Acute necrotizing pancreatitis is a frequent and potentially lethal disease, especially in case of infected pancreatic necrosis (IPN). IPN usually occurs after the first week of evolution. The step up approach is now widely recommended for the management of IPN. In fact, in case of suspicion of IPN, a drainage percutaneous or transgastric is recommended at first, supported by probabilist antibiotherapy. 1/3 of patients won't require any other interventions. For 2/3 of patients, an additional necrosectomy is necessary. Necrosectomy was formally realized by open laparotomy. Since de last decade, mini-invasive technics have emerged: transgastric necrosectomy, video-assist retroperitoneal debridement. laparoscopy and permitted a decreased of morbidity and mortality. Recently, Hollemans et al. developed a nomogram based on 4 variables (sex, multi-organ failure, % of necrosis and collections heterogeneity) which are negative predictors for success of catheter drainage in IPN with an receiver operating characteristic (ROC) curve at 0.76. The aim of this study is to validate on a large retrospective cohort Hollemans nomogram in predicting catheter drainage success. Secondary aims are to evaluate possible others predictors for success of catheter drainage in IPN and to evaluate the impact of antibiotherapy on microbiological results and on the need for an additional necrosectomy regarding its type and duration, as well as the emergence of multiresistance organism.

Using electronic patient data monitoring systems, the investigators reviewed all patients with a diagnosis of acute necrotizing pancreatitis admitted to digestive liver disease unity or ICUs at the Nantes University Hospital, Rennes, Angers and Brest University Hospital from January 1, 2012, to december 31, 2017. For enrolled patients, general clinical characteristics were collected. Patients were categorized into two groups according to success of catheter drainage or failure of catheter drainage, and the differences of these characteristics between two groups were evaluated. The Hollemans nomogram is evaluate and potential risk factors will be collected and studied by using multiple logistic regression analysis.

Inclusion Criteria:

• Adult patients (age >18 years) with acute necrotizing pancreatitis
• IPN proven or highly suspected (proven=positive cultures on pancreatic collections or gas on CT, highly suspected= sepsis without any others infection)
• The need for a catheter drainage

Non inclusion Criteria:

• patients under 18 years
• lack of radiologic evidence to diagnose acute necrotizing pancreatitis,
• no suspicion or confirmation of IPN
• no need for a drainage
• unable to receive a drainage
• pregnancy

Major patient, hospitalized with a diagnosis of acute necrotizing pancreatitis between 2012 and 2016, whatever the etiology, complicated by a necrosis infection infection (either proven or strongly suspected) requiring the initiation of antibiotic therapy and a drainage gesture of this CCM.