| Sumario: | Chronic kidney disease (CKD) is a major public health issue. Estimating CKD prevalence using Electronic Health Records (EHR) of General Practitioners (GPs) is of value to guide patient care and healthcare system policies. Using clinical guidelines, we developed two different CKD electronic phenotypes (e-phenotypes) fit for EHR data from GPs: a laboratory-derived e-phenotype based on the estimated glomerular filtration rate and albumin-to-creatinine ratio, and a diagnosis code-derived e-phenotype. We assessed them based on the CKD prevalence retrieved. Each e-phenotype identified a different number of CKD cases, with the combined approach capturing the largest number of cases. Notably, many laboratory-derived CKD cases lacked a CKD diagnosis code, indicating room for data quality improvement. Our findings highlight the importance of integrating laboratory and diagnostic data in CKD e-phenotype algorithms and provide guidance for researchers, clinicians, and guideline developers to improve CKD detection and care in GP settings.
|
|---|