A prospective evaluation of the predictive value of serum amylase levels in the assessment of patients with blunt abdominal trauma

Abstract

Introduction: Less than ten percent of patients who sustain blunt abdominal trauma will suffer a significant intra-abdominal injury. Identifying these patients is difficult and this study reviews the results of routine serum amylase levels in a series of patients with blunt abdominal trauma.

Methods: All patients admitted, by the primary author from November 2010 to November 2012, with a diagnosis of blunt abdominal trauma were included. All these patients had a serum amylase level measurement performed on admission.

Results: One hundred and three patients were selected, with an age range from 3 to 68 years. There were 33 females and 70 males. Imaging was obtained in 47 patients (38 CT scans and 9 ultrasounds). Nine (19%) of the patients who were imaged required a laparotomy due to the radiological findings, and 38 (81%) of this sub group underwent successful conservative management. Eighteen patients had a laparotomy on clinical grounds. Intra-abdominal injuries were identified in 38 patients on imaging and/or at laparotomy. In five patients laparotomy did not reveal any injuries. The remaining 38 patients were admitted for serial abdominal observation. They were all discharged home and their symptoms resolved. The serum amylase level ranged from 34 U/L to 3 156 U/L, with a mean of 227 U/L (standard deviation 456 U/L). The levels were raised in 60 patients (58%) of whom 19 (32%) had a significant intra-abdominal injury. The serum amylase level was normal in 43 patients (42%), of whom 19 (44%) had a significant intra-abdominal injury. There were eight pancreatic injuries in the group (pancreatitis (1), pancreatic contusion (3), laceration (1), and transection (3). The serum amylase level was normal in two and mildly elevated in one of the patients with contusions (91, 92 and 129 U/L respectively), mildly elevated in the patient with pancreatitis (121 U/L), and significantly raised in the others (340 U/L with the pancreatic laceration; 3 156, 472, and 1 497 U/L in those with a transected pancreas). Four patients had a serum amylase level of greater than 1 000 U/L. Two of these had pancreatic injuries (3 156 and 1 497 U/L) and had hospital stays of six and sixteen days respectively. In the other two (3 042 and 1 454 U/L) no intra-abdominal injury was found.

Conclusion: The routine use of serum amylase level in the investigation of patients with blunt abdominal trauma cannot be supported as a mildly raised serum amylase level is common following blunt abdominal trauma, is of uncertain clinical significance, and does not have any predictive value. A markedly raised serum amylase level is associated with major pancreatic injury but is in itself a non-specific finding.