Prospective Randomized Evaluation of the WATCHMAN™ LAA Closure Device in Patients With Atrial Fibrillation Versus Long Term Warfarin Therapy


The PREVAIL trial was designed to confirm the results of the PROTECT AF trial and validate the safety of the implant procedure, including a minimum of 20% of randomized patients enrolled at institutions that did not participate in a previous WATCHMAN study (PROTECT AF or CAP Registry) and a minimum of 25% of randomized patients enrolled by new operators at any institution.

A total of 407 patients were randomized 2:1 device vs. warfarin control in the PREVAIL trial. Patients enrolled in PREVAIL were warfarin eligible and had a mean CHADS2 score of 2.6 ± 1.0.

Data continues to accumulate on the WATCHMAN Device from patient populations studied in the clinical trials and registries. The WATCHMAN Device is an investigational device and is not approved in the United States.  No claims as to the safety or efficacy of the device may be made.

The pre-specified co-primary endpoints for the PREVAIL trial were as follows:

  • Primary Efficacy Endpoint: Comparison of composite of stroke, systemic embolism, and cardiovascular/unexplained death at 18 months follow-up
  • Late Ischemic Primary Efficacy Endpoint: Comparison of ischemic stroke or systemic embolism occurring from greater than 7 days post randomization to 18 months follow-up
  • Early Safety Endpoint: Acute (7-day) occurrence of death, ischemic stroke, systemic embolism and procedure or device related complications requiring open cardiovascular surgery or major endovascular intervention in device group only


PREVAIL trial enrolled patients with higher risk of stroke as determined by the CHADS2 scoring system

  • PREVAIL CHADS2 score was 2.6 versus 2.2 in PROTECT AF.
  • Average age of the patient in PREVAIL was 74 years compared to 71 years in PROTECT AF.
  • 33.8% of patients had diabetes in PREVAIL vs. 24.4% in PROTECT AF.
  • 27.5% of patients had prior stroke compared to 17.7% in PROTECT AF.

Demographics: Device Patients



The PREVAIL trial results below are based on the final pre-specified data analysis from January 2013 and have been published in the peer-reviewed Journal of American College of Cardiology (JACC).  An ad-hoc analysis including additional data from the PREVAIL trial was presented by Dr. Vivek Reddy at the 2014 TCT Conference and can be accessed via TCTMD here.

Primary Efficacy Endpoint: Non-inferiority WAS NOT ACHIEVED for the pre-specified criteria for the primary efficacy endpoint (composite of all stroke, systemic embolism, and cardiovascular/unexplained death at 18 months).  The ad-hoc analysis presented by Dr. Vivek Reddy at the 2014 TCT Conference included additional data on this endpoint and can be accessed via TCTMD here.

  • At 18 months, the WATCHMAN Device demonstrated a rate of 0.064 vs. 0.063 for the control resulting in a rate ratio of 1.07. The observed upper bound of the rate ratio was 1.89, slightly greater than the pre-specified criterion of 1.75 (95% CI).
  • Dr. Holmes, et al, stated that an over performing control group and wide credible intervals likely contributed to the missed endpoint and that the longer-term follow-up of the PROTECT-AF trial has confirmed the efficacy of the therapy.
  • The PREVAIL treatment group performed similarly to prior WATCHMAN studies, however the observed rate of stroke and systemic embolism (0.7 per 100 patient-years) in the PREVAIL Control group was lower than warfarin control groups in recently published studies (1.6-2.2 per 100 patient-years):

Late Ischemic Primary Efficacy Endpoint: Non-inferiority WAS ACHIEVED for the pre-specified late ischemic primary efficacy endpoint occurrence of late ischemic stroke and systemic embolism (>7 days post randomization) at 18 months.  The additional data from the PREVAIL trial included in the ad-hoc analysis presented by Dr. Vivek Reddy at the 2014 TCT Conference now indicates that non-inferiority was not achieved for this endpoint. The analysis can be accessed via TCTMD here.

  • The 18-month event rate was 0.0253 per 100 patient-years, resulting in an 18-month rate difference between the device and control group of 0.0053 per 100 patient-years. The upper bound of 0.0273 was less than the pre-specified criterion of 0.0275 (95% CI).

Early Safety Endpoint:
 The pre-specified endpoint criteria WAS ACHIEVED for the early safety endpoint (procedural related safety <7days).

  • The observed adverse event rate was 2.2%, resulting in an observed upper bound of 2.652% which is less than the pre-specified criterion of 2.67% (95% CI).

Additional safety analysis of PREVAIL shows:

  • A significant increase in implant success rate (95.1%) compared to PROTECT AF (90.9%). Additionally, new implanter success rate was 93.2% with no significant differences in complication rates between experienced and new implanters, demonstrating the success of our physician training program.  

  • Following successful implantation, 99.3% of patients (141/142) were able to discontinue warfarin after 12 months

Warfarin Cessation

Study 45 Day 12 Month
PROTECT AF 86.6% 93.2%
PREVAIL 92.2% 99.3%

  • Seven-day serious procedure/device related vascular complications occurred in 4.5% of subjects in PREVAIL vs. 8.7% of subjects in PROTECT AF, a 52% reduction.
  • Pericardial effusions requiring intervention occurred at a rate of 1.9% in PREVAIL vs. 4.0% of patients in PROTECT AF, a 52% reduction and at a rate comparable to other left atrial procedures.6
  • Procedural/Device related strokes occurred at a rate of 0.4% in PREVAIL vs. 1.1% in PROTECT AF, a 64% reduction.
  • Device embolization was infrequent, occurring in only four patients (2 in PROTECT AF, 2 in PREVAIL).