Trial suspended on 26 February, following the publication of a trial with negative results (Ann Intern Med 146(4), 2007). Pending ethics committee re-approval.

Blood glucose levels increase in response to stress, infection or other conditions faced by patients in the hospital. This occurs commonly among patients with known diabetes, but also among non-diabetic hospitalized patients. Tight glucose control, the maintenance of blood glucose levels within normal limits (80-120 mg/dl), has been shown to improve patient outcomes in the hospital in several settings, mainly among critically ill patients hospitalized in intensive care units.

We plan to assess the importance of tight glucose control during open-heart surgery. The prevalence of hyperglycemia (elevated blood glucose) during this operation is high. Hyperglycemia may be associated with increased vulnerability to surgical site infections, neurological damage, cardiac and renal injury. Conversely, tight glucose control may be associated with hypoglycemia (pathologically low glucose levels) that may results in neurological injury. We hypothesize that tight glucose control will improve patient outcomes following surgery.

Current evidence supports intensive glucose control for patients in the intensive care unit post-cardiac surgery. The risk-benefit ratio of tight glucose control using continuous insulin infusion during surgery has not been established. Pros for tight control include the association of hyperglycemia with neurological injury, cardiac ischemia, white blood cell dysfunction and renal failure. The cons include adverse effects, mainly hypoglycemia and hypokalemia. As with any intervention in medicine, tight intra-operative glucose control should be assessed in a randomized controlled trial.

Objectives:to assess whether tight intra-operative tight glucose control using continuous insulin infusion reduces morbidity and mortality following cardiac surgery, defined as the incidence rate of surgical site infections, adverse neurological events, renal failure and 30-day mortality following CABG.

Additional outcomes will include the effect of continuous insulin infusion on longer-term mortality; other infectious complications and antibiotic use during hospitalization; cardiovascular outcomes; the need for re-operations; length of hospital stay; readmission; hypoglycemia and other adverse events.

Design: randomized controlled trial, with blinding of outcome assessors.

Participants: all consecutive patients >18 years undergoing CABG, without or without additional valve or other surgery at Rabin Medical Center; Beilinson campus, providing informed consent.

Exclusion criteria: patients with diabetic ketoacidosis, or hyperosmolar coma.

Intervention: Continuous insulin infusion throughout the operation aimed to maintain normoglycemia using a nomogram

Control: Glucose management according to the discretion of the anesthesiologist (continuous or bolus infusion)

During the early post-operative period (ICU-stay following surgery), all patients will be treated with intensive glucose control targeting glucose levels between 80-110.

Inclusion Criteria:

• All consecutive patients >18 years undergoing CABG, without or without additional valve or other surgery at Rabin Medical Center; Beilinson campus, providing informed consent.

Exclusion Criteria:

• Patients with diabetic ketoacidosis, or hyperosmolar coma