Effectiveness of Sliding-Scale Insulin in Inpatients with Diabetes Mellitus

Objective: To compare the incidence of hypoglycemia and hyperglycemia among sliding-scale insulin (SSI) regimens and determine patient characteristics that influence glycemic control in inpatients with diabetes mellitus. Methods: Main outcome measures for this retrospective chart review were rates of hyperglycemic (≥200 mg/dL) and hypoglycemic (≤60 mg/dL or symptomatic) episodes. Patients were excluded if a diabetic complication led to hospitalization. SSI regimens as well as patient demographics and clinical characteristics were evaluated for influence on glycemic control. Results: Seventy-four patients with 1020 capillary glucose measurements were analyzed. Regarding overall glycemic control, 80% of patients had ≥2 hyperglycemic events and 4% had ≥2 hypoglycemic events. Factors that significantly increased the risk of hyperglycemia were insulin use prior to admission (p ≤ 0.0001), corticosteroid use (p ≤ 0.0001), antibiotic use (p = 0.009), and dextrose-containing intravenous fluids (p = 0.004). Implementation of an SSI order form (which was associated with selection of more aggressive SSI regimens) significantly decreased the incidence of hyperglycemia (p ≤ 0.0001) without increasing the incidence of hypoglycemia. When analyzed for interactions among variables, insulin use prior to admission and concurrent corticosteroid use remained as independent predictors of hyperglycemia. Conclusions: Suboptimal glycemic control is common in hospitalized patients with diabetes. Patients receiving insulin prior to admission, corticosteroid therapy, dextrose-containing intravenous fluids, or treatment for an active infection are at increased risk. An aggressive SSI regimen does not appear to increase the incidence of hypoglycemia and may lead to better glycemic control. Whether initiating treatment at blood glucose concentrations <150 mg/dL would improve control remains unknown.