Background: Severe hypoglycemia caused by glucose-lowering medications is a known public health and patient safety issue. Identifying patients with T2D at risk of severe hypoglycemia might facilitate interventions to offset that risk.
Study design: Prospective cohort.
Setting: Kaiser Permanente Northern California (derivation and internal validation cohort); Veterans Affairs Diabetes Epidemiology Cohort and Group Health Cooperative (external validation cohorts).
Synopsis: Through EHR data, 206,435 eligible patients with T2D were randomly split into derivation (80%) and internal validation (20%) samples. EHR data were reviewed for preselected clinical risk factors for hypoglycemia with a primary outcome of ED visit or hospital admission with a primary diagnosis of hypoglycemia over the ensuing year. A predictive tool was built based on six variables: prior hypoglycemia episodes, number of ED encounters for any reason in the prior year, insulin use, sulfonylurea use, presence of severe or end-stage kidney disease, and age. Predicted 12-month risk was categorized as high (greater than 5%), intermediate (1%-5%) or low (less than 1%). In the internal validation sample, 2.0%, 10.7%, and 87.3% of patients were categorized as high, intermediate, and low risk, respectively. Observed 12-month hypoglycemia-related health care utilization rates were 6.7%, 1.4%, and 0.2%, respectively. The external validation cohorts performed similarly.
Bottom line: A simple tool using readily available data can be used to estimate the 12-month risk of severe hypoglycemia in patients with T2D.
Citation: Karter AJ et al. Development and validation of a tool to identify patients with type 2 diabetes at high risk of hypoglycemia-related emergency department or hospital use..
Dr. Hoegh is a hospitalist at the University of Colorado School of Medicine.