Early prediction of severity in acute pancreatitis

OBJECTIVE: To identify the score that best predicts early severity in patients with acute pancreatitis. MATERIALS AND METHODS: A prospective cohort study was performed from December 2009 to December 2011. Cases of pancreatitis were classified as mild or severe based on the organ failure criteria and/or local complications according to the Atlanta Symposium. APACHE-II, BISAP and urea were calculated using data from the first 24 hours. Ranson was calculated using data from the first 48 hours. Hemoconcentration was assessed 24 hours from admission, after adequate hydration. To assess the predictive value of the different scores the area under the curve ROC was used. RESULTS: 329 patients with acute pancreatitis were included: 214 women (65%) and 115 males (35%). The most common etiology was biliary 245 (74.3%). Two hundred and forty were mild cases of pancreatitis (73%) and 89 were severe (27%). The areas under the curve ROC were: APACHE-II: 0.74; BISAP: 0.78; Ranson: 0.48; urea: 0.70; hemoconcentration: 0.73. There were 10 deaths (mortality: 3%). There were no significant differences between BISAP and APACHE-II, urea or hemoconcentration, but a significant difference existed with Ranson score. CONCLUSION: BISAP is a simple score that applied early in the emergency room identifies patients with severe pancreatitis with an efficacy similar to APACHE-II, urea and hemoconcentration, and KEY WORDS: acute pancreatitis, BISAP, severity, prognostic scoring.