Acute pancreatitis: interobserver agreement and correlation of CT and MR cholangiopancreatography with outcome.

Abstract

To assess the correlation between and the interobserver agreement of contrast medium-enhanced computed tomography (CT) and nonenhanced and contrast-enhanced magnetic resonance (MR) imaging findings in patients with acute pancreatitis and to correlate these findings with outcome. Two blinded reviewers separately assessed contrast-enhanced CT and nonenhanced and contrast-enhanced MR images in 30 patients with acute pancreatitis and established a severity index based on the presence of peripancreatic fluid collections and pancreatic necrosis. The Spearman rank correlation coefficient and weighted kappa statistic were used to assess the correlation between each imaging technique and the interobserver agreement, respectively. Correlation between hospitalization days, morbidity, and severity indexes were assessed by using linear correlation. A strong correlation existed for both reviewers when comparing contrast-enhanced CT with nonenhanced (r = 0.82, 0.79) or contrast-enhanced (r = 0.82, 0.79) MR cholangiopancreatography or when comparing nonenhanced and contrast-enhanced MR cholangiopancreatography (r = 0.99, 1.00). The interobserver agreement in staging was stronger with nonenhanced (kappa = 0.76) and contrast-enhanced (kappa = 0.78) MR cholangiopancreatography than with contrast-enhanced CT (kappa = 0.70). There was no linear correlation between the severity index for contrast-enhanced CT and outcome, while there was a linear correlation between nonenhanced or contrast-enhanced MR cholangiopancreatographic staging and the patient morbidity rate. MR cholangiopancreatography could be an alternative to contrast-enhanced CT for the initial staging of acute pancreatitis.