Implementation of mentor-assisted colorectal endoscopic submucosal dissection in Sweden; learning curve and clinical outcomes.

OBJECTIVE

It's still challenging to introduce colorectal (CR) ESD in Western countries. We assessed the feasibility of introducing and implementing CR-ESD in Sweden with hiring Japanese expert as a supervisor.

METHODS

We analyzed 71 consecutive CR-ESD cases performed by two endoscopists who had no (endoscopist A (E-A)) or 20 cases (endoscopist B (E-B)) of experience in ESD. E-A performed rectal lesions while E-B performed lesions in any locations. Factors associated with failure in en bloc resection and in self-accomplishment were analyzed.

RESULTS

Overall en bloc and R0 resection rates were 80.3% and 70.4%. Adverse event occurred in 7.0% including two perforations, two post-operative hemorrhage and one delayed perforation. Only case with delayed perforation underwent surgical treatment. Total self-accomplishment rate was 50% (10/20) for E-A, and 37.3% (19/51) for E-B. Dividing each performer's cases into three learning phases, self-accomplishment rates increased from 42.9% to 83.3% for E-A, and from 29.4% to 70.6% for E-B, as well as en bloc resection rates from 71.4% to 100% for E-A, and from 52.9% to 94.1% for E-B. Multivariate analysis revealed that location upper than rectum, lesions with formerly taken biopsy and lesions larger than 30mm were significantly associated with en bloc resection failure.

CONCLUSIONS

Implementation of CR-ESD with hiring Japanese supervisor for certain period was safe for patients and effective for good learning curve.