Zhifei Wang1, Dongsheng Huang1, Junwei Liu1, Defei Hong, MD, FACS1, Minen Xu2, Xuecong Lai2. 1Zhejiang Provincial People’s Hospital, 2Hangzhou Dianzi University
To introduce a series of dry lab models simulating the pancreatojejunostomy (PJ) for da vinci robotic surgery. 3 types of dry lab models were designed and applied for the training for surgeon for the pancreatic reconstruction surgery under da vinci surgical robot, ie. model 1#, for the anastomosis of the commercially available pancreatic duct stent made of silicon gal ;model 2#, the pancreatic duct stent was inserted into a silicon gal model from Ethicon Company as to simulate the PJ of duct to mocusa .The procedure simulating a typical duct to mucosa pancreatojejunostomy; in model 3#, a human pancreatic stump was produced with silicon gal according to the data of a thin slice CT scan with pancreatic duct. The procedure required was same as for model 2#. Three surgeons, A,B and C,with similar background of laparoscopic surgery and open surgery experience were enrolled in the training program. After the first round of warm up training for robotic surgery with conventional tasks. Surgeon A and B entered the second round of training on model 1# using da vinic surgical robotic(Si) with end-to-end anastomosis task for a total of 6 hours. The times of practicing before reaching a perfect anastomosis was counted and compared. Then surgeon A entered the third round of training with 2# under da vinci for totally 6 hours, surgeon C did not received any additional training but the first round. Then the three surgeons were examined for their proficiency for robotic reconstruction techniques on dry lab model 3# which includes the time consuming for the task, the security of the three anastomosis. The procedure was evaluated by an independent expertized surgeon of robotic surgery. Score ranged from A to D with A the best. Also, for each surgeon, the learning curve for each  the same records, ie the time consuming and the evaluation were applied to the first case for each dry lab model and the last one, which were also compared. Results There was no significant difference between the two surgeons for completing consecutive 3 times of satisfactory pancreatic stent anasotomosis ( 8 and 9), the time for completing a perfect anasotomosis of pancreatic stent was similar ( 31 to 33min). On the 3# dry lab model, the times of attempts of training before a consecutively 3 times of satisfactory anastomosis were achieved were counted and compared, for surgeon A, 6 cases, the cases to reach a satisfactory performance similar to A was 20 to 25 cases for surgeon B and surgeon C, respectively. Under such performance, the examination was all satisfactory. For each surgeon with the same task, the time consuming decreased significantly with the training, similarly, the security of the anastomosis improved. The coordination of the three robotic arms and the smoothness of the procedure increased too. Conclusions The specifically designed series of dry lab training models are rated favorably for both anatomic realism and potential as a training tool for advancing the robotic pancreatojejunostomy through quality improvement exercise in dry lab.