Management of intraoperative hemorrhage during NOTES: a prospective, randomized comparison.

Background: The optimal strategy to manage intraoperative hemorrhage during NOTES is unknown. A randomized comparison of three instruments for hemorrhage control was performed [prototype endoscopic bipolar hemostasis forceps (BELA) vs. prototype endoscopic clip (E-CLIP) applier versus laparoscopic clip (L-CLIP) applier].

Methods: A hybrid transvaginal NOTES model in swine was used, with hemorrhage induced in either the gastroepiploic (GE) arteriovenous bundle (vessel diameter ~3 mm) or in distal mesenteric vessels (vessel diameter ~1-2 mm). Hemostasis was attempted three times per vessel using each instrument in a randomized order. Full laparoscopic salvage was performed if hemorrhage persisted beyond 10 min. Outcomes included primary success rate (PS), primary hemostasis time (PHT), number of device applications (DA), and overall hemostasis time (OHT, including salvage).

Results: Seventy hemostasis attempts were made in 12 swine. PS was 42-67 % for the GE vessels, with no difference between instruments. PHT and OHT also were similar between instruments, with the BELA and L-CLIP having a higher number of DA. PS was (80-100 %) in mesenteric vessels, with the BELA and L-CLIP resulting in a shorter mean PHT compared with the E-CLIP.

Conclusions: All three instruments had similar effectiveness in achieving primary hemostasis during hybrid NOTES. Management of small vessel bleeding (1-2 mm) in a porcine model is effective using all three instruments but may be most efficient with the BELA or L-CLIP. Large vessel bleeding (≥3 mm) may be best managed by adding laparoscopic ports for assistance while maintaining a low threshold for conversion to full laparoscopy.