A 60-year-old Chinese man was diagnosed with atrial fibrillation 2 years earlier, when he presented with mild dyspnea. His medical history included hypertension and mild heart failure due to nonischemic cardiomyopathy, but no diabetes or stroke. Digitalis was prescribed, and beta-blockade was optimized for control of ventricular response. The patient's symptoms resolved. He was prescribed anticoagulant therapy with warfarin (Coumadin) to reach an international normalized ratio (INR) target range of 2 to 3.

The patient was followed at an anticoagulation clinic, which monitored his prothrombin time and adjusted his warfarin dosage. His INR over the previous 2 years ranged from 1.6 to 3.5, with approximately 60% of values within the target range. He remained in his baseline state of health from the time he was diagnosed with atrial fibrillation until he presented with intracranial hemorrhage (ICH) resulting in coma. Computed tomography scans showed no other pathology, and there was no antecedent head trauma. His INR on admission was 2.8.

The National Registry for Atrial Fibrillation reported the CHADS2 (congestive heart failure, hypertension, age > 75 years, diabetes, and previous stroke or transient ischemic attack) score for stroke risk assessment in patients with atrial fibrillation. Each of the following characteristics is assigned 1 point: age ≥ 75 years, hypertension, diabetes, and heart failure. Patients with prior stroke or transient ischemic attack are assigned 2 points. The patient described in this case report had a history of hypertension and heart failure, giving him a CHADS2 score of 2, which placed his annual estimated risk of stroke at 3% to 5%. At this level of risk, anticoagulation with warfarin to an INR target of 2.5 (range 2-3) is recommended by the American College of Cardiology/American Heart Association guidelines on atrial fibrillation. However, anticoagulation therapy is associated with bleeding, the most severe of which is ICH. Age and anticoagulation intensity are the most powerful predictors of severe bleeding. This patient was younger than most patients in clinical practice (mean age, 75 years), whereas his achieved anticoagulation intensity was similar to that reported in cohort studies. Based on published reports, his annual risk of ICH was estimated to be about 1%.

Very few studies on the risk and benefit of anticoagulation therapy for patients with atrial fibrillation have included nonwhite patients. Whether the estimates of this patient's risk of ICH based on published literature were accurate remain unknown.