Reducing hyperglycemia hospitalwide: the basal-bolus concept.

Background: Hyperglycemia has been identified as potent and independent risk factor for adverse outcomes for patients. An initiative was undertaken to reduce hyperglycemia hospitalwide in adults.

Methods: In a multistep process, insulin protocols were implemented hospitalwide via an electronic provider order entry system. Education regarding basal bolus insulin delivery preceded implementation. Protocols were modified in an ongoing manner on the basis of clinical staff feedback and blood glucose monitoring. Key practice changes included intravenous insulin for initial management in ICU patients, insulin replacement based on the basal bolus approach, elimination of sliding-scale insulin, standardization of blood glucose monitoring before meals, adjustment of prandial dose insulin based on food consumed, administration of prandial dose after the meal, evening snacks ordered based on insulin type, and a glycosolated hemoglobin (A1C) determination for patients with admission glucose > 180 mg/dL. Median inpatient glucose levels in patients with diabetes were assessed using statistical process control methodology.

Results: Between January 2004 and September 2007, median glucose for all inpatients with diabetes decreased 15% from 159 mg/dL to < 135 mg/dL. The percentage of inpatients with diabetes who experienced a day with a glucose measurement above 180 mg/dl decreased from 66% to 53%. Frequency of hypoglycemia (< 60 mg/dL) did not change following protocol implementation.

Discussion: Major improvements in hospitalwide blood glucose control are feasible and safe, employing standard protocols based on the basal-bolus concept. Improvement was sustained during a four-year period with ongoing institutional support, multidisciplinary education, collaboration between clinical services, and monitoring of clinical outcomes on a quarterly basis.