Intensive versus conventional insulin therapy in critically ill neurologic patients.

BACKGROUND: Previous studies of glycemic control in non-neurologic ICU patients have shown conflicting results. The purpose was to investigate whether intensive insulin therapy (IIT) to keep blood glucose levels from 80 to 110 mg/dl or conventional treatment to keep levels less than 151 mg/dl was associated with a reduction of mortality and improved functional outcome in critically ill neurologic patients. METHODS: Within 24 h of ICU admission, mechanically ventilated adult neurologic patients were enrolled after written informed consent and randomized to intensive or conventional control of blood glucose levels with insulin. Primary outcome measure was death within 3 months. Secondary outcome measures included 90-day modified Rankin scale (mRS) score, ICU, and hospital LOS. RESULTS: 81 patients were enrolled. The proportion of deaths was higher among IIT patients but this was not statistically significant (36 vs. 25%, P = 0.34). When good versus poor outcome at 3 months was dichotomized to mRS score 0-2 versus 3-6, respectively, there was no difference in outcome between the two groups (76.2 vs. 75% had a poor 3-month outcome, P = 1.0). There was also no difference in ICU or hospital LOS. Hypoglycemia (<60 mg/dl) and severe hypoglycemia (<40 mg/dl) were more common in the intensive arm (48 vs. 11%, P = 0.0006; and 4 vs. 0%, P = 0.5, respectively). CONCLUSION: There was no benefit to IIT in this small critically ill neurologic population. This is the first glycemic control study to specifically examine both critically ill stroke and traumatic brain injury (TBI) patients and functional outcome. Given these results, IIT cannot be recommended over conventional control.