Characterization of Symptoms and Symptom Clusters for Type 2 Diabetes Using a Large Nationwide Electronic Health...

Area of Science:

• Endocrinology and Metabolism
• Health Informatics
• Clinical Epidemiology

Background:

• Type 2 diabetes presents with diverse symptoms, often occurring in clusters.
• Understanding symptom prevalence, severity, and temporal patterns is crucial for managing chronic diseases.
• Electronic health records (EHRs) offer a valuable resource for studying large patient populations.

Purpose of the Study:

• To characterize type 2 diabetes-related symptoms using a nationwide EHR database.
• To identify symptom clusters and their temporal dynamics in type 2 diabetes patients.
• To uncover both common and rare symptoms associated with type 2 diabetes.

Main Methods:

• Utilized the Cerner Health Facts nationwide EHR database.
• Identified a cohort of 1,136,301 type 2 diabetes patients using a rule-based phenotype.
• Employed a multistep procedure with ICD codes to identify 59 type 2 diabetes-related symptoms and symptom clusters.

Main Results:

• Analyzed data from 1,136,301 patients, identifying 8,008,276 symptom occurrences.
• Most frequent symptoms included pain, heartburn, shortness of breath, fatigue, and swelling (21-60% prevalence).
• Over-represented symptoms included difficulty speaking, confusion, memory problems, weakness, and drowsiness, some rarely captured in patient-reported outcomes.

Conclusions:

• This is the first study to characterize type 2 diabetes symptoms and temporal patterns using a nationwide EHR.
• Identified 59 distinct symptoms, encompassing both common and rare manifestations of type 2 diabetes.
• Findings provide a comprehensive overview of the symptom landscape in type 2 diabetes.