Background: This study aimed to analyze the relation of hyperglycemia to outcome in cases of severe traumatic brain injury, and to examine factors that may be responsible for the hyperglycemic state.
Methods: A retrospective analysis in an intensive care unit of a level 1 trauma center investigated 77 patients with severe traumatic brain injury. Patients with a Glasgow Coma Scale (GCS) of 8 or lower who survived more than 5 days were reviewed. Serum glucose, base deficit, GCS, use of steroids, and amounts of insulin and carbohydrates were recorded for 5 days, along with age. The Injury Severity Score (ISS) and the Abbreviated Injury Score (AIS) for the head, chest, and abdomen also were recorded. A hyperglycemia score (HS) was calculated as follows. A value of 1 was assigned each day the glucose exceeded 170 mg/dL (range, 0-5). A hyperglycemia score for days 3, 4, and 5 (HS day 3-5) also was calculated (range, 0-3). Outcomes included mortality, day 5 GCS, intensive care unit length of stay, and hospital length of stay.
Results: Of the 77 patients, 24 (31.2%) died. Nonsurvivors had higher glucose levels each day. The HS was higher for those who died: 2.4 +/- 1.7 versus 1.5 +/- 1.4 (p = 0.02). Univariate analysis showed that only HS and ISS correlated with all four outcome variables studied. Cox's regression analysis showed that mortality was related to age and ISS. Head AIS and HS were independent predictors of lower day 5 GCS, whereas HS 3-5 and day 4 GCS were related to prolonged hospital length of stay. Older age, diabetes, and lower day 1 GCS were associated with higher HS, whereas carbohydrate infusion rate, ISS, head AIS, and steroid administration were not.
Conclusions: Early hyperglycemia is associated with poor outcomes for patients with severe traumatic brain injury. Tighter control of serum glucose without reduction of nutritional support may improve the prognosis for these critically ill patients.