Background:In patients with nonvalvular atrial fibrillation (NVAF),embolic stroke is thought to be associated with left atrial appendage (LAA) thrombi.The WATCHMAN LAA Occlusion Device has been shown to be noninferior to conventional oral anticoagulation with warfarin for stroke prevention in patients with NVAF.This study aimed to evaluate the procedural feasibility,safety and 12-month outcomes of the WATCHMAN LAA Occlusion Device in NVAF patients with high risk for stroke in China.Methods:The clinical data of 106 NVAF patients,who were consecutively underwent LAA closure with the WATCHMAN Device between April 2014 and May 2015,were collected.Patients were followed up at l,3,6,and 12 months after discharge.A transesophageal echocardiograph was performed at 45 days after implantation and repeated in case of an unexpected event during the follow-up period.Results:This study included 106 NVAF patients with a mean age of 64.2 ± 8.6 years (ranging from 50 to 88 years),and the mean CHA2DS2-VASc score of all patients was 3.6 ± 1.6 (ranging from 2 to 9).Among those 106 NVAF patients,100 (94.3％) patients were implanted with the device successfully.The procedural success rate was 94.3％ (100/106),and the occlusion rate was 100.0％ (100/100).There were one tamponade,one ischemic stroke,and eight minor pericardial effusions during hospitalization.During 12-month follow-up period,two patients developed a thrombus layer on the device that resolved with additional anticoagulation:one with visible device-thrombus experienced transient ischemic stroke,and one had a hemorrhagic stroke.There were no deaths in this study.The overall survival rate was 100.0％,and nonmajor adverse event rate was 95.0％ (95/100).In this study,the expected annual rate ofischemic stroke risk in these patients according to the CHA2DS2-VASc score was 4.0％,while the observed ischemic stroke rate was 2.0％ per year.Conclusions:LAA closure with the WATCHMAN Device was feasible,efficient,and safe for NVAF to prevent the accidence of stroke in Chinese patients.During the 12-month follow-up period,the observed ischemic stroke rate (2.0％ per year) in our study was lower than the predicted annual stroke risk (4.0％) using the CHA2DS2-VASc score.