Polypectomy technique, electrocautery current, and colonic bleeding risk
Before any snare touches tissue, the endoscopist must characterize the lesion. Shape (Paris classification), size, and surface pit pattern together predict histology, resectability, and — critically — the baseline risk of bleeding once the polyp is removed.
The Paris classification describes superficial gastrointestinal neoplasms by gross morphology: 0-Ip (pedunculated, on a stalk), 0-Is (sessile, broad-based mound), 0-IIa (flat elevated), 0-IIb (flat), 0-IIc (flat depressed), and 0-III (excavated).
Morphology directly predicts the vascular anatomy the endoscopist will encounter. Pedunculated (0-Ip) polyps often contain a discrete feeding artery running up the stalk — sometimes visible as a central vessel up to 1–2 mm in diameter — that can bleed briskly if transected without preparation. Sessile and flat lesions (0-Is, 0-IIa) instead have a diffuse submucosal capillary and venular plexus spread across the base, producing more of an oozing pattern than a discrete spurt.
Because stalk vessels in large pedunculated polyps can approach arterial caliber, thick stalks (>10 mm) are frequently treated with a detachable snare loop (endoloop) or through-the-scope clip placed at the base before resection, mechanically occluding the feeding vessel in advance.
Size is the single strongest predictor of both malignant potential and bleeding risk, and is estimated against a reference object (open biopsy forceps ~7 mm, open snare) rather than by eye alone, since visual estimation is notoriously inaccurate and tends to under- or over-call size by several millimetres.
Clinically meaningful size bands: diminutive (≤5 mm), small (6–9 mm), large (10–19 mm), and giant (≥20 mm). Diminutive and small polyps are removed with minimal bleeding risk regardless of technique. Once a lesion exceeds 10 mm, submucosal vessel caliber increases and thermal or piecemeal resection becomes more likely — both of which raise bleeding risk. Above 20 mm, immediate bleeding rates can approach 8–10% and delayed bleeding rates 6–10% in some series, especially in the right colon.
Narrow-band imaging (NBI) and the NICE/Kudo classifications let the endoscopist predict histology in real time — distinguishing hyperplastic from adenomatous or invasive-appearing surface pit patterns — before committing to a resection strategy.
This assessment feeds directly into technique selection, the subject of Stage 3: small, low-risk adenomas are strong candidates for cold snare polypectomy, while large or morphologically complex lesions are triaged toward hot snare endoscopic mucosal resection (EMR), piecemeal technique, or referral for advanced endoscopic or surgical resection if deep submucosal invasion is suspected.
A wire snare is advanced through the colonoscope working channel, opened around the target, and closed to capture a defined ring of tissue. Precise placement — enough margin for complete resection, not so much that muscularis propria is captured — is the mechanical foundation of a safe polypectomy.
The snare — a braided or monofilament wire loop on a sheathed catheter — exits the accessory channel (typically 3.7 mm internal diameter), is opened wider than the lesion, and lowered over it. The endoscopist then closes the loop circumferentially at the base, aiming to capture the polyp plus a 1–2 mm rim of normal mucosa.
Capturing too little tissue risks an incomplete resection with residual adenoma; capturing too much — particularly pulling deep into the submucosa or muscularis propria — raises the risk of both perforation and, because larger submucosal vessels live deeper in the wall, more significant bleeding. Gentle "tenting" of the tissue away from the bowel wall before full closure is a key safety maneuver, especially in the thin-walled right colon.
For flat or sessile lesions, especially those ≥10–15 mm, a fluid cushion is injected into the submucosa beneath the lesion before snaring — classic endoscopic mucosal resection (EMR). The injectate is typically normal saline mixed with dilute adrenaline (around 1:100,000–1:200,000) and a marker dye such as indigo carmine or methylene blue to outline lesion margins.
The cushion lifts the mucosa and lesion away from the muscularis propria, both reducing perforation risk and making the tissue easier to ensnare completely. Failure of a lesion to lift ("non-lifting sign") is a red flag for submucosal fibrosis or invasive cancer and should prompt reconsideration of endoscopic resection.
Dilute adrenaline in the lifting solution provides a degree of local vasoconstriction at the injection site, but its effect on reducing delayed bleeding is modest — mechanical and thermal hemostasis at the time of resection matter far more.
Whenever feasible, en bloc (single-piece) resection is preferred: it yields a complete specimen for histopathologic margin assessment and carries a low local recurrence rate. This is realistic for lesions under roughly 20 mm.
Larger sessile lesions often require piecemeal EMR (EPMR) — sequential snare captures of the same lesion. Piecemeal resection sacrifices some margin certainty and is associated with meaningfully higher local recurrence at first surveillance colonoscopy (historically on the order of 10–20%, though thorough margin treatment with snare-tip soft coagulation has been shown to cut this substantially). Piecemeal technique does not by itself raise bleeding risk as much as it raises recurrence risk, but the larger total resection area does increase the number of transected vessels the endoscopist must manage.
Whether — and how — electrical current is applied during transection is the single most consequential technical decision in this simulation. Cold snare relies on pure mechanical shear; hot snare uses electrosurgical current whose waveform (cut, coagulation, or blended) trades immediate hemostasis against deeper thermal injury and delayed bleeding.
A cold snare — a specifically thin, mechanically efficient monofilament wire loop — is closed around the polyp base with no electrical current applied. Tissue is transected by pure mechanical shear at the submucosal level, capturing the lesion plus a small margin of normal mucosa and shallow submucosa, typically in under a second.
Because colonic submucosal vessels in this size range are usually well under 1 mm in caliber, mechanical transection alone is generally sufficient: cut vessel ends retract and thrombose spontaneously, much as a shallow skin nick stops bleeding on its own. Visible immediate oozing is actually somewhat more common with cold technique than hot (roughly 2–5% of cases show transient bleeding at the site), but it is almost always self-limited within seconds to a couple of minutes and rarely requires any intervention.
Because no thermal energy crosses the bowel wall, cold snare polypectomy carries a near-zero risk of the delayed, thermally-mediated bleeding and perforation that can follow hot snare resection — the central reason professional societies now favor it as first-line for polyps under 10 mm, with expanding use up to 15–19 mm.
When current is applied, the electrosurgical generator can deliver several waveform types: pure cutting current produces rapid tissue vaporization and a fast, clean transection with minimal coagulation depth; pure coagulation current desiccates tissue more slowly, denaturing vessel walls for strong immediate hemostasis but at the cost of a broader and deeper zone of thermal spread; blended current (alternating cut and coagulation, e.g. an automated "Endocut"-style mode) is the most widely used compromise, balancing immediate hemostasis against controlled, shallower thermal injury while also reducing tissue sticking to the snare wire during transection.
Generator power (wattage) and the duration of current application both scale directly with the depth of thermal spread — longer or higher-power coagulation bites reach deeper into the bowel wall.
Electrocautery coagulates submucosal vessels at the time of resection, sealing the cut ends and largely preventing immediate bleeding. But that seal is a thermally denatured eschar sitting over the vessel stump, not a permanent repair. As the resection site heals over the following days, the eschar can slough away — typically around day 5–7 — re-exposing the previously sealed vessel and producing classic delayed post-polypectomy bleeding.
Excessive coagulation carries its own separate hazard: if thermal injury extends through the full thickness of the bowel wall without frank perforation, it can produce post-polypectomy electrocoagulation syndrome — localized peritonism, fever, and leukocytosis mimicking early perforation, occurring in roughly 0.5–1% of hot snare/EMR procedures, usually managed conservatively with bowel rest and antibiotics rather than surgery.
| Product | Indication | Trial Design | Key Result |
|---|---|---|---|
| Cold snare (no current) | Mechanical guillotine shear | Vessels retract & thrombose spontaneously; no thermal seal | Near-zero delayed bleeding & perforation |
| Pure cutting current | Rapid tissue vaporization | Minimal coagulation depth, fast transection | Low deep thermal injury; slightly more immediate ooze |
| Blended (cut + coag) | Alternating automated cycling | Balances hemostasis with controlled thermal depth | Most widely used default for hot snare |
| Pure coagulation current | Slow tissue desiccation | Deep, broad thermal spread; strong immediate seal | Best immediate hemostasis, highest delayed-bleed/perforation risk |
The instant the specimen separates, the endoscopist watches the defect for bleeding. Immediate bleeding is managed on the spot — with clips, dilute adrenaline, or targeted coagulation — before the scope is withdrawn, and this real-time response strongly shapes the overall complication rate reported for a procedure.
Immediate (intraprocedural) bleeding is bleeding observed during the procedure itself, before the scope is withdrawn. It occurs in roughly 1–2% of polypectomies overall, but rises to 8–10% for large (>20 mm) or thick-stalked pedunculated polyps.
Two patterns dominate: arterial spurting, from an unsealed feeding vessel — most classically a thick pedunculated stalk containing a true feeding artery up to 1–2 mm in diameter — and venous or capillary oozing, from the diffuse submucosal plexus of a sessile or flat lesion base. Spurting bleeds are visually dramatic but usually respond quickly to targeted therapy; oozing bleeds are often self-limited but can be harder to localize precisely.
Through-the-scope mechanical clips directly appose or compress the bleeding point or the cut edges of the resection site; they are the preferred first response to a discrete spurting vessel and are also used prophylactically (see Stage 5).
Injection of dilute adrenaline (typically 1:10,000 to 1:20,000) into the stalk or surrounding mucosa causes local vasoconstriction and tissue tamponade — a fast temporizing measure, often used pre-emptively before resecting a large pedunculated polyp, though as a stand-alone measure it does not meaningfully reduce later delayed bleeding.
Targeted thermal coagulation — with coagulation forceps or gentle "soft coagulation" via the snare tip — seals a visible vessel stump under direct vision. For very thick pedunculated stalks (>10 mm), a detachable nylon loop (endoloop) can be placed around the stalk before resection, mechanically ligating the feeding artery in advance rather than relying on cautery alone.
Combination therapy — clip plus injection, or clip plus coagulation — controls the large majority of immediate polypectomy bleeds without need for hospital admission, blood transfusion, or interventional radiology.
Beyond treating active bleeding, clips can be placed prophylactically to close the mucosal defect after resection of large polyps, even when no bleeding is occurring at the time. The landmark multicenter randomized trial (Pohl et al., JAMA 2019, ~919 patients with non-pedunculated polyps ≥20 mm) found that prophylactic clipping significantly reduced postpolypectomy bleeding for proximal (right-sided, cecum-to-transverse) lesions — roughly 7% down to 3% — but showed no significant benefit for distal, left-sided lesions.
Current guidance from ASGE and ESGE therefore suggests considering prophylactic clip closure preferentially for large, right-sided resection defects, particularly in patients who will resume antithrombotic therapy soon after the procedure — rather than applying it routinely to every polypectomy.
Delayed bleeding happens after the patient has left the endoscopy unit — from hours out to roughly two weeks later — as thermally sealed vessels are re-exposed by sloughing eschar. Size, location, technique, and antithrombotic therapy combine to stratify an individual patient's risk, and that stratification drives both technique choice and post-procedure counseling.
Polyp size is the strongest single predictor — risk climbs progressively above 10 mm and sharply above 20 mm. Location matters almost as much: the right colon (cecum, ascending, and transverse colon) has a thinner muscularis propria and less overlying protective musculature than the left colon and rectum, and carries roughly 2–3 times the delayed bleeding risk of an equivalent left-sided lesion.
Resection technique is directly modifiable: hot snare / hot EMR with coagulation current carries substantially higher delayed bleeding risk than cold snare technique for comparable lesions. Patient-level factors add further risk — anticoagulant therapy confers the highest incremental risk, followed by dual antiplatelet therapy and then single antiplatelet therapy (low-dose aspirin is usually continued through the procedure). Chronic kidney disease, cardiovascular comorbidity, and villous histology are secondary contributors.
Because polypectomy is classified as a high-bleeding-risk procedure, society guidelines (ASGE 2016; ESGE 2021 update) weigh procedural bleeding risk against each patient's individual thrombotic risk before deciding how to manage anticoagulants and antiplatelets.
Low-dose aspirin is generally continued throughout. P2Y12 inhibitors such as clopidogrel are typically held for about 5–7 days beforehand when it is safe to do so (in consultation with cardiology for patients with recent coronary stents). Warfarin is usually held roughly 5 days pre-procedure, with bridging anticoagulation reserved for high thrombotic-risk patients, and is resumed the same or next day. Direct oral anticoagulants (apixaban, rivaroxaban, dabigatran, edoxaban) are held for one to a few doses (roughly 24–48 hours, longer with renal impairment) beforehand, and resumed anywhere from 24 hours to about a week afterward depending on the size of the resection and the assessed bleeding risk of the defect.
Timing of anticoagulant resumption is a direct clinical lever on delayed bleeding risk: resuming too early, especially after a large right-sided hot resection, meaningfully raises the chance of bleeding during the day 5–7 eschar-slough window.
Delayed bleeding is defined as bleeding after the patient leaves the endoscopy unit, occurring anywhere from hours to about 30 days later, though the great majority present within the first week and incidence peaks around day 5–7 — matching the time course of eschar separation over a thermally sealed vessel.
It typically presents as hematochezia (passage of blood per rectum), occasionally with hemodynamic instability in a minority of more severe cases. Management begins with resuscitation as needed, followed by repeat colonoscopy: endoscopic hemostasis (clips, thermal coagulation, or injection at the bleeding site) is successful in over 90% of cases. Surgery or interventional radiologic embolization is required only rarely, when endoscopic control fails or bleeding is massive.
Kept in perspective, polypectomy remains an extremely safe procedure: colonoscopy with polypectomy is performed millions of times per year worldwide, procedure-related mortality is well under 0.1%, and the great majority of both immediate and delayed bleeding events are fully managed endoscopically without lasting harm.
The practical prevention strategy that emerges from the evidence reviewed across all five stages: prefer cold snare technique for polyps under roughly 10–15 mm; when hot snare is necessary for larger lesions, favor blended over pure coagulation current where hemostasis needs allow; consider prophylactic clip closure for large, right-sided resection defects; manage antithrombotic therapy according to guideline-based bridging and resumption timing; and counsel each patient on their individualized bleeding risk, built from the combination of size, location, technique, and their own medication profile.
| Product | Indication | Trial Design | Key Result |
|---|---|---|---|
| Low risk | <10 mm, cold snare, either location | No thermal injury, small vessel caliber | ~0.1–0.5% delayed bleeding |
| Moderate risk | 10–19 mm, hot blended, left colon | Some thermal spread, thicker-walled segment | ~1–2% delayed bleeding |
| High risk | ≥20 mm, hot snare/EMR, right colon | Deep thermal injury + thin wall + large area | ~6–10% delayed bleeding |
| Highest risk | ≥20 mm right-sided + early anticoagulant resumption | Combines all major risk factors | Prophylactic clipping strongly considered |