AF-ALERT: More Oral Anticoagulation, Fewer Clinical Events Using On-Screen Prompts

Marlene Busko

November 19, 2018

CHICAGO — Patients hospitalized with atrial fibrillation (AF) who had not been on an oral anticoagulant for stroke risk reduction were significantly more likely to receive one if their physicians received an on-screen electronic prompt to write the prescription, a single-center study found.

The on-screen alert "nearly tripled prescription of anticoagulation in hospitalized AF patients at increased risk for stroke," said Gregory Piazza, MD, Brigham and Women's Hospital, Boston.

But the study's "most dramatic finding" associated with prompts from the computerized decision support (CDS) system was a significant reduction in 90-day risk for stroke or myocardial infarction (MI), which were secondary outcomes, Piazza observed.

Those patients were also significantly less likely to have a major adverse cardiovascular event (MACE) — a composite of death, MI, stroke, transient ischemic attack (TIA), systemic embolism — at 90 days, he noted. The benefit was driven by reduced rates of stroke and MI.

The study randomized 458 patients cared for by clinicians who either received or did not receive the electronic reminders (based on an Epic electronic health record system) to put them on oral anticoagulation.

It suggests that the CDS intervention "has the potential to be a powerful tool in prevention of cardiovascular events in patients with AF," Piazza said.

He presented the results of the Alert-Based Computerized Decision Support to Increase Anticoagulation Prescription Prevents Stroke and Myocardial Infarction in High-Risk Hospitalized Patients With Atrial Fibrillation (AF-ALERT) trial here at the American Heart Association Scientific Sessions 2018.

"That's a pretty scalable intervention across an Epic-based platform that many health systems in the United States use," invited discussant Mark D. Huffman, MD, MPH, told | Medscape Cardiology.

"I would be interested in the dissemination of these results in other health systems," said Huffman, from Northwestern University Feinberg School of Medicine, Chicago.

"You clearly moved the needle on anticoagulant prescriptions and struck gold there, which is really great," said session comoderator Mintu Turakhia, MD, Palo Alto Health Care System and Stanford University, California. But, he asked, why weren't the patients receiving an anticoagulant in the first place?

The study was restricted to inpatients with a CHA2DS2-VASc score of at least 1 who did not have an order entry for an anticoagulant, Piazza explained.

"Despite a multitude of evidence-based clinical practice guidelines and five effective oral anticoagulants for stroke prevention," said Piazza, "at least 30% of AF patients remain unprotected."

Indeed, a previous observational analysis from the researchers showed that of 5000 hospitalized AF patients, 40% of high-risk patients were not on antithrombotic therapy, which was associated with a 36% increase in risk for MACE.

This "highlights a critical gap in implementation of best clinical practice among providers," said Piazza.

The group had previously shown an improvement in putting hospitalized patients at risk for venous thromboembolism on anticoagulant prophylaxis with the help of a CDS. So they hypothesized a similar alert would improve the rate of anticoagulant prescription and outcomes in hospitalized patients with AF, he said.

The current study involved 458 adults hospitalized for AF at their center with a CHA2DS2-VASc score of at least 1 who had not been prescribed anticoagulant therapy for stroke prevention. Average age of the study participants was 75 years, 55% were men, median CHA2DS2-VASc score was 4, and median HAS-BLED score was 3.

The patients were randomly assigned to a group for which the attending physicians received an alert designed to increase anticoagulant prescription (248 patients) or to a control group without such alerts (210 patients).

The alert displayed the patients' CHA2DS2-VASc scores and annual stroke risk and provided three options:

  • Click on accept to open the order set and select an anticoagulation order

  • Click on a link to review evidence-based clinical practice guidelines

  • Provide a reason for not proceeding with an anticoagulation order, by clicking on one of the following choices:

    • bleeding risk is too high

    • stroke risk is not high

    • patient is at high risk for falls

    • patient refuses anticoagulation, or

    • other (leave comment)

Of the clinicians in the alert group, 35.4% ordered an oral anticoagulant, 0.8% (two clinicians) chose to read the AF guidelines, and the remaining 63.7% provided a reason for not prescribing an anticoagulant. The latter clinicians were mainly concerned about bleeding risk (50%) and falls (12%).

In the CDS alert group, 19.4% of patients were prescribed an anticoagulant during hospitalization, at discharge, or in the subsequent 90 days, compared with 7.1% of control patients (< .001).

The 90-day rate of MACE was 11.3% in the CDS alert group and 21.9% in the control group (= .002).

Similar numbers of patients in the two groups died (10.1% vs 14.8%) or had a major or clinically relevant nonmajor bleed (4.4% vs 7.6%) in the 90 days after discharge.

However, notably fewer patients in the CDS alert group had an MI (1.2% vs 8.6%; < .0001) or stroke, TIA, or systemic embolic event (0% vs 2.4%; P=.02) in the 90 days after discharge.

Expressed differently, at 90 days, patients in the physician-alert group had half the risk for MACE (hazard ratio [HR], 0.50; 95% CI, 0.31 - 0.81; = .004) and about half the risk for MACE combined with a major or clinically relevant nonmajor bleed (HR, 0.48; 95% CI, 0.31 - 0.73; = 0.001).

The electronic alert seemed to preferentially increase the prescription of direct oral anticoagulants over warfarin, Piazza observed.

AF-ALERT was supported by Daiichi-Sankyo. Piazza discloses receiving research grants from EKOS/BTG, Janssen, Bristol-Myers Squibb, and Daiichi-Sankyo; and serving as a consultant or on an advisory board for Portola and Bayer. Huffman discloses receiving research support from Boehringer Ingelheim, Novartis, the American Heart Association, Verily, and AstraZeneca. Turakhia discloses receiving research grants from Janssen, Medtronic, AstraZeneca, Veterans Health Administration, Boehringer Ingelheim, and Cardiva Medical; receiving honoraria from Medtronic, St. Jude Medical, Boehringer Ingelheim, Precision Health Economics, iBeat, Akebia, Cardiva Medical, Medscape, and Amazon; and holding ownership interest in AliveCor, Zipline Medical, iBeat, and iRhythm.

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

For more from, follow us on Twitter and Facebook


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.