Attempting to maintain normoglycemia during cardiopulmonary bypass with insulin may initiate postoperative hypoglycemia

We attempted to develop an insulin administration protocol that maintains n ormoglycemia in patients undergoing cardiac surgery and to study the effect s of intraoperative blood glucose management on serum levels of creatine ph osphokinase isoenzyme BE (CK-BB) and S-100 protein. Twenty nondiabetic pati ents were randomly allocated to receive either "tight control" of blood glu cose with a standardized TV insulin infusion intraoperatively (Group TC) or "no control" of blood glucose intraoperatively (Group NC). Perioperative s erum levels of glucose, CK-BB, and S-100 protein were determined in all pat ients. Group TC patients received 90.0 +/- 49.2 units of insulin, whereas G roup NC patients received none. Despite insulin, both Group TC (P = 0.00026 ) and Group NC (P = 0.00003) experienced similar significant increases in b lood glucose levels during hypothermic cardiopulmonary bypass. However, mea n blood glucose level upon intensive care unit ar rival was significantly d ecreased in Group TC, compared with Group NC (84.7 +/- 41.0 mg/dL, range 32 -137 mg/dL vs 201.4 +/- 67.5 mg/dL, range 82-277 mg/dL, respectively; P = 0 .0002). Forty percent of Group TC patients required treatment for postopera tive hypoglycemia (blood glucose level <60 mg/dL). Substantial interindivid ual variability existed in regard to insulin resistance. The investigation was terminated after we realized that normoglycemia was unattainable with t he study protocol and that postoperative hypoglycemia was unpredictable. Ai l patients in both groups experienced similar significant increases in post operative serum levels of CK-BB and S-100 protein. These results indicate t hat "tight control" of intraoperative blood glucose in nondiabetic patients undergoing cardiac surgery was unattainable with the study protocol and ma y initiate postoperative hypoglycemia. Implications: The appropriate intrao perative management of hyperglycemia and whether it adversely affects neuro logic outcome in patients after cardiac surgery remains controversial. This investigation reveals that attempting to maintain normoglycemia in this se tting with insulin may initiate postoperative hypoglycemia.