Comparison of Apidra to Regular Insulin in Hospitalized Patients

OBJECTIVES: To compare the rapid acting insulin analogue Apidra with regular insulin in addition to insulin Glargine in hospitalized patients in terms of efficacy and safety, namely Glycemic control and frequency of hypoglycemia. Glycemic control, and incidence and rate of hypoglycemia during the hospitalization shall be the primary endpoints; length of hospital stay according to the short acting insulin used shall be the secondary endpoint.

RESEARCH DESIGN: Randomized, prospective study.

METHODS: Inpatient single center study, planning to enroll 600 patients with type II diabetes admitted to medical or surgical non-ICU service for three days or longer. Subjects will be randomized to Apidra or regular insulin in a 1:1 fashion. Insulin Glargine will be given once a day for basal insulin in all subjects. An algorithm to determine the initial doses of insulin and dose adjustments is as follows: Lean subjects (BMI less than 25 kg/m2) will initially receive a total of 0.4 units/kg/day, overweight subjects (BMI 25-30 kg/m2) 0.5 units/kg/day and obese subjects (BMI greater than 30 kg/m2) 0.6 units/kg/day. Fifty percent of the total amount of insulin will be given as Glargine and 50% as regular insulin or Apidra. Supplemental short-acting insulin will be given for hyperglycemia before meals. Automated order sets shall be generated to minimize errors in order entries. Glucose concentrations will be measured before each meal and at bedtime, and if symptomatic. In addition, eight-point blood glucose profiles will be obtained every three days starting on day 2. Dose adjustments will be made to keep blood glucose concentrations between 80 and 120 mg/dl pre-prandially and less than180 mg/dl after meals. In addition, HbA1c, lipid profile and a fasting plasma C-peptide will be obtained. Two days prior to the anticipated discharge, another HbA1c will be done.

The incidence and the rate of hypoglycemia in each category shall be determined. During the hospitalization, the average of blood glucose measurements at each time point of an 8-point blood glucose profile will be compared; after the hospitalization the HbA1c shall be used. Glycemic control will be compared between groups using ANCOVA adjusting for baseline HbA1c. Hypoglycemic events will be compared between groups using logistic or Poisson regression; length of stay will be compared between groups using survival analysis or the Mann Whitney U test.