Objective: To improve control of blood glucose concentrations in critically ill patients through use of a bedside, nurse-managed, intravenous insulin nomogram.
Design: Retrospective, before-after cohort study.
Setting: Fifteen-bed mixed medical/surgical intensive care unit in a tertiary, teaching hospital.
Patients: A total of 167 intensive care unit patients requiring intravenous insulin infusions during two 9-month periods.
Intervention: The sliding scale group was treated using ad hoc sliding scale infusion therapy. The intervention group was treated using a dosing nomogram that allowed the nurse to adjust the insulin infusion rate based on current glucose concentration and concurrent insulin infusion rates. The adjustments were made independent of physician input.
Measurements and main results: Time from initiating the insulin infusion to initial control of glucose concentration (<11.5 mmol/L) was determined. Effectiveness of glucose control was determined retrospectively by measuring the area under the curve of blood concentrations >11.5 mmol/L versus time of insulin infusion, divided by total duration of insulin infusion. The median time to initial control of glucose (<11.5 mmol/L) was 4 hr (range 1-38 hr) for the baseline and 2 hr (range 1-22 hr) for nomogram group (p =.0004). The median area under the curve of glucose concentration divided by duration of insulin infusion was 0.9 (range 0.0-5.9) for sliding scale group and 0.3 (range 0.0-11.1) for nomogram (p =.0001), without any increase in the frequency of episodes of hypoglycemia.
Conclusion: Use of an insulin nomogram in critically ill patients improves control of blood glucose concentrations and is safe.