Abstract T P162: Predicting Recurrent Stroke: Comparison of CHADS2 to CHA2DS2-VASC Scale Among Patients With and Without Atrial Fibrillation at a Designated Primary Community Hospital Stroke Center
Background: CHADS2 and CHA2DS2-VASC are predictors of thromboembolic events for atrial fibrillation (AF). There is limited data on using CHA2DS2-VASC to evaluate stroke recurrence in community hospitals. This study compared the utility of CHADS2 and CHA2DS2-VASC to predict recurrent stroke risk in a community hospital setting.
Methods: • The study population included all ischemic stroke cases from July 2006 to July 2014.
• Demographics including age, gender, ethnicity; co-morbidities including hypertension, dyslipidemia and diabetes and neurological-impairment-on-arrival were considered as confounders. The outcome was number of recurrent strokes over the eight year period.
• CHADS2 and CHA2DS2-VASC were analyzed as continuous variables.
Results: • Of 1,986 ischemic stroke patients, 468 had AF either prior to hospitalization, at admission or at discharge.
• In our patient population, CHADS2 was a more sensitive predictor recurrent stroke predictor compared to CHA2DS2-VASC (c-statistic: 0.85 and 0.78 respectively).
• Average CHADS2 score of patients with recurrent stroke was 2.9 vs. 1.4 in patients with no recurrent stroke.
• Average CHA2DS2-VASC score of patients with recurrent stroke was 4.2 vs. 2.7 in patients with no recurrent stroke.
• With every one point increase in CHADS2 patients were six times as likely to have a recurrent stroke, while with every one point increase in CHA2DS2-VASC, patients were 2.5 times as likely to have a recurrent stroke. o Vascular disease (p=0.62) and gender (p=0.87) did not affect recurrent stroke. Hypertension (p<0.0001) and congestive heart failure (p<0.0001) were most significant predictors of recurrent stroke.
Conclusion: • At our community hospital, CHADS2 was the better predictor of recurrent strokes.
• While composite scores are clinically useful, vascular disease and gender were not strong predictors of recurrent stroke in our population.
Author Disclosures: A. Jain: None. L. Sheikh: None. D. Van Houten: None.
- © 2015 by American Heart Association, Inc.