Reviewed by Stephen C. Clement, MD

Editor's note: Studies demonstrating the adverse effects of glucose control on the immune system date back many decades. Although it seems intuitive that better glycemic control during surgery leads to improved outcomes, intensive perioperative glycemic control is not the standard of care, and systems of care that optimize glucose control from the time of hospital admission have not yet been adopted.

STUDY
Golden SH, Peart-Vigilance C, Kao WH, Brancati FL: Perioperative glycemic control and the risk of infectious complications in a cohort of adults with diabetes. Diabetes Care 22:1408-14, 1999.

SUMMARY
Objective. To evaluate the relationship between perioperative glucose control and subsequent risk of infection.

COMMENTARY
Drs. Golden, Peart-Vigilance, Kao, and Brancati have made a significant contribution to the growing literature regarding the association between poor glycemic control and risk for postoperative infection. Their hypothesis was that hyperglycemia in the immediate postoperative period is positively associated with excess risk for postoperative infection.

The authors reviewed the records of all patients with diabetes undergoing coronary artery bypass graft (CABG) surgery between the years 1990 and 1995. Of the 510 patient records available, 411 were extensively reviewed with regard to the level of postoperative glycemic control. Glycemic control was based on capillary blood glucose tests done four times a day using bedside glucose meters. Patients were classified as having developed a postoperative infection if a documented infection occurred >36 hours following surgery.

The most common infections were leg wound infections (10.9%), urinary tract infections (6.6%), sternal wound infections (5.6%), and pneumonia (4.6%). The authors demonstrated a slight but statistically significant association between degree of postoperative glucose control and risk for infection. Patients in the highest quartile of glucose levels had a relative risk of infection of 1.7 compared to patients in the lowest quartile of glucose levels.

The authors of this study took great efforts to assess postoperative glycemic control using the methods available to them—bedside glucose monitoring of capillary blood glucose. By relying only on this methodology, I suspect the authors may have underestimated the association between glucose control and infection.

The measurement of whole blood glucose by standard glucose meters is affected by hematocrit and hypoxia, when compared to laboratory glucose measurements. Most blood glucose meters are recommended for use in patients with hematocrits in the range of 35­50%. A hematocrit of <30% can result in an overestimation of the actual blood glucose by as much as 20%.

The impact of transfused blood, which often occurs postoperatively in CABG patients, has never been assessed. This potential bias may have caused several of the patients who had good or fair glucose control (based on laboratory values) to be categorized as having poor glucose control in this study. This potential bias may have prevented the authors from seeing an even stronger association than was found.

In support of the authors' findings, Furnary and associates from Portland, Ore., recently demonstrated a reduction in the incidence of deep sternal wound infections in their diabetic patients undergoing cardiac procedures after instituting a protocol for the use of continuous intravenous insulin infusion in the intraoperative and postoperative periods.¹ The authors should be commended for adding to the mounting literature that aggressive glucose control is critical to reducing the incidence of postoperative infection in patients with diabetes. The potential for cost savings in reduced hospital days, as well as the reduction in morbidity and mortality, is substantial.

REFERENCE

¹Furnary AP, Zerr KJ, Grunkemeier GL, Starr A: Continuous intravenous insulin infusion reduces the incidence of deep sternal wound infection in diabetic patients after cardiac surgical procedures. Ann Thorac Surg 67:352-62, 1999.

Stephen C. Clement, MD, is an associate professor of medicine at Georgetown University Medical Center in Washington, D.C.