Abstract 13114: Predictive Value of the Has-Bled Score in Patients with Atrial Fibrillation and Chronic Kidney Disease Using Vitamin K-Antagonists
Aim The HAS-BLED score enables a risk estimate of major bleeding in patients with atrial fibrillation (AF) on vitamin K-antagonists (VKA) treatment, but has never been validated in patients with chronic kidney disease (CKD). We analyzed the predictive value of the HAS-BLED score in CKD compared with non-CKD patients.
Methods Medical records of 416 CKD(eGFR0-30 ml/min, 30-60 ml/min) and 300 non-CKD patients starting VKA treatment for AF between 1997-2005 were searched for items on the HAS-BLED score (hypertension; renal or liver disease; stroke; major bleeding or anemia; labile INR, (time within therapeutic range<60%), age>65 years; use of NSAIDs, anti-platelet therapy or alcohol) and major bleeding events. Areas under the curves (AUC) of the receiver operating characteristic (ROC) were calculated for the total population and CKD patients.
Results Mean HAS-BLED score in CKD patients was 3.1 versus 2.6 in non-CKD patients(P<0.01). Major bleeding occurred in 115/716 (16.1%,95%CI12.8-19.9%) patients. The AUC of the ROC analysis in the total population was 0.50 (95%CI0.44-0.56); 0.53 (95%CI0.43-0.62) in patients with an eGFR 30-60 ml/min, and 0.35 (95%CI0.21-0.49) in patients with an eGFR<30 ml/min. In a Cox regression analysis performed in all patients, renal impairment, labile INR and age>65 years were predictive of major bleeding, hazard ratios 2.5 (95%CI1.2-5.2), 2.2 (95%CI1.2-3.7) and 5.2 (95%CI1.6-16.8), respectively.
Conclusion Performance of the HAS-BLED score was limited in the total population and further reduced in patients with an eGFR< 30 ml/min. Further research is needed before the HAS-BLED score can be used in CKD patients.
- © 2012 by American Heart Association, Inc.