What’s the best way of monitoring blood glucose in intensive care units?

"Diabetes of injury" is a term that has been used to describe the insulin resistance and resulting hyperglycemia that is a frequent feature of severe trauma and critical illness. It was once supposed that hyperglycemia is a consequence of critical illness; an adaptive, protective stress response, which is important for survival.

This dogma allowed the judgment that a moderate degree of hyperglycemia among the critically ill should be tolerated, and remain untreated. Then in 2001 a landmark study was published that challenged this benign view of hyperglycemia.

The study showed that intensive insulin therapy, directed at maintaining blood glucose concentration within the normoglycemic range, 80-110 mg/dL (4.4-6.1 mmol/L), significantly reduced both morbidity and mortality associated with critical illness. Now that the toxicity of hyperglycemia is recognized, intensive insulin therapy has become a standard part of care for many patients in intensive care units.

The success of the new therapy depends crucially on frequent and accurate blood glucose measurement within intensive care units, and there is now, as never before, great interest in identifying the most accurate methods for blood glucose point-of-care testing. This was the subject of a study presented at the annual meeting of the American College of Chest Physicians in November last year.

The authors of this study submitted 234 arterial, venous and capillary blood samples for blood glucose measurement by three methods: reference (gold standard) laboratory method, and two point-of-care testing modalities, a hand-held glucometer and the blood gas analyzer sited in their intensive care unit. 

Mean venous blood glucose determined using the laboratory reference method was 129 mg/dL (7.1 mmol/L) with a range of 54 mg/dL (3.0 mmol/L) to 350 mg/dL (19.2 mmol/L). Capillary blood glucose determined by glucometer was higher than the laboratory reference method by a mean of 20.95 mg/dL (1.2 mmol/L). This difference was essentially the same if either arterial or venous blood was tested.

The mean of blood glucose determined by the blood gas analyzer was only 3.07 mg/dL (0.17 mmol/L) higher than that using the laboratory reference method. The authors conclude that the blood gas analyzer is the more accurate POCT method for blood glucose estimation.