Objective: To evaluate the incidence of hypoglycemia, hyperglycemia and blood glucose (BG) variability in brain-injured patients and their association with clinical outcomes.
Methods: Retrospective cohort study of brain-injured patients admitted to an 11-bed neurosciences intensive care unit (ICU) from January 1 to December 31, 2003.
Results: We included 606 patients. Mean age was 52.3 years, 60.6% were male, 11.9% had diabetes mellitus, and 64% were post-operative. Seventy-five (12.4%) received intensive insulin therapy (IIT) for a median (IQR) 72 (24-154) hours. Hypoglycemia and hyperglycemia occurred in 4.6% (96.4% receiving IIT) and 9.6% (77.6% receiving IIT). Median number of episodes per patient was 3 (75% with > or = 2) and 4 (81% with > or = 2) for hypoglycemia and hyperglycemia. Variable glycemic control occurred in 3.8% (100% receiving IIT) with median number of 13 episodes per patient. In-hospital mortality was 16.7%, median (IQR) ICU and hospital lengths of stay were 2 (1-5) and 8 (3-19) days. Hypoglycemia, hyperglycemia and BG variability showed non-significant but consistent associations with hospital mortality and prolonged lengths of ICU and hospital stay. The rate of recurrence of episodes showed stronger and significant associations with outcome, in particular for BG variability and hyperglycemia.
Conclusions: Hypoglycemia, hyperglycemia and BG variability are relatively common in brain-injured patients and are associated with IIT. An increased frequency of episodes, in particular for BG variability and hyperglycemia, was associated with greater risk of both hospital death and prolonged duration of stay.