ABSTRACT
Illness severity is intrinsically linked to more severe insulin resistance, and numerous observational studies have found a U-shaped association between spontaneous blood glucose concentrations and outcome of critically ill patients, whereby the lowest mortality risk associates with blood glucose concentrations in the normal fasting range. A landmark randomized controlled trial (RCT) performed in an adult, surgical intensive care unit in 2001 in Leuven, Belgium, found that lowering blood glucose to the healthy fasting range with insulin significantly reduced morbidity and mortality as compared to tolerating hyperglycemia up to the renal threshold. Use of validated decision-support software was found to result in a high time in target range, with a very low risk of hypoglycemia. The outcome differences between the pioneer RCT and largest multicenter RCT are likely explained by important methodological differences, including differences in blood glucose target, feeding strategies, glucose measurement technology, and insulin administration protocol.
