SAN ANTONIO, TEX—Until recently, it was widely believed that in critically ill patients, hyperglycemia is an adaptive response that provides extra glucose for vital organs and tissues. Thus, the conventional approach was not to treat hyperglycemia in these patients unless blood glucose levels became excessively high (greater than 215 mg/dL).

About a year ago, though, surprising data reported by Greet Van den Berghe, MD, and colleagues suggested that hyperglycemia actually prolongs the need for intensive care and predisposes the critically ill to complications and death.[1] The study compared strict to conventional glycemic control in nearly 1,600 mechanically ventilated adults.

At this year’s annual meeting of the Society of Critical Care Medicine, Dr. Van den Berghe provided additional information about the study—why and how it was done and its implications for critically ill patients.[2] She also detailed the results of other investigations that demonstrate the need for strict glycemic control.

“Strict glycemic control—below 110 mg/dL—using exogenous insulin … reduces ICU [intensive care unit] and hospital mortality and prevents the very typical ICU complications of the prolonged critically ill,” said Dr. Van den Berghe, who heads the Department of Intensive Care Medicine at the University of Leuven in Belgium. She described the “prolonged critically ill” as those who are in the ICU for more than five days.

INITIAL OBSERVATIONS

The first discovery that led Dr. Van den Berghe and colleagues to their study was their finding that high serum concentrations of insulin-like growth factor binding protein 1 (IGFBP-1) predicted mortality in patients with prolonged critical illness. This, said Dr. Van den Berghe, suggests that critically ill patients, particularly nonsurvivors, are lacking an insulin effect. This hypothesis is further supported by the fact that IGFBP-1 concentrations correlate inversely with insulin levels.

Hyperglycemia is common in critically ill patients; traditionally, it has been ascribed to insulin resistance, particularly in the liver but also in skeletal muscles and the heart. But what if the traditional explanation was wrong? What if hyperglycemia predisposes patients to specific complications, prolonged ICU dependency, and death? This is the question Dr. Van den Berghe’s group set out to examine.

INSULIN ADMINISTRATION AND FEEDING

Upon ICU admission, the study participants were randomized to strict or conventional glycemic control and started on a standard feeding regimen. This regimen called for 7 to 9 kcal/kg/d initially with a gradual increase to 25 kcal/kg/d by the end of the first week.

Both groups received their insulin by continuous infusion with no bolus injections. Critical care nurses were responsible for blood glucose monitoring and insulin dose adjustments. The nurses monitored blood glucose levels every one to four hours, performing more frequent checks when steep rises or falls in those levels occurred.

The two groups were comparable in age, sex, body mass index, initial illness severity, and rates of preadmission diabetes, preadmission malignancy, and hyperglycemia on admission. Furthermore, their feeding regimens were similar throughout the study.

REDUCED MORBIDITY, MORTALITY, AND COSTS

Total mortality was 4.6% in the strict glycemic control group and 8% in the conventional treatment group. This difference could not be attributed to any acute effect; in fact, mortality on day 5 was similar in the two groups. Instead, the difference was due entirely to declines in mortality among patients with prolonged critical illness.

In this subgroup, 10.6% of those given strict glucose control died versus 20.2% of those treated conventionally. This effect of strict glycemic control on mortality persisted throughout the hospital stay and was independent of illness severity. “Even in the most severely ill group … ICU mortality was significantly suppressed [by strict control],” Dr. Van den Berghe reported.

Strict control also significantly improved ICU morbidity, decreasing bacteremia by 46%, critical illness polyneuropathy by 44%, and acute renal failure by 41%. It reduced transfusion requirements from two to one per patient and favorably altered inflammatory markers, such as the C-reactive protein level, white blood cell count, and body temperature. In addition, among patients with prolonged critical illness, it shortened the length of ICU stay by three days.

Dr. Van den Berghe conservatively estimated that such a decline in length of stay would translate to about $2 million in annual savings. Her estimate assumed cost savings of $1,400 per day, the daily charge to stay in the ICU at her university’s hospital.

Not surprisingly, the rate of hypoglycemia was much greater in the strict glycemic control group (5.2% vs 0.8%). However, hypoglycemic events “were always very short and were never clinically associated with relevant side effects, such as seizures, hemodynamic instability, or anything else,” Dr. Van den Berghe said. In fact, she noted, these events were usually due to human error (eg, someone forgetting to turn down the insulin infusion when reducing the feeding).

WHY GLYCEMIC CONTROL WORKS

Was it strict glycemic control or merely insulin administration that provided the favorable results in this study? Dr. Van den Berghe strongly doubts that it is the insulin dose per se; she and her colleagues found that both blood glucose concentrations and insulin levels were strong positive risk factors. “The more insulin [that] was given, the higher the risk of death, which excludes a direct effect of insulin per se,” she concluded. “It appears to be that blood glucose levels—or another marker that is mimicked by blood glucose levels—seem to be important in explaining the effect on mortality.” Blood glucose levels correlated strongly and directly not only with mortality but also with the incidence of polyneuropathy.

Strict glycemic control’s main benefit with regard to mortality appeared to be the prevention of death from sepsis-associated multiorgan failure. Strict control may have achieved that effect by suppressing inflammation and enhancing immune function, Dr. Van den Berghe suggested.

—Timothy Begany

References
1. Van den Berghe G, Wouters P, Weekers F, et al. Intensive insulin therapy in the critically ill patients. N Engl J Med. 2001;345:1359-1367.
2. Van den Berghe G. Strict metabolic control with insulin to improve outcome of critical illness. Presented at: Annual Meeting of the Society of Critical Care Medicine; February 1, 2003; San Antonio, Tex.

Return to table of contents