Scholarlink

Background <p>Guidelines recommend cold snare polypectomy (CSP) for colorectal polyps 10&#xa0;mm or smaller, but data for larger polyps are limited. We compared delayed post-polypectomy bleeding (DPPB) and resection outcomes after CSP versus conventional endoscopic mucosal resection (EMR) for polyps 15&#xa0;mm or smaller.</p> Methods <p>We retrospectively analysed 2,673 polyps from 2,083 patients undergoing screening or surveillance colonoscopy between January 2022 and December 2025 (CSP, 1,726; EMR, 947). DPPB was defined as post-discharge bleeding requiring an emergency department visit, hospitalization, transfusion or repeat endoscopy; every candidate event was audited against this definition. One-to-one propensity score matching on seven covariates yielded 652 pairs. Because no bleeding followed CSP (complete separation), the matched analysis was interpreted alongside full-cohort Firth penalized regression, inverse-probability weighting and a patient-level analysis. Follow-up was 30 days.</p> Results <p>Six DPPB events met the definition, all after EMR (6/947, 0.6%); four further post-discharge bleeds were self-limited and met no criterion. All six events were right-sided. In the matched cohort DPPB occurred in 0/652 versus 2/652 (0.3%; <i>p</i> = 0.50) – underpowered. In the full cohort the difference was significant (0/1,726 versus 6/947; <i>p</i> = 0.002); Firth penalized regression gave an adjusted odds ratio of 14.6 for EMR (95% confidence interval 1.1–201; <i>p</i> = 0.045) – a very wide, imprecise estimate. Larger polyp size independently predicted bleeding (adjusted odds ratio 1.36 per mm; <i>p</i> = 0.002). At patient level, DPPB occurred in 0 of 1,258 CSP-only and 5 of 702 EMR-only patients (0.7%; <i>p</i> = 0.006); the sixth arose from an EMR polyp among 123 patients treated by both. En bloc resection with a tumour-free margin was similar (98.6% versus 99.7%; <i>p</i> = 0.064); CSP required less clipping (40.6% versus 54.8%) and shorter procedures (16.9 versus 20.9&#xa0;min; both <i>p</i> &lt; 0.001).</p> Conclusions <p>Delayed bleeding was infrequent, and none followed CSP. The matched comparison was underpowered, whereas full-cohort, adjusted, weighted and patient-level analyses consistently favoured CSP. Given the few events, the wide confidence interval and the retrospective single-centre design, these findings are hypothesis-generating rather than evidence of superiority and require prospective confirmation. Larger polyp size independently predicted bleeding, so caution is warranted for larger lesions.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Cold snare polypectomy versus conventional endoscopic mucosal resection for colorectal polyps 15 mm or smaller: delayed bleeding and resection outcomes in a retrospective cohort

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年8月4日

查看全文
  • Hyun Young Kim

摘要

Background

Guidelines recommend cold snare polypectomy (CSP) for colorectal polyps 10 mm or smaller, but data for larger polyps are limited. We compared delayed post-polypectomy bleeding (DPPB) and resection outcomes after CSP versus conventional endoscopic mucosal resection (EMR) for polyps 15 mm or smaller.

Methods

We retrospectively analysed 2,673 polyps from 2,083 patients undergoing screening or surveillance colonoscopy between January 2022 and December 2025 (CSP, 1,726; EMR, 947). DPPB was defined as post-discharge bleeding requiring an emergency department visit, hospitalization, transfusion or repeat endoscopy; every candidate event was audited against this definition. One-to-one propensity score matching on seven covariates yielded 652 pairs. Because no bleeding followed CSP (complete separation), the matched analysis was interpreted alongside full-cohort Firth penalized regression, inverse-probability weighting and a patient-level analysis. Follow-up was 30 days.

Results

Six DPPB events met the definition, all after EMR (6/947, 0.6%); four further post-discharge bleeds were self-limited and met no criterion. All six events were right-sided. In the matched cohort DPPB occurred in 0/652 versus 2/652 (0.3%; p = 0.50) – underpowered. In the full cohort the difference was significant (0/1,726 versus 6/947; p = 0.002); Firth penalized regression gave an adjusted odds ratio of 14.6 for EMR (95% confidence interval 1.1–201; p = 0.045) – a very wide, imprecise estimate. Larger polyp size independently predicted bleeding (adjusted odds ratio 1.36 per mm; p = 0.002). At patient level, DPPB occurred in 0 of 1,258 CSP-only and 5 of 702 EMR-only patients (0.7%; p = 0.006); the sixth arose from an EMR polyp among 123 patients treated by both. En bloc resection with a tumour-free margin was similar (98.6% versus 99.7%; p = 0.064); CSP required less clipping (40.6% versus 54.8%) and shorter procedures (16.9 versus 20.9 min; both p < 0.001).

Conclusions

Delayed bleeding was infrequent, and none followed CSP. The matched comparison was underpowered, whereas full-cohort, adjusted, weighted and patient-level analyses consistently favoured CSP. Given the few events, the wide confidence interval and the retrospective single-centre design, these findings are hypothesis-generating rather than evidence of superiority and require prospective confirmation. Larger polyp size independently predicted bleeding, so caution is warranted for larger lesions.