Developing a Computable Phenotype for Identifying Children, Adolescents, and Young Adults With Diabetes Using...

Area of Science:

• Diabetes research
• Public health surveillance
• Health informatics

Background:

• The Diabetes in Children, Adolescents, and Young Adults (DiCAYA) network aims to establish a national electronic health record (EHR)-based surveillance system for diabetes.
• Accurate identification of prevalent diabetes cases is crucial for effective public health monitoring and intervention.

Purpose of the Study:

• To develop and validate a DiCAYA-wide EHR-based computable phenotype (CP) for identifying prevalent diabetes cases in youth and young adults.
• To assess the accuracy of the CP in identifying diabetes and its subtypes (Type 1 and Type 2 Diabetes).

Main Methods:

• Conducted network-wide chart reviews of 2,134 youth and 2,466 young adults with suspected diabetes.
• Developed and compared three alternative CPs, using chart review diagnoses as the gold standard.
• Evaluated CP performance using sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV).

Main Results:

• The final DiCAYA CP requires at least one diabetes diagnosis code and uses diagnostic code ratios for subtype classification.
• The CP demonstrated >90% sensitivity, specificity, PPV, and NPV for overall diabetes detection in both age groups (with slightly lower metrics for young adults).
• The CP achieved >90% accuracy for Type 1 Diabetes (T1D) classification and >80% for Type 2 Diabetes (T2D) identification.

Conclusions:

• The DiCAYA CP is effective for identifying overall diabetes and T1D in youth and young adults.
• Refinements are needed for T2D misclassification in youth.
• The CP's simplicity facilitates widespread implementation across various EHR systems for enhanced diabetes surveillance.