Objectives: The objective of the study was to compare the accuracy of predictive methods for mortality in patients with severe acute pancreatitis (SAP) on admission and on the day when infection was confirmed.
Methods: Medical records of patients admitted for SAP in our hospital during January 2000 to November 2010 were retrospectively reviewed. Among those with infectious complications, time when infection was confirmed (TIC) and Acute Physiology and Chronic Health Evaluation II (APACHE II) score on admission and at the time when infection was confirmed (APACHE II OTIC) were studied. The correlations among the APACHE II score on admission, APACHE II OTIC score, and TIC were analyzed. The predictive accuracy was assessed by the area under the receiver operating characteristic curve.
Results: Time when infection was confirmed correlated negatively with the APACHE II score on admission and the APACHE II OTIC score (P < 0.05). The optimum cutoff value and the corresponding areas under the receiver operating characteristic curve for APACHE II score on admission, APACHE II OTIC score, and TIC were greater than 8, greater than 5, 12 days or less, and 0.67 (95% confidence interval [CI], 0.54-0.77), 0.84 (95% CI, 0.73-0.91), and 0.73 (95% CI, 0.61-0.82), respectively. Compared with the APACHE II score on admission, the APACHE II OTIC score was more accurate in predicting mortality (P = 0.029).
Conclusions: The time when infection is confirmed is a better time point for the reassessment of the outcome in patients with SAP.