The effect of prophylactic hemoclips on the risk of delayed post

The practice of prophylactic application of hemoclips for the prevention of delayed bleeding is increasing [30]. However, placement of hemoclips has been associated with mixed results in preventing delayed EMR bleeding and there are no uniform guidelines to support its routine application for this purpose. Furthermore, the cost of hemoclips are rather substantial with price range of $ 125–277 per clip in the current market (average cost for the four different models of clips used for this study cohort was $ 193), and this does not include added time spent to perform clip application thus raising concerns about the cost-effectiveness of routine prophylactic clip placement [25, 31]. Parikh et al. searched existing published data on post-polypectomy bleeding rates and performed a decision analysis to examine the cost-effectiveness of routine, prophylactic clip placement after colon polypectomy [31]. The researchers determined that prophylactic clip placement seems to be a cost-effective strategy in patients who are taking anti-platelet or anti-coagulation medications at the time of the procedure. However, for those patients not receiving anti-platelet or anti-coagulation therapy, placement of prophylactic clips after polypectomy was not cost-effective, particularly for smaller sized polyps [31]. Dokoshi et al. conducted a randomized study on the effectiveness of prophylactic clipping during endoscopic resection of colon polyps for the prevention of delayed bleeding [26]. They concluded that there was no difference in delayed bleeding rates between patients who received prophylactic clipping and those who did not. In this particular study, anticoagulant treatment made no difference between the clip and non-clip group in delayed bleeding rates provided the anticoagulant was stopped prior to the procedure [26]. Mori et al. carried out a prospective trial comparing snare cauterization to clip closure for the prevention of post-EMR delayed bleed. They concluded that snare cauterization was superior to clip closure in terms of procedure time and medical costs. Moreover, their findings showed no difference in delayed bleeding between the two groups [27].

Our study is a large retrospective cohort study that examines the effect of hemoclip application in preventing delayed post-EMR bleed for both upper and lower GI tract lesions. There was no statistically significant association between delayed post-EMR bleeding and use of prophylactic hemoclips. The total cost of clips for prophylactic purposes was $173,893 while there was no statistically significant difference between clip use and several clinically-relevant variables (hospital length of stay, need for repeat procedure at the time of rebleed, and blood transfusion requirement) that have the potential to accrue additional medical cost, meaning the amount listed here was the majority of the direct cost of prophylactic clip usage. Given these findings, one can then question if there is no place for prophylactic hemoclip placement at all in the setting of upper/lower EMR or certain subgroups within these EMR procedures. When we performed analyses on delayed bleeding rate in the context of different variables without taking clipping into consideration, delayed bleed did occur more frequently in those cases involving larger lesions (whether the lesion size was treated as a continuous variable or dichotomized as  90% of the cases were proximal polyps, 51% were > 40 mm, and 36% of the patients were on antiplatelets [33]. Intention to treat (ITT) analysis showed a trend for lower DB in clip group. In per protocol (PP) analysis, DB was significantly lower in complete clip closure group (as opposed to the partial closure subgroup and the failed closure subgroup). However, these polyps that were successfully clipped were also smaller, had better accessibility, and in shorter and easier EMR procedures. Contrary to the result of these two studies mentioned, the study by Feagins et al. showed no statistical difference in BD between hemoclip group and no hemoclip group in their cohort of 1098 patients [30]. As opposed to the Pohl study, they found no difference in bleed rates between proximal polyps that underwent prophylactic hemoclip placement versus no prophylactic hemoclip. While these recent RCTs have overcome one of the limitations of our study, its retrospective design, our study still has an advantage since it includes both upper GI and lower GI/colon lesions. Also, unlike the study by Albeniz et al., which included only high-risk lesions, our study included all lesions regardless of their risk status, and this may potentially make our study more generalizable for all EMRs. The Feagins study supports the findings of our results that the prophylactic placement of hemoclips does not affect the proportion of DB, also questioning the widespread, expensive practice of routinely placing prophylactic hemoclips after polyp resection.