Predictors for early readmission in acute pancreatitis (AP) in the United States (US) – A nationwide population based study

Abstract

Background & AimsPopulation based data on the burden and patterns of acute pancreatitis (AP) early readmissions (≤30-days) are limited. Methods2013 Nationwide Readmission Database (NRD) was queried. AP etiology was determined using associated diagnoses codes. Proportion, reasons for readmission, and associated costs were evaluated. Multivariate logistic regression analysis was performed to identify independent predictors for 30-day readmission. ResultsAfter exclusions, we identified 178,541 patients with primary diagnosis of AP (mean age 53 ± 17 years, 51% male). 13.7% were readmitted ≤30 days [7.1% in acute biliary pancreatitis (ABP) patients with index cholecystectomy (CCY), 16.3% in ABP patients without CCY, and 14.3% in non-biliary AP patients (p < 0.0001)]. Reasons for readmission included AP, chronic pancreatitis, Pseudocyst/walled off necrosis, biliary tract disease, smoldering symptoms and others. On multivariate analysis male gender, comorbidity status (≥3), non-biliary etiology, organ failure, Pseudocyst/walled off necrosis complications, and patients discharged to extended care facilities were associated with increased risk of readmission. ABP patients with index CCY had a significantly lower risk of early unplanned readmission (odds ratio 0.45, p < 0.0001) but ABP patients with index ERCP did not (p = 0.96). ConclusionsAbout 1 in 7 AP patients had a 30-day readmission after index hospitalization and about half of these were related to AP. Our data confirms the higher risk of readmission in alcohol and idiopathic AP and a lower risk in ABP. Risk of early unplanned readmission is significantly lower in ABP patients who underwent CCY and not ERCP during index hospitalization. Cholecystectomy should be performed in all ABP patients as per recommended guidelines.