PANCREATIC TRAUMA - A 10-YEAR MULTIINSTITUTIONAL EXPERIENCE

Our objective was to determine the incidence, management, and outcome of traumatic pancreatic injury. A retrospective review was performed o f all patients with pancreatic injury admitted to two Level I trauma h ospitals over a 10-year period. Comparisons were made with Chi square or Fisher's exact tests. Of 16,188 trauma admissions, 72 patients (0.4 %) had pancreatic injury. The mean age was 30 years, and 50 patients ( 69%) were male. Mechanism of injury was gunshot in 32 (45%), blunt in 27 (37%), and stab wound in 13 (18%). The pancreas was involved in 1.1 per cent of patients with penetrating injuries compared to 0.2 per ce nt with blunt injuries (P < 0.01). There were 18 grade I (25%), 32 gra de II (45%), 16 grade III (22%), and 5 grade IV (7%) injuries. Initial diagnosis was made intraoperatively in 63 patients and by computed to mography in 8. The mean injury grade was significantly lower on comput ed tomography compared to surgical exploration (0.4 vs 2.0; P < 0.05). Operative procedures included distal pancreatectomy in 23 (32%), expl oration only in 22 (31%), external drainage in 13 (18%), pancreatorrha phy in 4, internal drainage in 2, and proximal resection in 2. Mortali ty was 16.6 per cent and was not related to the mechanism or grade of injury. Mean Injury Severity Score and transfusion requirements were s ignificantly greater in patients who died (P < 0.05). Morbidity occurr ed in 30 patients (42%), including pancreatic fistula (11%), pancreati tis (7%), and pancreatic pseudocyst (3%). Six patients (8%) developed intra-abdominal abscesses, and all had associated liver or intestinal injuries. In patients with grade I and II injuries, morbidity was high er with external drainage compared to exploration without drainage. Pa ncreatic injury is infrequent and is more often associated with penetr ating trauma. Diagnosis is most commonly made by exploration and canno t be excluded by computed tomography. Drainage of low-grade injuries m ay not be necessary. Morbidity and mortality in patients with pancreat ic trauma is significant and is primarily due to associated injuries.