The decision support tool uses the patient's age, gender and history of stroke or bleeding, along with whether the individuals had vascular disease, a history of myocardial infarction, alcoholism, intracranial hemorrhage, hypertension, congestive heart failure, abnormal liver and other health aliments. The model assigns points based on these characteristics and offers a recommendation along with guidelines on treatment of atrial fibrillation from the American College of Cardiology/American Heart Association/Heart Rhythm Society.

Atrial fibrillation, an irregular often rapid heart rhythm that leads to poor circulation, affects up to 10 percent of the population age 70 and older, explains Eckman. It also presents a strong risk factor for stroke due to clots forming in the left atrium that then travel to the rest of the body including the brain.


The average risk of stroke for patients with atrial fibrillation is just over 4 percent per year. Patients who have strokes because of atrial fibrillation have a higher death rate than patients who have a stroke due to hypertension for instance, says Eckman.

National as well as local data suggest that physicians aren't providing appropriate blood thinning therapy for a large enough portion of patients with atrial fibrillation, says Eckman. Less than 50 percent of patients with atrial fibrillation are getting blood thinning therapies with anticoagulants so there is under treatment when it comes to stroke prevention.

"The reason for that, in part is because the treatment is very dangerous," he says. "These are blood thinning therapies, also known as oral anticoagulants like warfarin, and they have risks themselves of causing major bleeding. The bottom line is there is a balance between risk and benefit. The challenge is to give the treatment to patients who on average would do better with stroke prevention versus bleeding complications. That's the nut of the decision support tool."

Eckman said the computerized tool allows physicians to see what the life expectancy of a patient may be based on whether they received antithrombotic therapy, aspirin or oral anticoagulant therapy as part of their regiment to combat the stroke risk of atrial fibrillation. The model can figure out how high a risk of stroke or bleeding in the brain the patient may face based on the regimens.

Eckman says the results of the trial have convinced physician-researchers to make the computerized model more accessible to physicians by allowing doctors who use the electronic health record, EPIC, on a routine basis to also be able to use the computerized decision support tool. Previously, the tool was available as a separate system and only a small group of physicians used it.

Medical Research Homepage