The hypothesis of this study is that using hospital admission to identify patients with poorly controlled diabetes (hemoglobin A1c levels >8%), and intervening during the hospitalization with targeted inpatient diabetes management will improve glycemic control at 3 and 12 months, with inpatient glycemic control, quality of life, and diabetes self-efficacy serving as secondary endpoints.

The study is a randomized, controlled trial of targeted inpatient diabetes management versus usual care in 70 general medical and surgical inpatients over age 18 with type 2 diabetes and hemoglobin A1c levels greater than 8% who are followed in Partners-affiliated primary care practices. In addition, we will enroll a prospective chart review cohort matched for age, sex, and HbA1c level whose course we will follow for one year after discharge to determine the patter of glycemia among patients who are not enrolled in a clinical trial.

Inclusion Criteria:

• Age > 18 years
• Known diagnosis of type 2 diabetes by history with outpatient prescription of oral hypoglycemic medication or insulin
• Hemoglobin A1c > 8.0% within the prior 12 months, or if not known, fasting blood glucose greater than 200 mg/dl on sliding scale regular insulin.
• Partners-affiliated primary care physician

Exclusion Criteria:

• Screening HbA1c returns less than 8%.
• Diabetic ketoacidosis (DKA) as a primary reason for admission (admission blood glucose > 250 mg/dl with arterial pH < 7.30 or serum bicarbonate level < 15 mg/dl), or development of DKA during admission.
• Hyperosmolar hyperglycemic syndrome as a primary reason for admission (admission blood glucose > 400 mg/dl and plasma osmolality > 315 mOsm/kg.
• Pregnancy, ruled out by urine HCG test at screening after consent is obtained in all women who continue to have menstrual cycles.
• Anemia with hemoglobin < 9 g/dl, recent blood transfusion, or need for blood transfusion (interferes with interpretation of hemoglobin A1c assay)
• End stage liver disease with prothrombin time > 15 seconds and albumin <3 mg/dl
• End stage renal disease: Stage IV (glomerular filtration rate <30 mg/dl) or V chronic kidney disease
• Treatment with corticosteroids
• ICU transfer
• Inability to self-administer insulin
• Hypoglycemia unawareness: patient lacks sensation of common signs of blood glucose < 60 mg/dl (tachycardia, diaphoresis, hunger, confusion, fatigue).
• Projected survival < 1 year