Background: The early identification of severe acute pancreatitis is important for the management and for improving outcomes. The bedside index for severity in acute pancreatitis (BISAP) has been considered as an accurate method for risk stratification in patients with acute pancreatitis. This study aimed to evaluate the comparative usefulness of the BISAP.
Methods: We retrospectively analyzed 303 patients with acute pancreatitis diagnosed at our hospital from March 2007 to December 2010. BISAP, APACHE-II, Ranson criteria, and CT severity index (CTSI) of all patients were calculated. We stratified the number of patiants with severe pancreatitis, pancreatic necrosis, and organ failure as well as the number of deaths by BISAP score. We used the area under the receiver-operating curve (AUC) to compare BISAP with other scoring systems, C-reactive protein (CRP), hematocrit, and body mass index (BMI) with regard to prediction of severe acute pancreatitis, necrosis, organ failure, and death.
Results: Of the 303 patiants, 31 (10.2%) were classified as having severe acute pancreatitis. Organ failure occurred in 23 (7.6%) patients, pancreatic necrosis in 40 (13.2%), and death in 6 (2.0%). A BISAP score of 2 was a statistically significant cutoff value for the diagnosis of severe acute pancreatitis, organ failure, and mortality. AUCs for BISAP predicting severe pancreatitis and death were 0.80 and 0.86, respectively, which were similar to those for APACHE-II (0.80, 0.87) and Ranson criteria (0.74, 0.74) and greater than AUCs for CTSI (0.67, 0.42). The AUC for organ failure predicted by BISAP, APACHE-II, Ranson criteria, and CTSI was 0.93, 0.95, 0.84 and 0.57, respectively. AUCs for BISAP predicting severity, organ failure, and death were greater than those for CRP (0.69, 0.80, 0.72), hematocrit (0.45, 0.35, 0.14), and BMI (0.41, 0.47, 0.17).
Conclusion: The BISAP predicts severity, death, and especially organ failure in acute pancreatitis as well as APACHE-II does and better than Ranson criteria, CTSI, CRP, hematocrit, and BMI.