byUniversity of North Carolina at Chapel Hill School of Medicine

Credit: CC0 Public Domain

Researchers at the UNC School of Medicine and the University of Vermont have developed a more precise way to assess stroke risk in people with atrial fibrillation (A-Fib), a condition that affects 10.5 million Americans and is a leading cause of stroke.

Their findings, published in two studies in theJournal of Thrombosis and Haemostasis, show that adding blood tests to an existing riskcalculatorcan help physicians better determine who truly needs blood thinners—powerful drugs that prevent strokes but can also cause dangerous bleeding.

"This will help doctors better select patients for anticoagulation, potentially saving lives and reducing health care costs," said Samuel Short, MD, a first-year hematology and oncology fellow at the UNC School of Medicine and lead author of the paper.

Short began this research as amedical studentat the University of Vermont's Larner College of Medicine alongside faculty mentor Mary Cushman, MD, MSc. Together, they used data from an ongoing national cohort study of 30,239 adults monitored for stroke and related diseases termed Reasons for Geographic and Racial Differences in Stroke (REGARDS).

Atrial fibrillation (A-Fib) is the most common arrhythmia that makes the top chambers of the heart quiver instead of pumping normally. The blood inside the heart moves more slowly, and this pooling of blood can lead toblood clots.

"Blood clots can be ejected by the heart to the brain—causing stroke," said Short "Blood thinning medicines, or anticoagulants, make it harder for the blood to clot, loweringstroke risk. However, not every patient can or should be on a blood thinner."

Blood thinners, also known as anticoagulant medications, can reduce stroke risk for some people with A-Fib, but some people who take anticoagulants experience a breakthrough stroke anyway, with the medications potentially causing dangerous side effects related to uncontrolled bleeding.

Stroke risk calculators aren't exactly new. Physicians currently use a conventional risk calculator, called the CHA2DS2-VASc score, which considers a patient's age, sex, andmedical historyto estimate a patient's stroke risk before prescribing anticoagulants.

However, the conventional risk calculator does not account for heart dysfunction, accelerated blood clotting, and inflammation, which are all important risk factors for stroke. According to Short, current calculators need to consider more data beforeblood thinnersare prescribed to high-risk patients.

"Current calculators are not particularly accurate and improving these tools can both decrease strokes and bleeding by offering the right medicines to the right patients," said Short.

The improved CHA2DS2-VASc-Biomarkers risk calculator accounts for these additional risks factors by adding results from the blood tests to the calculation.

In thefirst study, researchers looked at nine blood tests that might predict the occurrence of stroke in 713 people who were taking an anticoagulant to prevent stroke. With a 12-year follow-up, 9%

In the second study, researchers looked at the same 9 blood tests in 2,400 people who were not taking an anticoagulant to prevent stroke. Over 13 years, 7% of them developed a stroke. Short and Cushman discovered that two of the blood tests increased the ability of the CHA2DS2-VASc score to predict astroke.

The researchers used these two blood tests to create a more accurate risk scoring system, called CHA2DS2-VASc-Biomarkers score. This improved scoring system can be used by clinicians to decide which patients withatrial fibrillationmight benefit the most from anticoagulant medication.

Physicians can use the newimproved stroke risk calculator onlinefor easy access. Although the calculator is not yet ready for use in clinics, Short is immensely excited and gratified that their research will someday improve how doctors care for all patients with A-Fib.

More information: Samuel A.P. Short et al, Circulating biomarkers and ischemic stroke risk in adults with atrial fibrillation taking anticoagulation: the Reasons for Geographic And Racial Differences in Stroke study, Journal of Thrombosis and Haemostasis (2025). DOI: 10.1016/j.jtha.2025.05.029 Samuel A.P. Short et al, Improving stroke risk prediction in atrial fibrillation with circulating biomarkers: the CHA2DS2-VASc–Biomarkers model, Journal of Thrombosis and Haemostasis (2025). DOI: 10.1016/j.jtha.2025.06.007 Access the stroke risk calculator Journal information: Journal of Thrombosis and Haemostasis

Provided by University of North Carolina at Chapel Hill School of Medicine