Validating an Electronic Health Record Algorithm for Diabetes Screening Eligibility in the Emergency Department.

Area of Science:

• Clinical Medicine
• Endocrinology
• Public Health

Background:

• American Diabetes Association (ADA) screening guidelines are widely used but their implementation varies.
• Adapting guidelines to specific settings impacts practical application and usage.
• Identifying patients eligible for hemoglobin A1c (HbA1c) testing in emergency departments (EDs) can be challenging.

Purpose of the Study:

• To validate a best practice advisory (BPA) screening algorithm informed by ADA guidelines.
• To identify patients eligible for HbA1c testing in the ED.
• To assess the algorithm's accuracy in detecting potential undiagnosed diabetes and prediabetes.

Main Methods:

• Cross-sectional study of adults in an urban ED in May 2021.
• Utilized sensitivity, specificity, likelihood ratios, and predictive values for algorithm validation.
• Manual chart review served as the reference standard; Area Under the Curve (AUC) was calculated.

Main Results:

• The BPA algorithm triggered in 27% of eligible ED patients, with 79% completing an HbA1c test.
• The algorithm demonstrated acceptable sensitivity (0.69) and high specificity (0.91).
• Positive predictive value was 0.75, negative predictive value was 0.88, and AUC was 0.74.

Conclusions:

• An EHR-based algorithm using ADA guidelines is a valid tool for ED-based HbA1c testing eligibility.
• The BPA algorithm effectively identifies patients with potential undiagnosed diabetes or prediabetes.
• Workflow integration and high yield make the BPA algorithm suitable for ED diabetes screening.