Background/aims: Acute pancreatitis is the most serious complication of endoscopic retrograde cholangiopancreatography (ERCP) but is not very common. A test that could predict the occurrence of pancreatitis would help to decide whether to discharge a patient after ERCP or not. The aim of this prospective study was to compare the value of serum amylase and elastase in the prediction of post-ERCP pancreatitis and its severity.
Methodology: Ninety-seven patients underwent ERCP. Serum samples were taken before, two and six hours after ERCP for amylase and elastase measurement. Fifty-four patients (group A) were treated with continuous intravenous infusion of octreotide, beginning 6 hours before ERCP and terminating 24 hours after. Forty-three patients (group B) received no preventive treatment.
Results: In group A, 9 patients (16.6%) developed pancreatitis, 8 of them (14.8%) mild and 1 (1.8%) severe. Two patients in group B developed mild pancreatitis (p = 0.1). In all patients the predictive accuracy in the second hour for amylase >3N, >5N and elastase >N was 79%, 87% and 86% respectively. The likelihood ratio of positivity (LRP) was 3.6, 6.5 and 6.1. In the sixth hour the respective values were 76%, 86%, 85% and 4, 7.3 and 6.4. In group A, the respective values in the second hour were 85%, 91%, 94% and 5, 25.2, infinity, and in the sixth hour 85%, 94%, 98% and 5.7, 11.5, infinity.
Conclusions: Serum amylase (with cutoff value >5N) and elastase (>N), 6 hours after ERCP, were the most accurate tests for the prediction of post-ERCP pancreatitis, especially in patients receiving octreotide. The measurement of serum elastase could supplement that of serum amylase in the prediction of more cases of post-ERCP pancreatitis.