Increasing evidence from clinical studies in Intensive Care Unit (ICU) settings indicates that insulin infusion can improve outcome measures for patients with hyperglycemia (high blood sugar) independent of a previous diagnosis of diabetes mellitus. This improvement in health could also apply to patients that have high blood sugars in various other non-critical care areas of the hospital as well. However, the data that shows improvement in health outcomes has been collected from wards that have a lower patient to provider and patient to nurse ratio, resulting in the ability for a much tighter control of the insulin infusion. We hypothesize that tight blood glucose control will provide the same benefits for patients in non-intensive care units settings but that these protocols may lead to a higher incidence of hypoglycemia (low blood sugar) and potentially to adverse outcomes in patients.

This study aims to determine the clinical outcome of patients treated with insulin infusion as well as the rate of hypoglycemic episodes in non-intensive areas. We will conduct a chart review of patients treated with insulin infusions in non-critical wards at Emory University Hospital during the period of 7/1/04 to 6/30/05. Medical records of all patients treated with intravenous insulin infusion protocols will be analyzed. Data on demographics, laboratory values, mortality rate, rate of hypoglycemic events, length of stay, as well as disposition at discharge will be analyzed.

Inclusion Criteria:

• All subjects over 18 years of age that received intravenous insulin treatment on general medical and surgical wards at Emory University Hospital during the period of 7/1/2004 to 6/30/2005

Cohort of hospitalized adult subjects in a teaching institution.