Good sugar control in postoperative hospitalized patient has been shown to improve wound healing and infection rates. However, sugar control is difficult to achieve and suboptimal use of insulin is thought to be a contributory factor. Though it is known that generally the consumption of carbohydrates alone raises the blood sugar, the usual practice of dosing meal-time insulin is based on the fraction of the total meal-tray eaten which includes proteins, fats and carbohydrates. This leads to an overestimation of insulin required for a patient who consumes a portion of mainly proteins and fats on their trays or an underestimation for those eating mainly the carbohydrates on their tray. Low sugars or high blood sugars can follow respectively.

Hypothesis: The purpose of this study is to see if dosing meal-time insulin based on grams of carbohydrates consumed will result in better sugar control compared to the usual practice of dosing meal-time insulin based on percent of total meal consumed in hospitalized patients.

100 research subjects are expected to be enrolled in this study, all recruited from the Rush University Medical Center surgical (non-critical) services.

As is the usual care at Rush University Medical Center, patients who have type 2 diabetes and are on insulin and or at least two oral diabetes medications are given only insulin during their inpatient stay to control their blood sugar. It is the normal practice to use only insulin the hospital instead of pills for treating your diabetes. The use of insulin would be temporary and just during your hospital stay. When you return home, you will resume using pills.

Before the study begins, a computer system will randomly assign you a number, and those numbers belong to one of two groups. The number you have been given will determine which group you will go to. Throughout the study, you will be given a standard dose of basal insulin (Lantus). This dose will be your home dose if you are already using insulin. If you are not using insulin, you will be given a dose based on your weight.

During your stay, you will be given meal trays for breakfast, lunch, and dinner, and each of the food items belonging to your meal tray will be marked with at a certain number of carbohydrates. For example, ½ cup of cottage cheese will contain 4 grams of carbohydrates and one slice of whole wheat bread will contain 14 grams of carbohydrates and one 5 oz apple will contain 21 grams of carbohydrates. A dietician will then count the total amount of carbohydrates that you consume with each meal.

If you are assigned to the investigational arm of the study, you will be given a dose of mealtime insulin (Aspart) based on the number of carbohydrates you have consumed. If you are assigned to the control arm, you will be given your standard dose of mealtime insulin, if you eat more than 50% of your tray as is the usual practice at Rush, regardless of how much carbohydrate you eat. This dose will be your home dose if you are already using insulin or calculated based on your weight.

In either arm, if meals are not given (due to additional procedures or other reasons) the insulin dose will not be given. For both arms of the study, you will be requested not to eat any snacks containing carbohydrates. Instead you may be request for snacks that do not contain carbohydrates such string cheese which the investigators will provide.

Response to these therapies will be monitored by fingerstick glucose readings, also called point-of-care glucose. Your nurses will draw capillary blood, (less than 1 ml) via fingerstick on 4 occasions: pre-breakfast, pre-lunch, pre-dinner, and post-dinner / bedtime. These values will be sent to the study investigators, who will then adjust the amount of insulin that you are given.

In addition, your doctors will check a blood test called an A1c. This test measures the average sugar level in your blood over the last 90 days. This test is done on all patients with diabetes who are hospitalized.

Inclusion Criteria:

• Adults aged > or = to 18 admitted to general surgical floors, excluding the ICU, at Rush University Medical Center, regardless of race, ethnicity, gender
• Clinical diagnoses of type 2 diabetes for more than 6 months prior to admission
• Treated with insulin and/or 2 or more oral diabetic agents
• Estimated length of stay 3 days or more
• Postoperative point of care blood glucose of > 180 mg/dL

Exclusion Criteria:

• Glomerular Filtration Rate < 60 based on MDRD equation
• Pregnant patients
• Receiving parenteral or enteral nutrition
• Patients with an admitting diagnosis of hypoglycemia
• Outpatient insulin < 0.5 units/kg/day
• Inability to give consent
• Severe liver disease
• Known hypopituitarism or adrenal insufficiency
• Treatment with Prednisone at dose > 5 mg daily or its equivalent

The study group had 8 subjects because due to randomization the study was selected for the next draw.

3 months

Early termination due to the PI becoming sick early in the study.