Can We Predict Disease Severity in Acute Pancreatitis?

The mean incidence of acute pancreatitis in the UK has been reported as being between 48 and 238 cases per year per million population. I Eighty per cent of patients make a rapid recovery with conservative treatment alone, but the remaining 20% of attacks are more severe and are associated with a 50llJo mortality.2 The complications associated with severe pancreatitis include adult respiratory distress syndrome, cardiovascular collapse, septicaemia, gastrointestinal haemorrhage and multiple organ failure. Sixty per cent of patients who die do so within 6 days of admission, predominantly from pulmonary complications;' Initial clinical assessment alone predicts only 39% of patients who will develop severe disease which rises to 73% 24 h after admission." Some means of identifying patients likely to develop severe pancreatitis in the first few hours after admission would be beneficial. Such patients could be carefully monitored from the outset and CT scanned to assess the extent of pancreatic necrosis with the possibility of surgical intervention.? Patients within this group may require endoscopic retrograde pancreoduodenostomy or endoscopic sphincterotomy if gall stones are present. 6 Such management could only be justified, and its value assessed, in patients with a high risk of developing severe pancreatitis, since the procedures are time consuming, expensive and not without risk. In 1944 Edmondson and Berne? reported that patients with severe pancreatitis developed hypocalcaemia, the mechanism of which still remains unclear. The late appearance and lack of specificity of hypocalcaemia as an index of severe pancreatitis has limited its clinical usefulness. Methaemalbumin in plasma or peritoneal fluid is a valuable pointer to haemorrhagic pancreatitis," but it may not appear until 24-48 h after the onset of the attack,