Carlsson and Tebbe mention several important points. Based on our study at a single institution, it is impossible to conclude that physicians throughout the United States are not routinely following the ACCP's recommendations regarding anticoagulation and cardioversion of atrial fibrillation, although our data are consistent with practices in Germany. Carlsson and Tebbe reported that only 48% of patients were receiving therapeutic anticoagulation at the time of cardioversion; it seems reasonable to assume that if patients were included who did not receive anticoagulation for a full 3 weeks before and 4 weeks after cardioversion, this percentage would be further reduced.
However, they also note that few recommendations in medicine are based on such strong evidence as the ACCP's recommendations regarding anticoagulation and atrial fibrillation, a statement with which we cannot agree (even though we strongly support the recommendations). Although the ACCP's recommendations regarding anticoagulation and cardioversion of atrial fibrillation are based