Objective

This retrospective study evaluates use of EndoTool to achieve tight glucose control for critical care patients receiving intravenous insulin and constant nutrition.

Setting

874-bed regional academic medical center.

Results

• Roughly half of blood glucose readings <40 mg/dL were associated with procedural errors, either a check more than 30 minutes late or no prior intravenous insulin infusion.
• Aggregate incidence of hypoglycemia <40 mg/dL was 0.11% of blood glucose readings and 2.9% of patients; when excluding procedural errors, it was 0.05% of blood glucose readings and 1.6% of patients.
• Mean blood glucose was 119.5 mg/dL (SD 35.9), and median blood glucose was 116 mg/dL.
• Incidence of hypoglycemia, mean blood glucose, nosocomial infection and length of stay decreased over the study period, given use of EndoTool.

Conclusions

The clinical benefits of tight glucose control in critical care units can be safely achieved with use of EndoTool software.