Fighting Alert Fatigue to Improve Patient Safety and Standardization of Care

As Associate Chief Medical Information Officers (ACMIOs) at Cleveland Clinic, Allison Weathers, MD and Eric Boose, MD focus on tailoring solutions within the Epic electronic health system (EHR). Their work helps clinicians deliver the best patient care possible. A key part of their job is leading the Cleveland Clinic Clinical Decision Support (CDS) Governance Workgroup, a multidisciplinary group of clinicians that regularly reviews EHR tools designed to enhance decision-making in clinical workflows.

CQD sat down with Drs. Weathers and Boose to get an inside look at the CDS Governance Workgroup at Cleveland Clinic and how these kinds of bodies are crucial to driving safe and standardized patient care.

What is CDS?

A large majority of healthcare professionals, especially clinicians, immediately associate CDS with interruptive interface alerts in the EHR — tools that act as fail safes, such as Epic BestPractice Advisories. But there’s lots of ways to implement decision support outside of alerting. For example, including questions within an order can guide the clinician towards taking the right steps. It can also be the way a collection of orders are grouped for a provider. The fields a physician is prompted to complete within a note template or the order of options in a list of selections are a couple more examples of CDS. Knowing that people’s first impulses are to pick from the top of a list, we can organize a list that makes sense and corresponds with what we want them to do clinically.

Why was the CDS Governance Workgroup formed?

Before we started the workgroup, on average, clinicians across our enterprise saw over three million BestPractice Advisory alerts in Epic per month. Over the years, we have become much more aware of the dangers and risks of alert fatigue. One analogy is alarms going off in an inpatient unit. Machines are beeping all the time, so people learn to tune them out. As a result, they sometimes end up missing important alarms. It’s the same for our EHR. If there’s too much alerting, or if it’s not done in the right way, we can end up ignoring truly significant items and patient safety risks.

To address this, we brought together a multidisciplinary and multispecialty group—including physicians, advanced practice providers (APPs), residents, trainees, nurses, pharmacists and researchers. We wanted broad representation to make sure we’re thinking very critically about every alert before we put it in place.

What are the goals of the group?

Every time we meet or think about CDS, we always focus on our driving philosophy of “the five rights.” Our mission is to ensure the right information goes to the right person at the right time, in the right place, format and channel. And we apply that philosophy to three “buckets”:

1. Review requests to build new alerts.

One of the first things we did was overhaul and standardize our intake process. The objective of this committee was to challenge requestors from the start to be thoughtful with what they’re trying to solve. We wanted to understand if there was a larger safety or security event behind the request, the scope of the impact and how they were going to measure the success of the functionality. From there, we would examine if the rationale made sense. We also checked if there were alternate, less-intrusive CDS tools to consider, that might be a better fit for the workflow. As a result of this new process, we were able to ensure the consistency of alerts across the enterprise to help guarantee that patients receive the same expert level of care, regardless of what location and specialty they visited.

For example, we received a request from a clinical institute to build an alert to remind providers when patients with certain chronic conditions were overdue for testing. After review, we found the issue to be more of a care gap, and a checklist to address existing patient care gaps already existed within our outpatient EHR workflows. By adding the test to the checklist, we avoided building an interruptive alert and made it easy for the provider to place needed orders during the office visit.

We’ve also been able to create better alerting to raise awareness around certain diseases. For example, concussion guidance and instructions change frequently. Because a consistent approach on patient instructions is critical for recovery, the committee approved an alert to streamline and remind providers to seamlessly add standardized concussion guidance for patients after their visit.

2. Review legacy alerts for relevance and efficacy, then deactivate or update accordingly.

Another important role of the committee is reviewing the alerts in our present ecosystem and challenging ourselves to clean them up. We’ve gone back and reviewed alerts that have been in circulation for a while to make sure they were accomplishing what they were intended. If they weren’t, we turned them off. We also partnered with our EHR software vendor to analyze our highest-use alerts for opportunities to redesign, optimize or include new functionality.

3. Review recently approved alerts and make sure they’re accomplishing their stated intention.

For the past couple years, we’ve focused primarily on building our intake process and cleaning up legacy alerts. Now we’re in a place where we can circle back on recently approved alerts to conduct a timelier review process. By having requestors come back and speak to their outcome data before the committee, we position them to make an intentional decision on whether to keep, deactivate or further optimize the alert based on the outcomes. This way, if an alert doesn’t meet its initial intentions, it doesn’t linger in the system for ten more years, disrupting workflows and cluttering our EHR

What is the feedback from clinicians in their day-to-day practice?

It’s been well-received—people are passionate about not having alerts that are nuisances. They want ones that are intentionally designed. For the first time, we’re truly getting down to the grassroots level of having people involved at all levels of our alerting. That’s a huge benefit of having a multidisciplinary committee.

We’re both involved in a wide number of similar committees and physician advisory groups, but for some reason, this one stands out. Everyone involved in the committee is excited and extremely engaged.