Introduction

In intraoperative bleeding during endoscopic submucosal dissection (ESD) by using a needle-tip type electrocautery knife, spray coagulation is considered helpful compared to swift coagulation. In this study, we compared the hemostatic effect of two coagulation modes (swift coagulation and spray coagulation) on intraoperative bleeding during gastric ESD in a propensity score analysis.

Aims & Methods

In an intraoperative bleeding strategy, we changed “swift-dependent” hemostasis to “spray-switching” one in 2021. Therefore, 202 bleeding events before April 2020 (33 lesions, Swift group) and 203 bleeding events after April 2022(36 lesions, Spray group) were consecutively collected. In the Swift group, hemostasis was attempted firstly by swift coagulation by using a retracted tip of the needle-type knife, and if hemostasis was not achieved, hemostatic forceps were used. In the Spray group, when hemostasis was attempted several times by swift coagulation but hemostasis was not achieved or when hemostasis was not expected by swift coagulation, spray coagulation was used. If hemostasis was not achieved by spray coagulation, hemostatic forceps were finally used. The two coagulation modes were determined by one endoscopist who retrospective observed the difference of sparks that occurred in hemostasis on video. We performed a post-hoc propensity score-matched analysis to compare the two groups.

Results

In comparison of patient background (age, gender, whether or not intraoperative antithrombotic medication was taken) and lesion background (site of lesion occupation, resection diameter, depth, presence or absence of ulcer, histological type, etc.) between the two groups, the larger number of patients in the Swift group had antithrombotic agents on ESD (p=0.016), while no significant differences were observed in other parameters. A propensity score matching was performed in defining intraoperative administration of antithrombotic agents, location, resection size, pathology, and presence or absence of ulcer as covariates. A total of 106 events were matched per event in the two group, respectively. There was no significant difference in the hemostatic duration per event in the two group (Swift vs. Spray: 48.2 sec. vs. 55.3 sec., p=0.22). The number of coagulations per event in the Swift group was less than that of Spray group (4.3 times vs. 5.8 times, p=0.008). The hemostatic forceps was significantly less frequently used in the Spray group than in the Swift group (31.1% vs. 17.0%, p=0.016). The cumulative success rate of hemostasis with electrocautery knives was likely to be higher in the Spray group (p=0.104), although there was no significant difference between the two groups. In the Swift group, two patients had postoperative hemorrhage and one patient had delayed perforation, whereas no adverse events were observed in the Spray group.

Conclusion

The data suggested that spray coagulation by the tip of the needle-type knife could reduce the use of hemostatic forceps. In gastric ESD, spray coagulation may facilitate the hemostasis of intraoperative bleeding.