Predrainage and Postdrainage Prognostic Nomograms to Predict Outcome of Percutaneous Drainage for Infected Pancreatic and Peripancreatic Necrotic Collections.

OBJECTIVES

This study aimed to identify factors affecting outcome of percutaneous catheter drainage (PCD) in management of infected pancreatic necrosis treated with step-up approach.

METHODS

This was a single-center retrospective cohort study that included patients with infected necrosis undergoing PCD as initial intervention. Patients who did not respond underwent necrosectomy. Predictors of PCD failure (ie, mortality or need for necrosectomy) were analyzed. Models were constructed for predrainage and postdrainage use and were internally validated.

RESULTS

Of 304 patients included, catheter drainage was successful in 59.8%, with overall mortality of 22%. Predrainage model consisted of Acute Physiologic and Chronic Health Evaluation II score at admission, early organ failure, and pancreatic necrosis of greater than 50%. Postdrainage model consisted of Acute Physiologic and Chronic Health Evaluation II at first PCD, early organ failure, pancreatic necrosis of greater than 50%, sepsis reversal within 1 week of PCD and Escherichia coli in PCD culture. Both models were internally validated with area under receiver operating characteristics curve of 71.2% for pre-PCD and 81.2% for post-PCD model. Prognostic nomograms were constructed using the models.

CONCLUSIONS

Percutaneous catheter drainage alone was successful in 59.8% with mortality of 22%. The nomograms can help in guiding treatment strategy and referral of high-risk cases.