EHR Alert Fatigue: An Anticoagulation Example

Most of us practice in an environment that relies on an electronic health record (EHR). The crown jewel of the medical records-the Inbox-is inundated with dozens of messages, including alerts, reminder, messages, copied charts, and lab results. My EHR is Epic, and it does everything from reminding me of “overdue results” when a patient doesn’t get a test I ordered to showing me an icon when a patient I am seeing in the office has a recent or upcoming birthday. 

So, what would you say if I said that I was going to one more alert to your already full-of-alarms-and-popups medical records? Do you think it would make you more or less efficient?

Time to Rx OAC?

An interesting absract presented at the AHA Scientific Sessions 2018 studied the utilization of oral anticoagulation (OAC) prescriptions for hospitalized patients among 458 providers; the clinicians were randomized in an Epic-based platform to either receive a popup reminder to put their patients on OAC or not to receive the popup. Attending physicians caring for patients in the trial who had a CHA₂DS₂-Vasc score of at least 1 received an alert on the EHR that presented the patient’s risk score and annual stroke risk and included a decision making tool that prompted the physician to do one of the following:

• Order OAC through an order set

• Review the guidelines on use of oral anticoagulation, or

• Explain why anticoagulation was not being used

Only 2 physicians actually reviewed the guidelines and 64% provided a good contraindication for not choosing OAC.

The trial, aptly named ALERT-AF, reported that 19.4% of patients with physicians in the alert group were prescribed an anticoagulant during, at discharge from, or after the hospitalization, compared to 7.1% of controls (P<.001) and this translated into a ~50% lower rate for MACE (11.3% vs. 21.9%, p=0.002) up to 90 days following discharge.

Pathognomonic for alert fatigue

Although these results are somewhat compelling as they reflect a significant increase in the uptake of OAC prescriptions and improvement in MACE for hospitalized patients, the major limitation of the study is its short time-line. I remember when the sepsis alert first started to appear in Epic for any patient who could be septic and benefit from guideline-recommended goal-directed therapy. For the first few months, it dramatically influenced clinician behavior. After about 6-12 months, it became yet another popup that prompted most clinicians to select the “other” option to escape it and there was significant attrition in the benefit of the popup-pathognomonic for alert fatigue.

What do you think?

Would this type of popup improve or hinder your efficiency and likelihood to prescribe OAC to a hospitalized patient? Data 1 or 2 years after implementing this intervention may help to answer this question and hopefully will be reported at the next AHA…

Source: American Heart Association (AHA) Scientific Sessions 2018: Abstract 17685. Presented November 10, 2018