RESUMO
BACKGROUND: Recent guidelines advocate a step-up approach for managing suspected infected pancreatic necrosis (IPN) during acute pancreatitis. Nearly half the patients require secondary necrosectomy after catheter drainage. Our primary objective was to assess the external validity of a previously reported nomogram for catheter drainage, based on four predictors of failure. Our secondary objectives were to identify other potential predictors of catheter-drainage failure. We retrospectively studied consecutive patients admitted to the intensive care units (ICUs) of three university hospitals in France between 2012 and 2016, for severe acute pancreatitis with suspected IPN requiring catheter drainage. We assessed drainage success and failure rates in 72 patients, with success defined as survival without subsequent necrosectomy and failure as death and/or subsequent necrosectomy required by inadequate improvement. We plotted the receiver operating characteristics (ROC) curve for the nomogram and computed the area under the curve (AUROC). RESULTS: Catheter drainage alone was successful in 32 (44.4%) patients. The nomogram predicted catheter-drainage failure with an AUROC of 0.71. By multivariate analysis, catheter-drainage failure was independently associated with a higher body mass index [odds ratio (OR), 1.12; 95% confidence interval (95% CI), 1.00-1.24; P = 0.048], heterogeneous collection (OR, 16.7; 95% CI, 1.83-152.46; P = 0.01), and respiratory failure onset within 24 h before catheter drainage (OR, 18.34; 95% CI, 2.18-154.3; P = 0.007). CONCLUSION: Over half the patients required necrosectomy after failed catheter drainage. Newly identified predictors of catheter-drainage failure were heterogeneous collection and respiratory failure. Adding these predictors to the nomogram might help to identify patients at high risk of catheter-drainage failure. CLINICALTRIALS: gov number: NCT03234166.