Sustainability of a Real-Time Informatics Alert to Prevent Inpatient Severe Hypoglycemia

Inpatient severe hypoglycemia (SH), blood glucose (BG) <40 mgs/dl, is associated with morbidity. We previously piloted a predictive model to identify and intervene in inpatients at risk for SH that decreased the rate of SH by 68%. Expanding the alert system-wide to 9 hospitals, succeeded in reducing the occurrence of severe hypoglycemia from 2.9% to 1.7% per 1000 at risk patient days during all visits, and from 4.4% to 2.5% during visits...