Prophylactic hemoclips halved delayed bleeding after thermal polyp resection

Design

Systematic review and meta-analysis of randomised trials with trial sequential analysis

Population

10 RCTs, 4,588 patients after thermal resection of nonpedunculated colorectal polyps of 5 mm or more

Primary outcome

Delayed post-polypectomy bleeding

Effect

3.4% vs 6.5%; RR 0.51 (95% CI 0.34 to 0.76)

A meta-analysis pooled 10 randomised trials with 4,588 patients comparing prophylactic hemoclip closure with none after thermal resection of nonpedunculated colorectal polyps of 5 mm or more. Delayed post-polypectomy bleeding occurred in 3.4% with clips against 6.5% without (RR 0.51, 95% CI 0.34 to 0.76; I-squared 45%; GRADE high certainty). For proximal polyps of 20 mm or more the rates were 4.4% and 8.1% (RR 0.54, 0.32 to 0.91; moderate certainty). Trial sequential analysis supported both findings, and other outcomes did not differ.

The authors note that patients on antithrombotic drugs may gain more in absolute terms because their baseline bleeding risk is higher. Subgroup results by histology were exploratory because interaction tests were not significant.

The relative reduction is clear; the absolute gain is about three percentage points. Clip use is worth weighing against lesion size and site, antithrombotic use, cost and availability, particularly where resources are limited.

• Consider prophylactic clips after thermal resection of polyps of 5 mm or more, especially large proximal lesions.
• Weigh antithrombotic use and other bleeding risk factors when deciding.
• Absolute risk fell from about 6.5% to 3.4%, so most patients would not have bled either way.
• Document the number of clips and advise patients on delayed bleeding symptoms.
• Subgroup findings by histology are exploratory.

Why it matters

It gives a high-certainty answer to a question that previous, smaller trials left open.

Don't overread it

The pooled trials concerned thermal resection only, and the histology subgroup findings are exploratory.

The statistics, in plain English

A risk ratio of 0.51 means bleeding was about half as likely with clips; the confidence interval of 0.34 to 0.76 excludes no effect. Moderate heterogeneity (I-squared 45%) means trials differed somewhat in technique and population, but the direction was consistent.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.