RESUMO
Chronic Kidney Disease (CKD), is highly prevalent in the United States. Epidemiological systems for surveillance of CKD rely on data that are based solely on the NHANES survey, which does not include many patients with the most severe and less frequent forms of CKD. We investigated the feasibility of estimating CKD prevalence from the large-scale community disease detection Kidney Early Evaluation and Program (KEEP, n = 127,149). We adopted methodologies from the field of web surveys to address the self-selection bias inherent in KEEP. Primary outcomes studied were CKD Stage 3-5 (estimated glomerular filtration rate [eGFR] <60 mL/min/1.73 m2, and CKD Stage 4-5 (eGFR <30 mL/min/1.73 m2). The unweighted prevalence of Stage 4-5 CKD was higher in KEEP (1.00%, 95%CI: 0.94-1.05%) than in NHANES (0.51%, 95% CI: 0.43-0.59%). Application of a selection model that used variables related to demographics, recruitment and socio-economic factors resulted in estimates similar to NHANES (0.55%, 95% CI: 0.50-0.60%). Weighted prevalence of Stages 3-5 CKD in KEEP was 6.45% (95% CI: 5.70-7.28%) compared to 6.73% (95% CI: 6.30-7.19%) for NHANES. Application of methodologies that address the self-selection bias in the KEEP program may allow the use of this large, geographically diverse dataset for CKD surveillance.
