Conventional and subtraction scanning of the pancreas: an assessment based on blind reporting

Abstract Conventional and subtraction pancreatic scans were performed in an unselected series of 221 patients. The results were assessed by blind reporting by a panel of three independent observers at three viewing sessions. At the first session, only the conventional scans were reviewed, at the second, only the subtraction scans and at the third, conventional and subtraction scans for each patient were reported on jointly. Fewer false positive results in normal patients were obtained at the third, combined, viewing session than when either the conventional or subtraction pancreatic scans were reported on alone. No false negative scans were found in chronic pancreatitis and few in patients with pancreatic tumours. A large number of abnormal scans was found in several categories of patients with no biochemical or radiological evidence of exocrine pancreatic disease; abnormal scans in these conditions must be interpreted with caution. The subtraction technique was not successful in all situations where the ...