Objective: Evaluation of the effects of tight glycemia control in critically ill patients should include temporal as well as punctual glycemia data.
Methods: Insulin drip was used to target intensive care unit (ICU) glucose levels between 80 and 126 mg dl⁻¹ in a consecutive series of adult cardiac surgery patients. ICU hourly glycemia was prospectively recorded. Glycemia standard deviation, hyperglycemia index (area under the curve for glycemia>126 mg dl⁻¹ divided by total hours in ICU), and hypoglycemic episodes were recorded and analyzed, together with outcomes.
Results: A total of 596 patients were included. Hypoglycemia occurred in 21% of the patients. In-hospital mortality was 2.6%. There was a univariate correlation between mortality and glycemia standard deviation, and hypoglycemia occurrence. At multivariate analysis, hypoglycemia was a determinant for mortality (p=0.002; odds ratio (OR)=20.0), respiratory failure (p=0.0001; OR=1.4), requirement of a tracheostomy (p=0.0001; OR=21.6), and hemodynamic instability requiring intra-aortic balloon pump (IABP) (p=0.01; OR=1.5). To clarify the determinants of hypoglycemia, a second multivariate model was built. Diabetes (p=0.0001; OR=23) and chronic renal failure (p=0.01; OR=25) were the sole determinants for hypoglycemia occurrence.
Conclusion: Iatrogenic hypoglycemia secondary to ICU tight glycemia control correlates with hospital mortality, respiratory, and cardiac morbidity in patients undergoing cardiac surgery. ICU hyperglycemia index and glycemia temporal variability have no independent correlation with outcomes. Higher glycemia targets should be advised in the perioperative management of patients with diabetes and renal failure, as both conditions independently increase the risk of hypoglycemia occurrence.
Copyright © 2010 European Association for Cardio-Thoracic Surgery. Published by Elsevier B.V. All rights reserved.