Use of an Electronic Medical Record Integrated Tool to Identify Patients at Highest Risk of Hypoglycemia

Inpatient hypoglycemia can increase patient length of stay, readmission risk, and mortality. Therefore, incidence of hypoglycemia after administration of an antidiabetic medication is a Centers for Medicare & Medicaid Services Quality Measure. Retrospective studies have identified factors associated with risk of hypoglycemia. Risk prediction tools can identify patients at high risk of hypoglycemia.

An electronic medical record-integrated hypoglycemia prediction tool was developed using known hypoglycemia risk factors to prospectively identify patients at high risk for hypoglycemia. Pharmacists used the tool and dashboard to identify high-risk noncritically ill inpatients in daily collaboration with the medical team. This prospective, single-center interventional cohort study included 809 control and 839 intervention group patients.

Occurrence of hypoglycemia in the low-risk control group was 16.1% (P = .001) compared with high-risk control. The occurrence of any hypoglycemia (blood glucose <70 mg/dL) among high-risk patients in the intervention group was 22.4% compared with 26.3% in the high-risk control group (P = .064). Use of the tool for clinical decision support resulted in a 46% increase in insulin therapy adjustments in high-risk patients. There was no increase in hyperglycemia across study groups.

An electronic medical record-integrated hypoglycemia prediction tool can accurately identify patients at high risk for hypoglycemia in the next 24 hours. Use of a risk assessment tool can increase insulin dose adjustments in high-risk patients.