Glycemic control in the ICU: a multicenter survey

Background. Intensive insulin therapy has recently been shown to decrease morbidity and mortality in the critically ill population in a large randomized clinical trial. Objective. To determine the beliefs and attitudes of ICU clinicians about glycemic control. Design. Self-administered survey. Participants. ICU nurses and physicians in five university-affiliated multidisciplinary ICUs. Results. A total of 317 questionnaires were returned from 233 ICU nurses and 84 physicians. The reported clinically important threshold for hypoglycemia was 4 mmol/l (median, IQR 3-4 mmol/l). In non-diabetic patients, the clinically important threshold for hyperglycemia was 10 mmol/l (IQR 9-12 mmol/l); however, nurses had a significantly higher threshold than physicians (difference of 0.52 mmol/l (95% CI 0.09-0.94 mmol/l, P=0.018). In diabetic patients, the clinically important threshold for hyperglycemia was also 10 mmol/l (IQR 10-12 mmol/l), and again nurses had a significantly higher threshold than physicians (0.81 mmol/l, 95% CI 0.29-1.32 mmol/l, P=0.0023). Avoidance of hyperglycemia was judged most important for diabetic patients (87.7%, 95% CI 84.1-91.3%), patients with acute brain injury (84.5%, 95% CI 80.5-88.5%), patients with a recent seizure (74.4%, 95% CI 69.6-79.3%), patients with advanced liver disease (64.0%, 95% CI 58.7-69.3%), and for patients with acute myocardial infarction (64.0%, 95% CI 58.7-69.3%). Physicians expressed more concern than nurses about avoiding hyperglycemia in patients with acute myocardial infarction (P=0.0004). ICU clinicians raised concerns about the accuracy of glucometer measurements in critically ill patients (46.1%, 95% CI 40.5-51.6%). Conclusions. Attention to these beliefs and attitudes could enhance the success of future clinical, educational and research efforts to modify clinician behavior and achieve better glycemic control in the ICU setting.