Another look at glycemic control in stroke patients

Introduction Although hyperglycemia is associated with poor outcomes in patients with cerebrovascular stroke, there is no consensus about the optimal glycemic control strategy in these patients. This study compares the effects of an intensive intravenous insulin regimen with a conventional subcutaneous one on the degree of glycemic control and neurological outcome in stroke patients admitted to the intensive care unit.Patients and methods Sixty stroke patients were randomized into either an intensive intravenous insulin therapy group (target blood glucose: 90–120 mg/dl) or a conventional subcutaneous insulin group (target blood glucose 150–180 mg/dl). Comparisons were done between groups as regards blood glucose levels, length of ICU and hospital stays, mortality rate, change in National Institute of Health Stroke Scale (NIHSS) score, incidence of infection, and incidence of hypoglycemia.Results The length of ICU and hospital stays in the intensive group were13.51±12.35 and 19.04±12.4 days, respectively, compared with 13.32±14.06 and 18.69±14.83 days in the conventional group (P>0.05), whereas the ICU and hospital mortality rates in the intensive group were 26.66 and 40%, respectively, compared with 23.33 and 36.66% in the conventional group (P>0.05). Mean blood glucose levels were lower in the intensive group (132.1±38.4 mg/dl) compared with the conventional group (156.24±42.1 mg/dl) (P<0.05). Incidence of hypoglycemia was higher in the intensive group (23.33%) compared with the conventional group (6.66%) (P<0.05). There was no significant differences between both groups as regards incidence of infection and NIHSS score change (P>0.05)Conclusion Intensive insulin therapy in stroke patients was associated with better glucose control at the expense of increased incidence of hypoglycemia compared with a conventional insulin regimen. Both strategies did not influence the length of ICU and hospital stays, mortality rates, incidence of infection, and NIHSS score change.