APACHE system is better than Ranson system in the prediction of severity of acute pancreatitis.

BACKGROUND: It has been suggested that addition of obesity score to the APACHE-II system can lead to more accurate prediction of severity of acute pancreatitis. However there is scanty information on the usefulness of the combined APACHE-O scoring system in Asian patients. This study aimed to compare the accuracy of Ranson, APACHE-II and APACHE-O systems in assessing severity of acute pancreatitis in a local Chinese population. METHODS: One hundred and one consecutive patients with acute pancreatitis were prospectively studied. Body mass index (BMI) was measured on admission. Ranson score, APACHE-II and APACHE-O scores were recorded on admission and at 48 hours. By adopting the cut-off levels and definitions advocated in the Atlanta consensus for severe disease, the diagnostic accuracy of the three scoring systems was compared by the area under the curve (AUC) under the receiver operator characteristic curve. RESULTS: Of the 101 patients, 12 (11.9%) patients suffered from severe pancreatitis. Obesity was uncommon and only two patients (2.0%) had BMI>30. Eighty-two (81.2%) patients were normal weight (BMI< or =25) whereas 17 (16.8%) were overweight (BMI 25-30). Overweight or obesity (BMI>25) was not associated with severe pancreatitis (P=0.40). The AUC for admission scores of Ranson, APACHE-II, and APACHE-O systems was 0.549, 0.904 and 0.904, respectively. The AUC for 48-hour scores of Ranson, APACHE-II and APACHE-O systems was 0.808, 0.955 and 0.951, respectively. CONCLUSIONS: The APACHE-II scoring system is more accurate than the Ranson scoring system of the prediction of severity in acute pancreatitis. Addition of obesity score does not significantly improve the predictive accuracy of the APACHE-II system in our local population with a low prevalence of obesity.