BMC Gastroenterology (Mar 2020)

The effect of prophylactic hemoclips on the risk of delayed post-endoscopic mucosal resection bleed for upper and lower gastrointestinal lesions: a retrospective cohort study

  • Karen Chang,
  • Brian S. Lee,
  • Timnit Tekeste,
  • Andrew Nguyen,
  • Mopelola Adeyemo,
  • Agathon Girgis,
  • Karl K. Kwok,
  • H. Michael Crowson,
  • Alicia O. Burris,
  • Rajeev Attam,
  • Charles T. Chaya,
  • Theodore E. Durbin,
  • Andrew Q. Giap,
  • Gordon C. Hunt,
  • John Iskander,
  • Kevin T. Kao,
  • Brian S. Lim

DOI
https://doi.org/10.1186/s12876-020-01199-x
Journal volume & issue
Vol. 20, no. 1
pp. 1 – 10

Abstract

Read online

Abstract Background Endoscopic mucosal resection (EMR) is a minimally invasive procedure used for the treatment of lesions in the gastrointestinal (GI) tract. There is increased usage of hemoclips during EMR for the prevention of delayed bleeding. This study aimed to evaluate the effect of hemoclips in the prevention of delayed bleeding after EMR of upper and lower GI tract lesions. Method This is a retrospective cohort study using the Kaiser Permanente Southern California (KPSC) EMR registry. Lesions in upper and lower GI tracts that underwent EMR between January 2012 and December 2015 were analyzed. Rates of delayed bleeding were compared between the hemoclip and no-hemoclip groups. Analysis was stratified by upper GI and lower GI lesions. Lower GI group was further stratified by right and left colon. We examined the relationship between clip use and several clinically-relevant variables among the patients who exhibited delayed bleeding. Furthermore, we explored possible procedure-level and endoscopist-level characteristics that may be associated with clip usage. Results A total of 18 out of 657 lesions (2.7%) resulted in delayed bleeding: 7 (1.1%) in hemoclip group and 11 (1.7%) in no-hemoclip group (p = 0.204). There was no evidence that clip use moderated the effects of the lesion size (p = 0.954) or lesion location (p = 0.997) on the likelihood of delayed bleed. In the lower GI subgroup, clip application did not alter the effect of polyp location (right versus left colon) on the likelihood of delayed bleed (p = 0.951). Logistic regression analyses showed that the clip use did not modify the likelihood of delayed bleeding as related to the following variables: use of aspirin/NSAIDs/anti-coagulants/anti-platelets, pathologic diagnoses (including different types of colon polypoid lesions), ablation, piecemeal resection. The total number of clips used was 901 at a minimum additional cost of $173,893. Conclusion Prophylactic hemoclip application did not reduce delayed post-EMR bleed for upper and lower GI lesions in this retrospective study performed in a large-scale community practice setting. Routine prophylactic hemoclip application during EMR may lead to significantly higher healthcare cost without a clear clinical benefit.

Keywords