Rationale:

• For patients requiring insulin who undergo complex major surgical procedures, insulin infusions tend to control blood glucose levels better than intermittent insulin therapy and likely reduce perioperative infections.

Evidence:

• In a prospective cohort study of 2402 diabetic patients undergoing coronary artery bypass surgery, 968 patients received treatment with correction dose insulin for the first 2 postoperative days, and 1499 patients received continuous intravenous insulin infusion. There was a significant reduction in the incidence of deep sternal wound infections in the group that received the insulin infusion (0.8% vs. 2.0%) (42). More recently, the same team has reported reduced mortality in diabetic patients undergoing coronary artery bypass grafting (46).
• In 737 patients undergoing cardiac surgery, intra- and postoperative use of an insulin infusion protocol designed to achieve blood glucose levels of <130 mg/dL more than 50% of the time was accompanied by a reduction in the rate of mediastinitis from 1.6% (before initiation of the protocol) to 0%. It is noteworthy that 57% of the patients meeting the 130 mg/dL threshold for insulin therapy did not have a perioperative diagnosis of diabetes (47).
• The favorable effect on wound infection of postoperative continuous insulin infusion has also been reported for 761 patients undergoing coronary artery bypass grafting, 37% of whom were diabetic. Target blood glucose was 120 to 160 mg/dL. Prior to the use of the infusion protocol, the rate of wound infections was significantly higher in the diabetic patients than in the nondiabetic patients. The use of the protocol was associated with a reduction in the rate of wound infections in the diabetic patients to that observed in the nondiabetic group (48).

Comments:

• Because of consistent evidence showing the effectiveness and safety of meticulous glycemic control in the intensive care setting, particularly in cases of cardiac surgery, and because of consensus among experts (49), one should use an insulin infusion protocol in such a setting whenever possible. One should consider its use even in patients without a pre-ICU diagnosis of diabetes when target glycemia has been exceeded.
• However, benefit from insulin infusion intraoperatively during cardiac surgery was not observed in a randomized, controlled trial comparing intensive therapy with conventional therapy (50). Thus, there is no support at present for intraoperative intensive insulin therapy of the sort that has been shown to improve clinical outcomes when used during the postoperative period.
• Note that potassium levels may need to be checked more frequently if the patient has renal failure or is taking an ACE inhibitor, angiotensin-receptor blocking agent, or potassium-sparing diuretics.
• There are several different types of intravenous insulin infusions, each with its own benefits and disadvantages.
  • There is no clearly superior choice in insulin infusions.
  • There are three different insulin infusion protocols: the Yale Insulin Infusion Protocol, the Lien-Spratt Insulin Intravenous Nomogram, and the Trence-Hirsch Example of an Intravenous Insulin Infusion.
  • In assessing glycemic response to calculated correction dose insulin and making dose adjustments, it is important to do so in anticipation rather than in reaction to glycemic needs.
  • Whether insulin infusion is used depends on the degree of hyperglycemia and the kind of intravenous fluid being used (dextrose, intravenous alimentation fluid) regardless of the preoperative regimen used for glycemic control.
• Additional published examples of insulin infusion protocols that have been used successfully appear in 51 and 52.