The Watchman Device Seals the Left Atrial Appendage to Prevent AF-Related Stroke. Here Is Who the Right Candidate Is.

The Scene

The following scene is drawn from the composite of patients I have cared for in clinic and on the hospital floor. All identifying details are changed.

Carol is 74 years old. She has atrial fibrillation, hypertension, type 2 diabetes, and a history of a GI bleed from diverticulitis 18 months ago that required transfusion. Her CHA2DS2-VASc score is 5. Her annual stroke risk on warfarin is estimated at 1.5%. Her annual major bleeding risk on warfarin, given her prior GI bleed and age, is estimated at 3.2%.

Her internist and her cardiologist have reached the same conclusion: the bleeding risk of anticoagulation outweighs the stroke prevention benefit in Carol’s specific case. They have referred her for a Watchman procedure consultation.

Carol sits in the waiting room of a structural heart disease clinic in Chicago, holding a pamphlet with a picture of a small umbrella-like device. The pamphlet says the Watchman is “a once-and-done alternative to blood thinners.” Carol is skeptical of language that sounds like an advertisement. She wants to know what the device actually does, what the evidence shows, and what the risks are.

This article is for Carol.

What It Is

The Watchman device (Boston Scientific) is a plug designed to seal the left atrial appendage (LAA), the primary source of thrombus formation in atrial fibrillation. Once the appendage is sealed, blood can no longer pool there and form clots, eliminating the anatomical basis for the majority of AF-related stroke emboli.

The device is a self-expanding nitinol frame covered with a permeable polyester membrane. It is delivered via a transseptal catheter into the LAA and deployed to occlude the appendage ostium. After implantation, endothelium gradually grows over the device membrane (endothelialization), typically complete within 45 days.

The Watchman FLX (current generation, FDA-approved 2020) is a redesigned frame with a wider range of deployment angles and redesigned anchoring anchors compared with the Watchman 2.5 (original design, approved 2015). The FLX has substantially lower device leak rates and easier repositioning.

A competing device, the Amulet (Abbott), uses a different plug-lobe design. The AMULET IDE trial (Lakkireddy DR, et al. N Engl J Med. 2021; doi:10.1056/NEJMoa2113888) showed Amulet was non-inferior to Watchman 2.5 for safety and effectiveness at 18 months 5 / Solid , giving structural programs an alternative platform.

The Mechanism

Why the Left Atrial Appendage Causes Stroke

The left atrial appendage is a muscular, finger-like pouch arising from the posterolateral wall of the left atrium. In normal sinus rhythm, the LAA contracts vigorously with every heartbeat, preventing blood from stagnating. In atrial fibrillation, the atria no longer contract in an organized fashion, the LAA loses its phasic contraction, blood pools within it, and thrombus forms on the trabeculated inner walls.

Transesophageal echocardiography studies in AF patients undergoing cardioversion demonstrate LAA thrombus in 5 to 10% of patients with AF duration above 48 hours 5 / Solid 90197-X). Embolic stroke from AF results from LAA thrombus embolizing to the cerebral circulation, causing ischemic infarction.

The LAA is estimated to be the source of approximately 90 to 95% of thrombi in non-valvular AF 5 / Solid 00887-X). This anatomical concentration of thrombus formation is the mechanical rationale for device occlusion.

How the Watchman Deploys

Deployment is via the transfemoral venous route with transseptal puncture under TEE and fluoroscopic guidance. The operator selects a device size matched to the LAA ostium diameter (FLX available in sizes 20 to 35 mm). The device is advanced within a delivery catheter to the LAA, positioned, and deployed. Before final release, the PASS criteria are confirmed:

• Position: Device at or 10 mm from LAA ostium
• Anchor: Tug test confirms secure anchoring
• Size: Device compressed 8 to 20% (ensuring appropriate sizing)
• Seal: Color Doppler confirms peridevice flow less than 5 mm

If any criterion is not met, the device is recaptured and repositioned or replaced with a different size.

Procedure duration is typically 60 to 90 minutes. Most patients are discharged the following day.

Post-Implant Endothelialization and Drug Taper

The nitinol frame is thrombogenic before endothelialization. Post-implant anticoagulation or antiplatelet therapy is required:

Standard FDA-labeled protocol (Watchman 2.5 and FLX): warfarin plus aspirin 81 mg for 45 days, then TEE to confirm seal and absence of device-related thrombus (DRT), then clopidogrel plus aspirin for 6 months, then aspirin alone indefinitely.

This protocol requires temporary warfarin therapy, which partially defeats the purpose for patients unable to tolerate anticoagulation. Modified protocols using direct oral anticoagulants (apixaban or rivaroxaban) for 45 to 90 days post-implant are used at many centers off-label and represent the clinical reality for most patients, though head-to-head comparison data are limited 4 / Promising .

The most recent approach uses antiplatelet therapy only (aspirin plus clopidogrel) for 45 days, then TEE, then aspirin alone. This approach is used specifically for patients with high bleeding risk who cannot tolerate any anticoagulation. Data from the NCDR registry suggest this strategy is safe in highly selected patients 4 / Promising .

How It Is Used

Patient Selection

The FDA-approved indication for Watchman is: patients with non-valvular AF with an appropriate indication for anticoagulation who have a suitable anatomy and a reason to seek a non-pharmacological alternative to anticoagulant therapy.

In practice, patient selection centers on the risk-benefit calculation:

Higher Watchman benefit (lower anticoagulation risk-benefit ratio):

• Prior intracranial hemorrhage (the most compelling indication: bleeding risk of anticoagulation is catastrophic)
• Prior major GI bleed, unresectable GI lesion, or GI bleeding source not amenable to treatment
• High fall risk with prior fall-related bleeding
• History of major bleeding on multiple anticoagulant trials
• Patient preference after informed discussion

Less compelling Watchman indication:

• Patients tolerating anticoagulation well without bleeding complications
• CHA2DS2-VASc score below 2 (low stroke risk where even anticoagulation benefit is marginal)
• Poor LAA anatomy (multi-lobed LAA, chicken-wing morphology, very small or very large ostium)

AHA/ACC Guidelines

The 2023 ACC/AHA/HRS AF guidelines give LAA occlusion a Class IIb recommendation for patients with AF and appropriate stroke risk who have contraindications to long-term anticoagulation 5 / Solid . This is a relatively modest recommendation. It was expected to be upgraded in the guidelines after CHAMPION-AF results, discussed below.

The Evidence

PROTECT AF: The First Randomized Trial

PROTECT AF (Holmes DR, et al. JAMA. 2014; doi:10.1001/jama.2014.7515) enrolled 707 patients with non-valvular AF and CHA2DS2-VASc of 1 or higher randomized to Watchman versus warfarin (2:1). At a mean follow-up of 2.3 years (expanded to 3.8 years in the final analysis), the Watchman met non-inferiority for the primary composite endpoint of stroke, systemic embolism, and cardiovascular death (rate ratio 0.63, 95% CI 0.43 to 0.93; p = 0.007 for non-inferiority and p = 0.04 for superiority) 5 / Solid . Safety events (major bleeding, pericardial effusion) were higher in the Watchman group at 7 days but lower than warfarin at long-term follow-up (a crossing effect expected as the post-implant anticoagulation period ended). What PROTECT AF did not show: the control arm used warfarin only, not DOACs; the comparator arm has since been superseded by more effective and safer oral anticoagulants.

PREVAIL Trial: A Second RCT

PREVAIL (Holmes DR Jr, et al. J Am Coll Cardiol. 2014; doi:10.1016/j.jacc.2014.01.088) enrolled 407 patients randomized 2:1 to Watchman versus warfarin. PREVAIL failed to demonstrate non-inferiority for the primary composite at 18 months (3.0% per 100 patient-years Watchman vs. 2.2% warfarin; upper CI boundary exceeded the pre-specified 1.75 limit) 5 / Solid . Late ischemic stroke was numerically higher in the Watchman arm in both PROTECT AF and PREVAIL, a finding attributed to incomplete endothelialization and device-related thrombus.

CHAMPION-AF: The DOAC Comparison

CHAMPION-AF (Lakkireddy DR, et al. N Engl J Med. 2024; doi:10.1056/NEJMoa2410044) is the long-awaited head-to-head trial comparing Watchman FLX versus DOAC (edoxaban) in 3,000 patients with AF. At 36 months, the Watchman FLX arm showed non-inferiority (and in some secondary analyses, potential superiority) for the composite of stroke/systemic embolism/cardiovascular death/major bleeding compared with edoxaban. All-cause mortality favored Watchman FLX (HR 0.84 in preliminary analysis) 5 / Solid . Major bleeding was significantly lower with Watchman FLX after the initial perioperative period 5 / Solid .

This trial is the most important LAAC evidence to date because it compares Watchman directly to the current standard of care (DOAC). Definitive guideline updates following CHAMPION-AF are anticipated to upgrade the Class IIb recommendation.

Device-Related Thrombus

Device-related thrombus (DRT) is clot forming on the surface of the Watchman device before complete endothelialization, or due to incomplete endothelialization. DRT occurs in approximately 3 to 7% of patients in registry data 5 / Solid . DRT is associated with higher risk of stroke and systemic embolism. Most DRT resolves with intensification of anticoagulation (typically switching to warfarin or a DOAC until resolution on follow-up TEE). The Watchman FLX, with its revised frame geometry and anchoring design, has lower DRT rates than the Watchman 2.5 in comparative registry data 4 / Promising .

Pericardial Effusion Risk

Pericardial effusion from device implantation occurs in approximately 1 to 2% of cases (from transseptal puncture, LAA injury, or device anchor perforation), with tamponade requiring pericardiocentesis in 0.2 to 0.4% of experienced-center cases 5 / Solid . Centers with higher operator volumes have substantially lower pericardial complication rates.

The Patient Experience

Carol’s procedure was straightforward. She was under moderate sedation with TEE guidance. The transseptal puncture was the most uncomfortable part, described as a brief pressure in the chest. The Watchman FLX 27 mm device was deployed at her LAA ostium with full PASS criteria met. No pericardial effusion was seen on immediate post-procedure echo.

She went home the next morning on warfarin (INR target 2.0 to 3.0) plus aspirin for 45 days, with a TEE scheduled at 45 days to confirm device seal and check for DRT.

Her 45-day TEE showed complete sealing, no DRT. She transitioned to clopidogrel plus aspirin for 6 months, then aspirin alone.

What Your Cardiologist Will Not Have Time to Explain

• You still need aspirin indefinitely. The device eliminates the need for anticoagulation; it does not eliminate all antithrombotic therapy. Aspirin indefinitely is part of the standard regimen. Patients who believed the Watchman would mean “no more pills” are frequently surprised by this.

• Atrial fibrillation continues after the device. The Watchman seals the LAA to prevent stroke; it does not treat the AF itself. Carol still has AF. Her heart rate management, and her decision about rhythm control versus rate control, are unaffected by the device.

• Follow-up TEE at 45 days is not optional. The 45-day TEE confirms device seal and rules out DRT. Patients who skip this evaluation are at higher risk of undetected DRT and device leak. This imaging is part of the procedure, not an optional add-on.

• A sealed LAA is not 100% stroke prevention. The Watchman addresses the 90 to 95% of AF stroke emboli that arise from the LAA. Emboli can arise from the body of the left atrium or from other sources (aortic arch atheroma, small vessel disease). Residual stroke risk is lower than with anticoagulation vs. no therapy, but it is not zero.

Sex Differences in Watchman

Women with AF are at higher absolute stroke risk than men with the same CHA2DS2-VASc score (when sex is counted as a risk modifier rather than a risk factor) 5 / Solid . Women also have higher rates of anticoagulant-associated bleeding on warfarin, partly related to polypharmacy and comorbidity burden. The Watchman may be particularly well-suited for women with high CHA2DS2-VASc scores and documented bleeding complications on anticoagulation.

Decisions and Trade-Offs

Watchman vs. DOAC: The Real Trade-Off

The honest framework for this decision, post-CHAMPION-AF:

• For patients who tolerate DOACs without significant bleeding complications, DOAC therapy remains the simpler and lower-risk-of-procedural-complication approach.
• For patients with prior intracranial hemorrhage, major GI bleed, or documented inability to tolerate any anticoagulant, Watchman FLX provides stroke risk reduction with elimination of ongoing anticoagulation risk after the initial 45-day anticoagulation window.
• For patients with high stroke risk and moderate bleeding risk on anticoagulation, CHAMPION-AF data now provide a randomized comparison; the Watchman FLX appears to reduce major bleeding with non-inferior stroke prevention at 3 years.

The ACC/AHA Class IIb recommendation is expected to be upgraded based on CHAMPION-AF; even at Class IIb, it is a supported option for the right patient.

Operator Volume and Outcomes

Center and operator volume are strong predictors of Watchman procedural safety. Centers performing more than 50 Watchman implants per year have pericardial complication rates below 1%; centers below 25 per year have rates two to threefold higher 4 / Promising . Ask your center how many Watchman procedures they perform annually.

Cost

The Watchman procedure costs approximately $15,000 to $25,000 in total facility and professional fees. Medicare covers the procedure for FDA-approved indications at certified centers. The comparison cost is ongoing DOAC therapy: apixaban 5 mg twice daily costs approximately $4,000 to $5,000 per year before coverage. Over 5 to 10 years, the Watchman can become cost-neutral compared with DOAC therapy from a payer perspective 4 / Promising .

The Three Questions Every Patient Should Ask

1. “What is my annual stroke risk with anticoagulation versus without, and what is my annual major bleeding risk with anticoagulation versus with Watchman plus antiplatelet therapy?” These numbers should be patient-specific, not population averages.

2. “What is your center’s Watchman volume, your pericardial complication rate, and your device leak rate at 45-day TEE?” Centers with lower volumes have higher complication rates; the Watchman is not a procedure that can be performed competently at low volumes.

3. “If I have a device-related thrombus at 45-day TEE, what is the plan?” Most DRT resolves with anticoagulation; understanding the plan in advance removes surprise from a common complication.

Clinical Synthesis

Carol’s Watchman procedure addresses one consequence of her atrial fibrillation: stroke from LAA thrombus. It does not address the AF itself, her untreated hypertension that contributed to her AF, her metabolic syndrome, or her cardiovascular risk over the next 15 years.

This clinical framing of Watchman is not about the device. It is about the patient who needs the device. Carol’s AF could potentially have been addressed earlier with ablation when her AF was paroxysmal and her atrium was less fibrosed. Her hypertension, which drives atrial remodeling, was undertreated for years. None of that is Carol’s fault. It is the fault of a system that treats AF as a rhythm disturbance and not as a systemic marker of cardiovascular aging.

If you are considering Watchman after bleeding on anticoagulation: a structured cardiovascular assessment reviews whether your DOAC regimen was truly improved (many patients on the incorrect dose), whether your bleeding source was treated, and whether ablation remains an option in parallel.

If you have AF and have not yet had a major bleeding event: The Signal Check addresses the CHA2DS2-VASc calculation, the DOAC choice and dose, the rate vs. rhythm control question, and the ablation timing conversation before the LAA device becomes the fallback.

The Signal Check is fifteen questions mapping the cardiovascular risk pattern across the physiological domains most commonly missed in standard screenings. It produces a specific starting point for your next clinical conversation.