Etiology and mortality in severe acute pancreatitis: A multicenter study in Japan
Hideto Yasuda,Masayasu Horibe,Masamitsu Sanui,Mitsuhito Sasaki,Naoya Suzuki,Hirotaka Sawano,Takashi Goto,Tsukasa Ikeura,Tsuyoshi Takeda,Takuya Oda,Yuki Ogura,Dai Miyazaki,Katsuya Kitamura,Nobutaka Chiba,Tetsu Ozaki,Takahiro Yamashita,Toshitaka Koinuma,Taku Oshima,Tomonori Yamamoto,Morihisa Hirota,Mizuki Sato,Kyohei Miyamoto,Tetsuya Mine,Takuyo Misumi,Yuki Takeda,Eisuke Iwasaki,Takanori Kanai,Toshihiko Mayumi
DOI: https://doi.org/10.1016/j.pan.2020.03.001
IF: 3.977
Pancreatology
Abstract:Background/objectives: Severe acute pancreatitis (SAP) has a high mortality rate despite ongoing attempts to improve prognosis through a various therapeutic modalities. This study aimed to delineate etiology-based routes that may guide clinical decisions for the treatment of SAP. Methods: Using data from a recent retrospective multicenter study in Japan, we analyzed the association between clinical outcomes, mainly in-hospital mortality and pancreatic infection, and various etiologies while considering confounding factors. We performed additional multivariate analyses and built decision tree models. Results: The 1097 participating patients were classified into the following groups by etiology: alcohol (n = 436, 39.7%); cholelithiasis (n = 230, 21.0%); idiopathic (n = 227, 20.7%); and others (n = 204, 18.6%). Mortality at hospital discharge was 8.4%, 12.2%, 16.7%, and 16.2% in the alcohol, cholelithiasis, idiopathic, and others groups, respectively. According to multivariable analysis, early enteral nutrition (EN) was significantly associated with reduced in-hospital mortality only in the cholelithiasis group. However, there was a consistent association between age and the need for mechanical ventilation and increased mortality, regardless of etiology. Our decision tree models presented different contributing factors depending on the etiology and patient background. Interaction analysis showed that EN and the use of prophylactic antibiotics may influence these results differently according to etiology. Conclusions: No study has yet used comprehensive models to investigate etiology-related prognostic factors for SAP; our results can, therefore, be used as a reference for improving clinical decisions.