The selection of pancreatic reconstruction techniques gives rise to higher incidences of morbidity: results of the 30th Japan Pancreatic Surgery Questionnaire Survey on pancreatoduodenectomy in Japan

Background/Purpose We aimed to determine the impact of the surgical strategy used for pancreatic reconstruction on morbidity after pancreatoduodenectomy (PD). Methods A questionnaire was sent to all surgeon members of the Japan Pancreatic Surgery Club in December 2002. Results We received 152 replies, and the data from all of them were used in the analysis. Thirty‐six percent of the 152 responders performed PD and selected from among two or more pancreatic reconstruction techniques (PRTs). PRT selection was used no more frequently in the high‐ and medium‐hospital‐volume institutions than in low‐hospital‐volume institutions (25% or 37% vs 35%). The incidence of both “all arterial hemorrhage” and “delayed arterial hemorrhage” after PD in the institutions that used multiple PRTs was significantly higher than that in the institutions where only a single PRT was used (4.2% vs 2.2%, and 3.3% vs 1.5%, respectively; P < 0.05). In the high‐ and medium‐hospital‐volume institutions, the incidence of all arterial hemorrhage after PD in the multiple‐PRT institutions was significantly higher than that in the single‐PRT institutions (4.0% vs 1.9%; P < 0.05). Furthermore, in the low‐hospital‐volume institutions, the incidence of delayed arterial hemorrhage, 7 or more days after PD, was clearly higher in the multiple‐PRT institutions than in the single‐PRT institutions (4.1% vs 1.4%; P = 0.056). Therefore, the hospital‐case volumes of PD were distributed as practice‐case volumes according to the PRT by the selection of PRTs, and PRT selection gave rise to higher incidences of morbidity as a result of pancreatic leakage after PD. Conclusions The hospital‐case‐volume — better outcome relation for PD was attributable to expert pancreatic reconstruction skills that can be mastered only through frequent repetition.