Villous adenoma in a blind duodenal loop.

Serum amylase levels in patients in this type of coma without clinical evidence of pancreatitis have seldom been reported and have usually been normal. The level reported in the present patient was raised but not enough to be diagnostic of acute pancreatitis. Although acute renal failure does produce raised levels it is unlikely to have been the cause here, because the amylase level fell to normal long before the blood urea began to fall. The intravascular haemolysis lasted for a few days only and subsequent screening for abnormal haemolytic mechanisms, such as cold agglutinins, Coombs' test, osmotic fragility and haemoglobinopathy, was negaiye. The fact that the platelet count never fell below normal made microangiopathic haemolysis unlikely. A more likely cause of the haemolysis became evident when considering the intravenous fluid used. In the first 6 hr the patient received 11 of 097% and 21 of 018%. saline. This was followed by a further 41 of0 18%0 saline in the next 24 hr. The large volume given was needed and the hypotonic solution used as it appeared appropriate. However, because of the rapidity of infusion this could have made the area around the tip of the cannula hypotonic and caused haemolysis. One-fifth 'normal', or 0 18%, saline has been used for correcting dehydration in babies for a long time, but is seldom used now as haemolysis sometimes occurred. We have no definite proof that the hypotonic saline infusion caused the haemolysis in our patient, but it seems likely. If large volumes of fluid are needed rapidly intravenously, 045%o saline or infusions of 09°/ saline and 5%/ dextrose alternately are suitable.