Major Implications for New Atrial Fibrillation Guidelines

A new analysis finds that updates to atrial fibrillation (AF) treatment guidelines will encourage American doctors to prescribe oral anticoagulants to almost 1 million more patients.

The increase stems entirely from a simple substitution in recommendations from the American Heart Association (AHA)/American College of Cardiology (ACC)/Heart Rhythm Society (HRS). In 2011, those groups urged physicians to measure stroke risk with CHADS2 scores. In 2014, they urged physicians to measure it with CHA2DS2-VASc scores.

To estimate the affect of this substitution on oral anticoagulant eligibility, researchers calculated its exact impact on a representative study cohort — in this case the Outcomes Registry for Better Informed Treatment of Atrial Fibrillation (ORBIT-AF) — and extrapolated their findings to the nation’s entire population of AF patients.

Their analysis found that that while the old guidelines recommended oral anticoagulants for most AF patients (71.8%), the new guidelines make such recommendations for an overwhelming majority of them (90.8%).Indeed, the new guidelines end up recommending oral anticoagulation almost universally to certain large subpopulations such as patients aged 65 and over (98.5%) and female patients of any age (97.7%).

A research letter synopsizing the analysis appears in JAMA Internal Medicine, along with an editorial that questions the justification for urging oral anticoagulants on so many more patients.

“It is important to realize that risk schemes for stroke have only modest predictive ability and, when directly compared with each other, neither risk scheme emerges as clearly superior,” wrote Margaret C. Fang, MD, an associate professor of medicine at the University of California, San Francisco.

“In fact CHA2DS2-VASc and CHADS2 have nearly identical discrimination based on their C statistics.”

The CHADS2 scale uses 5 factors to evaluate stroke risk: Congestive heart failure, Hypertension, Age 75+, Diabetes mellitus, Stroke or TIA or thromboembolism. Patients receive 1 point for each of the first 4 risk factors and 2 points for a history of strokes. Scores thus vary from 0 (very low risk) to 6 (very high risk).

The old AHA/ACC/HRS guidelines recommended anticoagulation for all patients with scores of 2 or more.

The CHA2DS2-VASc scale assigns an extra point to female patients and another to those with a history of vascular disease. It also refines the age score such that patients who are under 65 get 0 age points, those who are between 65 and 74 get 1 age point and those who are 75 and older get 2 age points. Scores thus range from 0 (low risk) to 9 (high risk).

All patients, therefore, score at least as high on the newer scale as on the older one (and most score higher), yet new guidelines still call for anticoagulation for all patients who score 2 or more points.

In her editorial, Chang argued that this was particularly inappropriate given that stroke rates have declined so sharply in recent years that any score on either scale indicates a far lower stroke risk today than it did when the scales were developed.

According to Fang, this overall decline magnifies the effect of the new recommendations and leads physicians to pursue an utterly untested strategy of prescribing anticoagulants to “high-risk” patients who are actually low-risk patients.

“It is possible that the new guideline, which recommends anticoagulation for a wider swath of the population, will result in lower stroke rates. However, more widespread anticoagulation will also increase the number of bleeding complications,” Fang wrote.

“Only studies that directly compare the tradeoffs of bleeding and stroke can determine which risk scheme is preferable.”