ABSTRACT
INTRODUCTION: Colorectal endoscopic submucosal dissection is not a widely adopted procedure due to its technical difficulties. This study aimed to share the experience in setting up this novel procedure and to report the learning curve for such a procedure at a low-volume district hospital in Hong Kong. METHODS: This case series comprised 71 colorectal endoscopic submucosal dissections that were performed by a single endoscopist without experience in gastric or colorectal endoscopic submucosal dissection. Lesion characteristics, procedure time per unit area of tumour, en-bloc resection rate, R0 resection rate, complications, and length of stay were recorded prospectively. Results were compared for two consecutive periods to study the learning curve. RESULTS: Overall, 41 (57.7%) tumours were located in the right colon, 21 (29.6%) in the left colon, and nine (12.7%) in the rectum. The median tumour area was 4 cm(2) (range, 0.25-16 cm(2)). The median operating time was 105 (range, 47-342) minutes. The median procedure time per unit area of tumour was 24.9 min/cm(2). There was one instance of intra-operative bleeding that required conversion to laparoscopic colectomy. There was no postoperative haemorrhage. The overall perforation rate was 15.5%, in which one required conversion to laparoscopic colectomy. The overall morbidity rate was 16.9% and there was no mortality. The median hospital stay was 1 day (range, 0-11 days). The overall en-bloc resection rate and R0 resection rate was 81.2% and 58.0%, respectively. Comparison of the two study periods revealed that procedure time per unit area of tumour decreased significantly from 31.5 min/cm(2) to 21.5 min/cm(2) (P=0.032). The en-bloc resection rate improved from 78.8% to 83.3% (P=0.15). The R0 resection rate improved significantly from 39.4% to 75.0% (P<0.01). CONCLUSION: Untutored colorectal endoscopic submucosal dissection is feasible with acceptable clinical outcomes at a low-volume district hospital in Hong Kong.
