Electronic Phenotype for Advanced Chronic Kidney Disease in a Veteran Health Care System Clinical Database:...

Area of Science:

• Nephrology
• Health Informatics
• Clinical Research

Background:

• Accurate identification of advanced (stages 4-5) chronic kidney disease (CKD) in clinical databases is often unreliable.
• Targeting this patient population is crucial for specialized clinical care and research initiatives.
• Existing methods for identifying advanced CKD in electronic health records (EHRs) have limitations.

Purpose of the Study:

• To develop and validate a system-based strategy for identifying prevalent Veterans with advanced CKD.
• To examine the accuracy of conventional diagnosis codes and estimated glomerular filtration rate (eGFR)-based phenotypes for advanced CKD in EHRs.
• To create a pragmatic EHR phenotype for real-time identification of advanced CKD cohorts within a Veterans health care system.

Main Methods:

• Extracted a cohort of Veterans with advanced CKD using a combination of latest eGFR ≤30 ml·min⁻¹·1.73 m⁻² or ICD-10 codes (N18.4, N18.5) within 12 months.
• Estimated advanced CKD prevalence using prior EHR phenotypes (diagnosis codes, single eGFR <30) and developed operational phenotypes (high-, intermediate-, low-risk).
• Evaluated phenotype accuracy by assessing the likelihood of sustained eGFR <30 ml·min⁻¹·1.73 m⁻² over a 6-month follow-up.

Main Results:

• Identified 1759 Veterans with advanced nondialysis CKD; prevalence varied from 1% to 1.5% based on EHR phenotype.
• Diagnosis codes identified 62.9%, while index eGFR <30 identified 79.1%; high-, intermediate-, and low-risk groups comprised 52.7%, 27.2%, and 17.9% respectively.
• The new phenotype showed high accuracy: 94.2% of high-risk, 71% of intermediate-risk, and 16.1% of low-risk groups maintained advanced CKD status at 6-month follow-up.

Conclusions:

• While CKD prevalence shows minor variation across EHR phenotypes, accuracy improves with stratified eGFR values.
• Developed a pragmatic EHR-based model for real-time identification of advanced CKD in Veterans.
• The tiered approach effectively targets patient groups at risk for progression to end-stage kidney disease.