• Percentage of blood glucose values within the target range of eighty to one hundred forty mg per dL
  

• Percentage of blood glucose values below fifty mg per dL.
  
• Percentage of blood glucose values above two hundred mg per dL
  

Critical illness causes an impairment of insulin secretion and insulin action, resulting in hyperglycemia even in normal individuals and a worsening of the hyperglycemia in patients with diabetes. Normalization of elevated glucose levels by intensive insulin infusion therapy in these critically ill patients has been proven to dramatically improve in-hospital mortality rates. After glucose levels have been controlled with insulin infusions, the best way to convert them to subcutaneous insulin regimens has not been demonstrated conclusively but the insulin regimen best suited is a combination of a basal insulin such as glargine (Lantus) insulin with premeal insulin boluses using a short-acting insulin such as Lispro or Aspart.

Subjects will be randomized into three groups, 25 subjects in each group, the groups differing according to the starting dose of glargine insulin as follows:

1. Multiply the total daily IV insulin dose, using the final 6 hour infusion rate to estimate the total daily dose, by 0.4 to get the starting dose of glargine, continuing the insulin drip for 5 hours after giving the glargine before stopping the infusion;
2. Multiply the total daily IV insulin dose, using the final 6 hour infusion rate to estimate the total daily dose, by 0.6 to get the starting dose of glargine, continuing the insulin drip for 5 hours after giving the glargine before stopping the infusion;
3. Multiply the total daily IV insulin dose, using the final 6 hour infusion rate to estimate the total daily dose, by 0.8 to get the starting dose of glargine, continuing the insulin drip for 5 hours after giving the glargine before stopping the infusion.

The glargine insulin will then be continued as the basal insulin, adjusting doses every 24 hours based on the fasting blood glucose level. In addition, patients will receive Lispro or Aspart insulin as prandial bolus insulins with the goals premeal being 80 - 120 mg/dl, including bedtime. Glucose measurements will be obtained every 1- 4 hours while patients are on their insulin infusions, depending upon the stability of their condition and the stability of their glucose levels. Following transfer to glargine insulin, they will have glucose measured at least four times per day, premeal and bedtime as per standard protocol in the hospital for patients receiving insulin. Glucose data will be obtained from such patients and this will be compared among the various glargine regimens.