Practice and yield of early CT scan in acute pancreatitis: a Dutch Observational Multicenter Study.

ABSTRACT

BACKGROUND: Early computed tomography (CT) (within 4 full days after symptom onset) may be performed to distinguish acute pancreatitis (AP) from other intra-abdominal conditions or to identify early pancreatic necrosis. We analyzed practice and yield of early CT in patients with an established clinical diagnosis of AP in a Dutch cohort (EARL study). METHODS: Multicenter observational study. Etiology, disease course, CT timing, Balthazar CT score, and clinical management were evaluated. RESULTS: First documented hospital admissions of 166 patients were analyzed. Etiology was biliary (42.8%), unknown (20.5%), alcoholic (18.1%), post-endoscopic retrograde cholangiopancreatography (11.4%), and miscellaneous (7.2%). In 89.2% (148/166), the disease course was mild. Out of 18 patients with severe AP, 11 eventually developed (peri)pancreatic necrosis. At least one CT (range 1-12) was performed in 47% (78/166) of all patients and in 62.8% (49/78) it was acquired within 4 full days after symptom onset. Practice, timing, and Balthazar CT score of early CTs were not significantly different between mild and severe AP. None of the early CTs showed necrosis and no alternative diagnoses were established. In 89.8% (44/49), clinical management was not altered after early CT. In 10.2% (5/49), prophylactic antibiotics were started, but in absence of necrosis. CONCLUSIONS: A CT scan was frequently acquired early in the course of AP, but its yield was low and had no implications with regard to clinical management. It seems prudent that clinicians should be more restrictive in the use of early CT, in particular in mild AP, to prevent unnecessary radiation exposure and to save costs.