Detecting Patients with Chronic Kidney Disease Using General Practitioner Electronic Health Records and Electronic...

Area of Science:

• Nephrology
• Public Health
• Health Informatics

Background:

• Chronic kidney disease (CKD) presents a significant global health challenge.
• Accurate prevalence estimation via Electronic Health Records (EHR) is vital for effective patient management and policy development in primary care settings.
• General Practitioners (GPs) are key in early CKD detection and management.

Purpose of the Study:

• To develop and evaluate novel electronic phenotypes (e-phenotypes) for CKD detection within GP EHR data.
• To compare the performance of laboratory-derived and diagnosis code-derived e-phenotypes.
• To assess the utility of integrating multiple data sources for improved CKD case identification.

Main Methods:

• Utilized clinical guidelines to construct two distinct CKD e-phenotypes: one based on estimated glomerular filtration rate (eGFR) and albumin-to-creatinine ratio (ACR), and another using diagnosis codes.
• Applied these e-phenotypes to GP EHR data to estimate CKD prevalence.
• Compared the number of CKD cases identified by each e-phenotype and a combined approach.

Main Results:

• Each e-phenotype identified a different subset of CKD cases, with the combined approach yielding the highest case ascertainment.
• A significant proportion of cases identified by the laboratory-derived e-phenotype did not have a corresponding CKD diagnosis code.
• The findings underscore potential data quality issues and the need for comprehensive data integration.

Conclusions:

• Integrating laboratory results (e.g., eGFR, ACR) and diagnosis codes is essential for developing accurate CKD e-phenotypes.
• Improved CKD detection algorithms in primary care can enhance patient care and inform healthcare policy.
• Guidance is provided for researchers and clinicians to optimize CKD identification in EHR systems.