In the first of our 2 part series, we discussed how digital tools support earlier detection and better monitoring of atrial fibrillation (AF). In this article, we explore the rise of risk scores in AF, another significant development in the field.
Summary
AF risk scores have evolved over the years and newer risk scores include biomarkers
Risk scores require time to develop and subsequent validation. Therefore, first movers stand an advantage
CHA2DS2-Vasc is recommended by guidelines and widely used to estimate stroke risk in AF
Whilst newer risk scores involving biomarkers outperform older ones, biomarker testing requires additional time and cost, which makes uptake challenging
Atrial fibrillation is a type of cardiac arrhythmia commonly encountered in clinical practice. Stroke is a common complication of AF, costing up to 45 billion per year in the European Union.
To mitigate the risk of stroke, patients are prescribed oral anticoagulants (blood thinners), such as warfarin. But this is done with caution as anticoagulants increase the risk of bleeding and major bleeding events, especially intercranial bleeds, may be devastating to the elderly.
Researchers have therefore developed risk scores, combining clinical, demographic, and even biomarker measures to predict the risk of stroke in AF patients. This has direct implications in the need for anticoagulants. Another group of risk scores help to predict the risk of bleeding.
Several risk scores are well-known in the field:
CHADS2
First published in 2001, the CHADS2 risk score is a relatively simple schema assigning 1 point each to three risk factors: congestive heart failure, hypertension and age 75 and above. Prior stroke/thromboembolism is assigned two points.
The total score corresponds to the 1-year risk of stroke, 0 points being 1.9% risk and a full 6 points being 18.2% risk.
Whilst known as a simple and widely applicable risk score in its time, the CHADS2 score classified the majority of patients as being at intermediate risk (including some patients who may be at low risk) and did not include newer risk factors such as being female.
CHA2DS2- VASc
The CHA2DS2- VASc score is the main score currently used to estimate stroke risk in AF patients. It is included in European and American guidelines and commonly used in clinical practice.
The score is essentially an upgrade of the CHADS2 score, which was refined as the understanding of stroke risk factors improved. It included new risk factors such as being female.
The score is actionable, which may account for good uptake. A score of 2 or more in men and 3 or more in women means oral anticoagulant therapy is recommended.
HAS-BLED
The HAS-BLED score is usually used to assess the 1-year risk of major bleeding in AF patients on anti-coagulants.
A high HAS-BLED score should not result in withholding oral anti-coagulants. But bleeding risk factors should be identified and treatable factors addressed.
Whilst HAS-BLED is mentioned in guidelines, bleeding and stroke risk factors overlap and a high score in one generally implies a high score in the other.
ABC-stroke and ABC-bleeding risk score
More recently developed, the ABC-stroke and ABC-bleeding risk score combines age, biomarkers and clinical history to estimate the risk of bleeding in AF patients.
The ABC-stroke score includes age, test results of 2 cardiac biomarkers NT-ProBNP, High Sensitivity Troponin, and clinical history. The ABC-bleeding score includes similar factors with the additional use of hemoglobin and the GDF15 biomarker.
Recent studies show that ABC-stroke outperformed CHA2DS2- VASc at predicting stroke and ABC-bleeding fared better than HAS-BLED at predicting bleeding. ABC-bleeding is mentioned in guidelines. However, logistically, the need for biomarker testing may mean increased costs and time incurred, unlike CHA2DS2- VASc which can be calculated on the spot.
Other commonly discussed risk scores include ATRIA, ORBIT and HEMORR2HAGES.
While risk scores enable better stratification of patients, it seems to be a competitive field in AF with established clinician preferences. So what makes a successful risk score?
Early birds might have an advantage. CHA2DS2-VASc rode of the success of CHA2DS2 and both have simple, easy-to-calculate schematics. Risk scores usually take years to develop and they are subsequently validated in various cohorts through multiple studies, giving first movers a natural advantage.
As with the CHA2DS2-VASc risk score, the link to actionability seems like a big plus. Based on the score, doctors know whether to prescribe anti-coagulants or not.
The ABC risk scores while outperforming their predecessors, also add new costs and time-incurring elements as biomarker testing is needed. This may mean an uphill battle in changing physician habits, a battle worth fighting if it leads to better patient outcomes.
As the population ages, so will the burden of AF on society. The good news is that innovation is on its way, as digital health and risk scores have shown.