Treating “upper endoscopy” as one uniform bleeding-risk procedure can lead to two avoidable harms: stopping warfarin when it is not needed, or arriving unprepared for a procedure that does require interruption. In an illustrative work-up for possible esophageal cancer in a patient with a mechanical mitral valve, the key question is what the endoscopist plans to do—not simply whether the procedure is called an EGD or EUS.
Consider an illustrative patient with progressive dysphagia under evaluation for possible esophageal cancer; malignancy is not yet established. He takes warfarin for atrial fibrillation and a mechanical mitral valve. The booking request has not specified whether the endoscopist plans diagnostic EGD with standard mucosal biopsy, EUS without sampling, or EUS-guided needle sampling. Prompt evaluation and a specific procedural plan are both important.
The case does not give the patient’s current INR, prescribed therapeutic range, renal function, history of embolism during a prior interruption, or complete medication list. Those details matter if interruption is being considered. So does baseline function and the patient’s current ability to maintain hydration and nutrition; neither should be inferred from the diagnosis under investigation.
Name the exact procedure before changing warfarin
Guidelines generally classify diagnostic upper endoscopy with standard mucosal biopsy as low risk for bleeding. EUS without sampling is also generally low risk. By contrast, EUS-guided fine-needle aspiration or biopsy (FNA/FNB) and esophageal stricture dilation are classified as higher-risk procedures.
| Planned procedure | Usual bleeding-risk category | Practical warfarin approach |
|---|---|---|
| Diagnostic EGD with standard mucosal biopsy | Low | Generally continue warfarin; confirm the INR is within the patient’s prescribed therapeutic range. |
| EUS without needle sampling or intervention | Low | Generally continue warfarin, with the INR in the therapeutic range. |
| EUS-guided FNA/FNB or stricture dilation | High | Warfarin interruption is commonly planned; the need for bridging requires individualized discussion for a mechanical mitral valve. |
For the first procedure, a reflex order to stop warfarin for several days is not the default. If the INR is above the prescribed range, or if the endoscopist expects an intervention beyond routine biopsy, the plan may change. Make that distinction before the patient arrives, rather than leaving it to an assumption in the procedure booking.
A mechanical mitral valve changes the interruption discussion
The valve makes the patient’s thrombotic risk an important part of the plan if warfarin must be held. But it does not change the bleeding-risk classification of a mucosal biopsy—and it does not mean every endoscopy requires heparin bridging. If warfarin is continued for a low-risk procedure, there is no interruption to bridge.
For high-risk endoscopy, recommendations are not fully aligned. BSG/ESGE recommends warfarin interruption with low-molecular-weight heparin (LMWH) substitution for patients at high thrombotic risk, including those with a mechanical mitral valve. ACG/CAG advises against routine bridging when warfarin is held for elective endoscopy but notes that bridging may be appropriate for selected patients with mechanical valves, after specialist consultation. CHEST’s 2022 guideline instead suggests against heparin bridging for patients with a mechanical valve who require warfarin interruption for an elective procedure. The 2024 AHA/ACC perioperative guideline supports selective bridging in patients at highest thrombotic risk, including those with mechanical mitral valves, when interruption is necessary. That cardiovascular guideline addresses noncardiac surgery, however; it is not an endoscopy-specific protocol.
The evidence gap helps explain the disagreement. The BRIDGE trial, often cited in arguments against bridging, studied patients with atrial fibrillation and excluded those with mechanical valves. PERIOP-2 included patients with mechanical valves, but all received preprocedural bridging; the trial randomized postoperative bridging, not whether to bridge before the procedure. Evidence therefore does not justify turning either “always bridge” or “never bridge” into a universal endoscopy rule.
If EUS-guided sampling is planned, the endoscopist and the clinician managing anticoagulation should agree on the interruption and resumption plan before the procedure. Endoscopy guidance such as BSG/ESGE commonly uses about a five-day warfarin interruption for high-risk procedures and recommends checking that the INR is below 1.5 before the procedure. For this patient, whether to bridge, when to start and stop heparin, and when to restart warfarin should be decided with the anticoagulation team and, when appropriate, cardiology. Renal function, bleeding history, the exact valve and anticoagulation indication, and the urgency of the procedure may affect that plan.
Make the booking match the clinical question
A useful procedural request specifies what is intended: “diagnostic EGD with mucosal biopsy; no dilation or EUS-guided sampling planned” is different from “EUS with possible needle sampling.” If both diagnostic endoscopy and higher-risk sampling may be needed, ask whether they can be staged. A separate plan can avoid interrupting warfarin merely in case a higher-risk maneuver becomes desirable, while still allowing the team to prepare safely if sampling is essential.
Keep urgency and anticoagulation planning in view at the same time. Progressive dysphagia warrants timely assessment. Substantial weight loss or nocturnal regurgitation with coughing would increase concern and should prompt attention to aspiration symptoms, oral intake, hydration, and functional decline. These features alone do not establish nutritional status, aspiration pneumonia, or the immediate need for a particular therapeutic maneuver.
Common traps are calling all endoscopy “high risk,” treating mechanical mitral valves as an automatic instruction to bridge, and assuming a negative history for embolism when the case has not supplied one. Another is ordering a warfarin hold without naming the intervention, the INR check, who will manage any bridge, and when anticoagulation will resume.
Practical takeaways
- Classify the planned maneuver, not just the endoscopy label.
- For diagnostic EGD with standard mucosal biopsy, warfarin is generally continued if the INR is within the prescribed therapeutic range.
- EUS without sampling is different from EUS-guided FNA/FNB; sampling and stricture dilation are higher-risk procedures.
- If warfarin interruption is needed, a mechanical mitral valve makes the bridging decision important—but not automatic. Guidance differs, and evidence is limited.
- Confirm the INR, renal function, valve and embolic history, other antithrombotic drugs, and a written stop/restart plan with the procedural and anticoagulation teams.
