Refining Stroke Prediction for African-American AF Patients

Study Questions:

Does the inclusion of African-American (AA) race to the CHA2DS2-VASc score improve the prediction of stroke in patients with atrial fibrillation (AF)?


Using Medicare administrative data, the authors identified patients with AF between 2009 and 2012. Medicare administrative data were also used to identify patients with stroke between 2010 and 2012. Race was ascertained using a validated code based on first and last name algorithms. Components of the CHA2DS2-VASc score were identified using International Classification of Diseases Clinical Modification (ICD) 9 codes. Time to event models were used to determine the relationship between CHA2DS2-VASc, race, and stroke. A CHA2DS2-VASc-R score was calculated by adding 1 additional point to the score for AA race. The CHA2DS2-VASc and CHA2DS2-VASc-R scores were compared and components of the CHA2DS2-VASc-R score (e.g., age, prior stroke, etc.) were evaluated to determine their relative importance to the overall score.


In the sample, there were 460,417 patients with AF and 5.4% were AA. The mean age was similar across racial and ethnic groups, but AA patients had more comorbidities, and therefore, higher CHA2DS2-VASc scores. AA patients had higher rates of stroke (38.0 per 1,000 patient years) than other racial and ethnic groups. Modeling showed that, when compared to whites, AA patients had an increase in stroke risk that was greater and 1 additional point, but less than 2 additional points on the CHA2DS2-VASc score. The addition of other racial or ethnic groups had no impact on the CHA2DS2-VASc score. By adding AA race to the CHA2DS2-VASc score (e.g., CHA2DS2-VASc-R) and assigning 1 point to AA race, the model better predicted stroke risk for AA patients and did not significantly change predictions for white patients. When the components of the CHA2DS2-VASc-R score were examined, only history of stroke, age ≥75 years, and female sex were more important than AA race in predicting stroke.


The addition of AA race to the CHA2DS2-VASc score improves its ability to predict stroke.


AF increases the risk of stroke, and strokes due to AF tend to be associated with more disability than other stroke subtypes. Studies have shown that AAs have a higher risk of stroke than whites, and the reasons for this disparity are not entirely clear. The CHA2DS2-VASc score is used to predict stroke risk in patients with AF and to make decisions about anticoagulation, which can dramatically reduce the risk of stroke in this patient population. This study is important because it shows that even though AA patients, on average, have more comorbidities that increase their risk of stroke, AA race is an independent predictor of stroke risk in patients with AF. While this study is limited by only including Medicare patients who are >65 years old, it provides support for future prospective studies that can determine if use of the CHA2DS2-VASc-R score improves clinical outcomes.

< Back to Listings