Cost-benefit analysis of the work-up for pancreatic cancer.

We reviewed the records of 126 patients with pancreatic cancer to assess the value of diagnostic tests. The most commonly performed studies were computed tomography (CT) (97% of patients), endoscopic retrograde cholangiopancreatography (ERCP) (44%), and fine-needle aspiration (FNA) (41%). Of 34 patients who were found to have a mass in the body or tail of the pancreas on CT, 13 underwent ERCP; the results found by ERCP did not affect the management of the patients, whereas the results of FNA in 12 patients eliminated the need for operation. Of 14 patients with suspected metastases as evidenced by CT, the results of 3 ERCPs had no impact, whereas 5 of 7 patients who had FNAs avoided operation. Of five patients with normal results on CT, three had an ERCP that identified tumor. Of 26 patients with atypical CTs, the results of 12 of 16 ERCPs and 3 of 5 FNAs confirmed cancer. In contrast, in 48 patients with a mass in the head of the pancreas and biliary dilatation, ERCP did not alter the patients' management; only 3 of 14 patients who had FNAs avoided operation. Thus, the results of ERCP rarely altered the management of the patient when the CT showed a mass but was useful when the scan was normal or atypical. FNA was helpful in patients with cancer in the body or tail of the pancreas or with suspected metastases and in confirming the diagnosis when the CT was inconclusive.