Single estimated glomerular filtration rate and albuminuria measurement substantially overestimates prevalence of chronic kidney disease.

Background/aims: Guidelines require repeatedly diminished estimated glomerular filtration rate (eGFR) and/or albuminuria to diagnose chronic kidney disease (CKD), and advise screening only in select populations. Many estimates of CKD prevalence have used single measurements. This longitudinal study assessed eGFR and albuminuria reproducibility, and impact on estimate of CKD prevalence, in factory workers.
Methods: A total of 512 white workers in a Belarusian industrial factory were initially tested, identifying 206 with abnormal (eGFR <59 ml/min/1.73 m(2) or albuminuria) or near-abnormal (eGFR up to 1 SD above abnormal) renal function. At 3 months, 142 of the abnormal/near-abnormal cohort were re-tested.
Results: Analysis of repeat samples revealed no significant change in eGFR in this population, however 21% individually changed CKD stage. Initial proteinuria was reproducible in only 48% at 3 months. This had a major impact on estimated CKD prevalence: a point prevalence of 8.2% halved with repeat testing. The predictive value of initially abnormal eGFR or albuminuria for repeat abnormality at 3 months was 0.5.
Conclusion: Non-targeted screening for CKD is inaccurate and can overestimate prevalence. This study emphasises the importance of confirming abnormal eGFR and proteinuria on at least one further sample 3 months apart before categorising the individual as having CKD. This has wide implications for screening in European general populations.
(Copyright © 2010 S. Karger AG, Basel.)