Sliding Scale in Hospitalized Insulin Treated Patients
Abstract Number: 1945-PO
Authors: BARRETT L. CHAPIN.
Institutions: Laconia, NH.
Results: OBJECTIVE: Sliding scale insulin coverage is commonly used in hospitalized patients with diabetes. Prospective data using this method of glucose management are scarce.
RESEARCH DESIGN AND METHODS: Sixty, insulin treated, hospitalized subjects with Type 2 diabetes mellitus were prospectively, randomized to four groups. Subjects were left on their outpatient insulin regimen (Home Group), placed on a standardized sliding scale insulin protocol using regular insulin (SS/reg Group), the standardized sliding scale regimen using insulin aspart (SS/asp Group), or the standardized sliding scale using regular insulin plus insulin glargine (SS/reg+gla). Data consisted of in-hospital finger stick glucose measurements. Primary endpoints were the frequency of hypoglycemia (glucose < 70 mg/dl) and extreme hyperglycemia (glucose > 350 mg/dl).
RESULTS: The table shows by group the percentage of glucose measurements with (95% Confidence Intervals):
*significantly different from the Home Group at p<0.05
|< 70 mg/dl||70-180 mg/dl||> 180 mg/dl||> 350 mg/dl|
|Home Group||5.9 (3.8-9.4)||45.4 (38.6-53.3)||46.5 (39.6-54.5)||1.9 (0.9-3.8)|
|SS/reg Group||1.0 (0.3-2.8)∗||40.0 (33.7-47.4)||58.8 (50.2-68.8)∗||2.2 (0.8-6.0)|
|SS/asp Group||1.4 (0.7-3.0)∗||34.7 (29.8-40.5)∗||63.1 (56.2-70.9)∗||9.0 (6.4-12.6)∗|
|SS/reg+gla||2.9 (1.6-5.3)||48.0 (41.4-55.8)||46.6 (39.8-54.4)||4.7 (2.8-8.0)∗|
The study was stopped after 50% of planned enrollment when the SS/reg Group had significantly fewer primary endpoints than the other 3 groups (p<0.05).
CONCLUSIONS: The subjects treated with the standardized sliding scale insulin protocol had significantly fewer occurrences of hypoglycemia and extreme hyperglycemia than subjects treated with their usual outpatient insulin regimen, the standardized sliding scale protocol using insulin aspart, or the standardized sliding scale protocol with the addition of insulin glargine.