Insulin infusions are commonly used in hospitalized diabetics to control blood sugar, and
they are effective. However, insulin infusions require the use of limited resources.
Insulin infusions are therefore changed to insulin shots as a patient recovers. Once an
insulin infusion is stopped and shots are started, blood sugar control is harder to
maintain. This is, in part, because physicians have different ideas on how to dose insulin
shots in hospitalized patients. A math equation has been developed by the research staff
that attempts to predict the effective doses of insulin shots in patients whose insulin
infusion have just been stopped. The math equation was developed for patients with type 2
diabetes mellitus. In this study, all patients will be treated with the same type of
insulin shots, with doses of the insulin shots chosen either by the math equation or by the
judgment of the patient's physician. The study will then follow blood sugar values for 24
hours to see if the math equation is effective. If the equation is proven to be effective,
a new tool will exist for physicians to determine the best dose of insulin shots for type 2
diabetics. Such a tool would, in turn, allow for widespread use of insulin infusions to
determine a patient's insulin needs before discharge from the hospital. Blood sugar control
for type 2 diabetics that are inpatient or outpatient would improve as a result, with
potentially far reaching public health benefits.
- Inpatients at an urban, mixed academic and community tertiary care hospital who were
on IV insulin were enrolled.
- Patients were taken from medical, general surgical, and cardiothoracic services, and
were located both inside and outside the intensive care unit (ICU).
- At time of enrollment, patients with type I diabetes mellitus,
- active acute or chronic pancreatitis,
- history of pancreatic surgery,
- use of a self-titratable insulin pump, or
- history of β-islet cell transplantation were excluded.
- At time of randomization, patients with insulin drip rates ≤ 2 units/hr, ∆ in serum
creatinine of > 20% in previous 24 hours, or
- those without caloric intake while on IV insulin were excluded.