Evaluation of the laparoscopic Whipple procedure

Abstract

In the Netherlands, approximately 1.750 people are diagnosed with pancreatic cancer each year. Approximately 20% of these new patients are operated with curative intent. This surgical treatment, called pancreaticoduodenectomy leads to a five-year survival rate of 5-10%. Laparoscopic pancreaticoduodenectomy (LPD) is a complex surgical procedure that pancreas surgeons have only started to adopt. In the Netherlands, the four pioneering hospitals in laparoscopic pancreatic surgery (Catharina Ziekenhuis (CTZ), Jeroen Bosch Ziekenhuis (JBZ), Amsterdam Medisch Centrum (AMC) and Onze Lieve Vrouwe Gasthuis (OLVG)) are still in the beginning of their learning curve. An early assessment of their operative technique could provide deep insight in the differences of the operative method between the hospitals. This is an experimental study to evaluate the LPD procedure with the aim to improve the operative techniques of the four pioneering hospitals in the Netherlands. In each of the four hospitals, the steps of the LPD procedure were defined based on the operation report. For each step of the procedure the risks and risk scores were identified. This was done with a multidisciplinary team per hospital according to the hazard analysis of the Health Failure Mode and Effect Analysis (HFMEA). In consultation with one surgeon of each of the four hospitals, the risks were converged to relevant risks s according to the adapted decision tree of the HFMEA method. The similar steps of the four hospitals were linked to each other to create an overview of the differences and similarities in process and risks. Finally, risk types and the corresponding causes were identified based on the relevant risks. In all included hospitals, relevant risks were found (CTZ: n = 10, JBZ: n = 16, AMC: n = 3 and OLVG: n = 13). The process steps which contained relevant risks by more than one surgical teams were (1) performing Kocher maneuver and exposing ligament of Treitz, (2) cholecystectomy, (3) mobilising portal vein, superior mesenteric vein and artery, (4) transection gastroduodenal artery, (5) pancreatojejunostomy (PJ), (6) hepaticojejunostomy (HJ) and (7) gastrojejunostomy (GJ). Eight out of 41 relevant risks were accepted by the surgeon of the corresponding hospital (no further action is warranted to diminish the risk). The remaining relevant risks were bleeding (n = 23), HJ failure (n = 4), PJ failure (n = 3) and GJ failure (n = 3). The prevalent reason for bleeding was unable to view or identify anatomical structures of the patient (33%). HJ and PJ failure originated from patient’s habitus and iatrogenic/operative technique. GJ failure originated from iatrogenic/operative technique solely. The HFMEA method provided an overview of the practiced operative methods during the LPD procedure for each of the included hospitals with detailed information of the surgical steps. There are clear differences in the order of several surgical steps between the four hospitals. This information could be used by the surgeons to learn from each other by sharing their considerations and knowledge about specific process steps.