospective cohort study showed that the likelihood of a provider’s acceptance of the alert fell by 30% for each additional reminder received per encounter.
Meanwhile, one study on public-health messaging suggested a consequence of alert fatigue can be significant. This study concluded that an increase of one local public-health message per week resulted in a 41.2% decrease in the likelihood of providers recalling the content of the message.
An argument can be made that EMRs should not provide clinical alerts until vendors agree on a universal set of standards for alert formats, and until researchers identify best practices for deploying alerts safely and efficiently.
No matter how intelligent and focused a physician is, there will come a time when they make a mistake that stands to cause a patient harm. Ideally, clinical alerts would appear only under limited circumstances, and the content would be laser-focused on the needs of the individual provider.
Instead, as the article outlines, the number of alerts a physician encounters during the course of their work is mushrooming, and as the volume of alerts goes up, their value goes down. Regardless of how we got here, we’ve reached a crisis point.
Not only that, but the sad truth is that the EMR your physicians use in the clinic may generate a different set of alerts at different times and under different circumstances. This makes it even more difficult for physicians to sort out useful alerts from relevant ones.
Fortunately, it may soon be possible to customize alerts for individual physicians, as well as providing specialized alerts for a given patient seen by that doctor.
Artificial-intelligence technologies will someday soon be able to draw on multiple sources of data—including EMR data, test results, patient diagnoses, and more—to make alerts far more relevant. Artificial intelligence will also make it possible to deliver highly targeted local research details to support physicians in the process of care. Such technologies are not mature yet, but they’re on the way.
In the meantime, as the author suggests, it may be time to admit that the current system is not working and that, at a minimum, healthcare organizations should scale back on the number of alerts they deliver, if possible. After all, there’s little value to delivering reminders if it is inevitable that they’ll be ignored.
Article appeared in the Physician Leadership Journal, November–December 2018