Glucose Measurement: Confounding Issues in Setting Targets for Inpatient Management

Van den Berghe et al. (1) reported a significant reduction in mortality with normoglycemia (target value 80–110 mg/dl) in patients whose medical intensive care unit (ICU) stay was >72 h and reduced morbidity in all patients, regardless of the duration of ICU stay. Although severe hypoglycemia did not occur in the Van den Berghe et al. study, 18.7% of patients in the intensive treatment group compared with 3.1% of those who received conventional therapy did experience hypoglycemia (defined as glucose <40 mg/dl), albeit with no adverse consequences reported. However, altered consciousness is common in the ICU, and even severe hypoglycemia may be unrecognized. Other studies (2,3) examining intensive insulin protocols in various inpatient settings have suggested benefits in clinical outcomes associated with improved glycemic control. In a mixed ICU population, Van den Berghe et al. (2) previously demonstrated reduced morbidity and mortality with three- to fourfold less hypoglycemia than the medical ICU population (2). Thus, careful assessment of glucose measurement and how it may impact the targets selected in the hospital are critical safety issues in intensive management of hyperglycemia. As a result of increasing evidence that tight glycemic control is beneficial in the management of inpatients with diabetes, the American Diabetes Association (ADA) currently recommends a glucose target “as close to 110 mg/dl as possible and generally <180 mg/dl” for critically ill patients (4). The American Association of Clinical Endocrinologists recommends the “upper limits for glycemic targets” of 110 mg/dl in critically ill patients (5). In practice, it may be difficult to obtain the level of glycemic control (average glucose 111 mg/dl in the intensively managed group) achieved by Van den Berghe et al. Though a wider range of glucose values has been targeted, rarely have mean glucose values between 80 and 110 mg/dl been achieved, …