Overview on Alert Fatigue- Dr. Nancy Gagliano & Wayne Thompson
Health Affairs from 2011 sums up the challenge well in their article “Clinical Decision Support Systems Could Be Modified To Reduce ‘Alert Fatigue’ While Still Minimizing The Risk Of Litigation”. As Meaningful Use incentivized the implementation of EHRs, providers faced a dilemma. Turn on all of the alerts provided, or turn them off and face possible legal challenges. Many organizations defaulted to turning off all but the most critical alerts in response to push-back from clinicians. Per the AHRQ, “Alert fatigue is now recognized as a major unintended consequence of the computerization of health care and a significant patient safety hazard.” The CMO, with patient care as a primary focus, and the CIO, charged with providing an efficient and yet compliant system in which to conduct that patient care, often find themselves as unintended combatants in a quest to strike the appropriate balance.
 Kesselbaum, Cresswell, et al. Health Affairs V30 No.12 2011
Besides alerts produced by the electronic health record, there are smart pumps, motion alarms, and vital signs monitors. Each alert is well-intentioned, but the combination leads to an alarming (pun intended) number of warnings, that an individual clinician must absorb.
This is compounded by the fact that many alerts are not clinically significant. The problem is that most of today’s systems lack sufficient ability to tailor alerts. As a result, Alert Fatigue is now widely recognized as a patient safety issue, a physician satisfaction/burnout issue, and an unintended consequence of expedited EHR implementations across the country. The good news is that improving the situation has a few common addressable components, such as reducing “interruptive” alerts, and eliminating clinically inconsequential alerts.
Organizations try to ensure that evidence-based content and best practice are “baked in” to the EHR implementation. Regulatory compliance, quality, patient safety, payer contracts, and population health, all add alerts to the fray. The resulting environment is burdensome, but it is usually the result of the layered collective needs.
The configuration team is often faced with a dizzying array of variables to reconcile, often without clear direction. In addition, the CIO is almost always working with a tool that is less flexible and granular than he or she would like. The complex logic inherent in clinician thinking and stakeholder requirements is often impossible to recreate, leading to compromises which nobody likes.
Lastly, testing often fails to fully engage all stakeholders. The result is that the real impact of alerts is first realized in production. This can lead to a typical reaction from clinicians of “Turn them off until we sort this out!”. If he or she is the perceived owner of the issue, the CIO must react, or turn to clinical leadership for an answer….
An Integrated Approach
An enlightened approach starts with the question of ownership. Who oversees clinical alerts ? Med Exec, the CMO, the CMIO, Patient Safety, the P&T Committee, each practice/service? We often hear 3 common answers to this question:
- Nobody, or I don’t know
- Several conflicting answers
Of course, none of these are desirable answers, so step one is to make ownership. The organization needs to know where to direct issues regarding alerts. IT is usually not the right answer. Step two is to break away from the tendency to have clinicians define what they want, and then have IT respond. Instead, start with the alerts that are most intrusive, or which fire most often. Then have a multi-disciplinary team use this fact-based list to make changes in brief work sessions. Leadership can quickly point to noticeable improvements for clinicians which further empowers the workgroup.
Alerts in and of themselves are neither good nor evil. The magic is in managing them within the boundaries of your technology platform, and your patient care environment.