Objective

This retrospective observational study evaluates the impact of using a computer-guided glucose management system (EndoTool) for patients admitted to intensive care and intermediate units, with an analysis of data from a period of seven years, 16,850 visits and 492,078 blood glucose readings.

Setting

900-bed tertiary care teaching hospital.

Results

• Average time to a target blood glucose <=180 mg/dL was between 1.5 and 2.3 hours.
• Average rate of glucose excursions (defined as a blood glucose >180 mg/dL after control was achieved) was 4%, at an average duration of 1.91 hours.
• Incidence of hypoglycemia <40 mg/dL was 0.03% of blood glucose readings.
• Incidence of hypoglycemia <70 mg/dL was 0.93% of blood glucose readings, a rate that decreased 96% over the study period, from 1.04% to 0.046%, concurrent to a near-doubling of visits.
• Among the subset of cardiovascular surgery patients, average time to a target blood glucose <140 mg/dL was between 4.5 and 4.8 hours and was achieved by ~98% of patients.
• Rate of hospital-acquired conditions per 1,000 under HAC-8 (manifestations of poor glycemic control) decreased 61%, from 0.083 pre-EndoTool to 0.032 post-EndoTool, for a ranking in the top 10% of U.S. hospitals.

Conclusions

Use of a computer-guided glucose management system (EndoTool) leads to significant improvements in overall glucose control and quality of care, with relatively rapid time to target, low rate of glucose excursions, minimal hypoglycemia and a marked decrease in the rate of hospital-acquired conditions under HAC-8 (manifestations of poor glycemic control).