masthead


Frequently Asked Questions

Glucose levels in the critically ill: how low should we go? Results of the NICE-SUGAR Study

Background
Hyperglycemia is a well known complication in critically ill patients. In response, the body releases stress hormones such as epinephrine and cortisol. The release of stress hormones, along with decreased activity, parenteral nutrition, and possible corticosteroid therapy all contribute to increasing glucose levels. Hyperglycemia can be dangerous because it decreases immune function, impairs wound healing, increases oxidative stress, creates a pro-coagulant state, and causes endothelial dysfunction.1 This phenomenon is seen in a variety of hospital settings, especially in critically ill patients in the intensive care unit (ICU). In the past, it has been noted that when stress levels decrease, or patients start to improve, glucose levels will self-regulate and return to baseline.

Literature Review
Multiple trials over the past 10 years have evaluated optimal glucose level targets in ICU patients. The most influential of these trials were the Van den Berghe trials conducted in Belgium. The first of the 2 single center trials investigated optimal glucose levels in surgical ICU patients and the second, 5 years later, evaluated the same outcome in medical ICU patients. Patients were randomized to receive either intensive insulin control, to a target glucose range of 80-100 mg/dL, or conventional glucose control group, which aimed for levels of less than 215 mg/dL.2,3 The first trial found that patients in the surgical ICU benefited from tight glucose control.2 Mortality rates were 4.6% in the intensive insulin groups versus 8.0% in the conventional group (p<0.04). Septecemia was reduced by 46%. These benefits, however, were seen in patients who were hospitalized for longer than 5 days (p=0.005). Hypoglycemia was more frequent in the intensive therapy group (p=not reported). The second trial demonstrated patients in the medical ICU benefited when they were hospitalized for greater than 3 days.3 Mortality was reduced in this group; 52.5% in the conventional group versus 43% in the intensive group (p=0.009). Again, hypoglycemia occurred more often in the intensive therapy group. Subsequent to the publication of these trials, institutions across the world have debated as to what optimal glucose levels in critically ill patients should be. Because of the need for stronger evidence regarding tight glucose control, difficulty in consistently adhering to strict glucose levels, and the risks of hypoglycemia, additional trials were necessary to confirm the Van den Berghe findings.

The most recently published trial, NICE-SUGAR (Normoglycemia in Intensive Care Evaluation- Survival Using Glucose Algorithm Regulation), investigated intensive versus conventional glucose control in patients who were in the ICU for greater than 3 days.4 This randomized, non-blinded, parallel group, multicenter, controlled trial strived to further investigate the Van den Berghe trial results. Patients were included if they were being admitted to the surgical or medical ICU and had a projected stay of greater than 3 days as predicted by their physician. The goal of the NICE-SUGAR trial was to evaluate all cause mortality at 90 days and determine if glucose control impacted this outcome.

Patients were randomized to either the intensive glucose control group, glucose between 81-108 mg/dL (n=3010), or to the conventional group, glucose less than or equal to 180 mg/dL (n=3012).4 Insulin was given by intravenous infusion to both groups according to an online algorithm. The conventional group was given insulin if glucose levels were above 180 mg/dL and stopped when glucose was below 144 mg/dL. Blood glucose measurements were conducted at the point of care from arterial samples in both groups. The intervention was discontinued when patients were discharged from the ICU or if they were able to eat food by mouth. The primary outcome was death from any cause at 90 days. There were a variety of secondary and tertiary outcomes including the incidence of hypoglycemia (blood glucose of <40 mg/dL).

At 90 days, 829 (27.5%) of the intensive group died versus 751 (24.9%) in the conventional group for an absolute difference of 2.6% (95% confidence interval [CI] 0.4 to 4.8; odds ratio [OR] 1.14, 95% CI 1.02 to 1.28, p=0.02).4 Statistical adjustments were made to account for any confounders, and the values remained significant after the appropriate adjustments. Causes of death in both groups were similar except for a significantly greater amount of cardiovascular deaths in the intensive group. As expected, there were more hypoglycemic episodes in the intensive group 206 (6.8%) when compared to the conventional group 15 (0.5%), OR 14.7, 95% CI 9 to 25.9, p<0.001). There were no differences in terms of nutrition administered between the groups that could have influenced glucose levels. Of note, more patients in the intensive group received corticosteroids 1042 (34.6%) versus 955 (31.7%) to treat septic shock (p=0.02). The authors concluded that intensive glucose control increased mortality in patients in the ICU.

There were several strengths and limitations to this trial. To begin with, this was a large, randomized trial that spanned numerous hospitals and institutions. The use of a standardized glucose control algorithm was a strong aspect that ensured a standard for practitioners to follow when administering insulin. The study was well organized, and when needed, clinicians could consult study coordinators for assistance with interpreting the insulin algorithms. Additionally, this study investigated the Van den Bergh trial results and was appropriately structured to do so. However, there were also limitations of the study. The first was that it was difficult to predict who would stay in the ICU longer than 3 days and who would not. The study investigators did their best with this aspect of the trial, but there was no standard for anticipating each patient’s length of ICU stay. The authors chose their glucose level targets by surveying various study sites across the world. Not all centers target the levels investigated in the trial, so the results may not be completely applicable in various medical centers.

In an accompanying editorial, Inzucchi and Siegel highlight the idea that many ICUs have established individual protocols for glycemic control that do not necessarily match those of the NICE-SUGAR trial.5 The editorialists appropriately conclude that lowering glucose to levels below 140-180 mg/dL does not appear to confer additional benefits and may be harmful. They further caution that these data do not support a return to sliding scale insulin protocols and less careful glycemic control.

Conclusion
Overall, this study poses more questions than solid conclusions as to the optimal glucose level debate. Now that the NICE-SUGAR evidence demonstrated significant harm to patients (number needed to harm 38) who received intensive insulin treatment, physicians may chose not to control glucose as vigorously as the Van den Berghe trials once suggested. It is difficult to assess exactly what glucose levels are optimal in ICU patients, but it seems reasonable to keep patients in the range of the conventional groups since conventionally managed patients fared better than those in the intensive control group.

References

  1. Inzucchi SE. Management of hyperglycemia in the hospital setting. N Engl J Med. 2006; 355(18):1903-1911.
  2. Van den Berghe G, Wouters P, Weekers F, et al. Intensive insulin therapy in the critically ill patients. N Engl J Med. 2001;345(19):1359-1367.
  3. Van den Berghe G, Wilmer A, Hermans G, et al. Intensive insulin therapy in the medical ICU. N Engl J Med. 2006;354(5):449-461.
  4. NICE-SUGAR Study Investigators, Finfer S, Chittock DR, Su SY, et al. Intensive versus conventional glucose control in critically ill patients. N Engl J Med. 2009;360(13):1283-1297.
  5. Inzucchi SE, Siegel MD. Glucose control in the ICU-How tight is too tight? N Engl J Med. 2009;360(13):1346-1349.

By Alana Dikopf