KDIGO 2012 Clinical Practice Guideline CKD classification rules out creatinine clearance 24 hour urine collection?

OBJECTIVES: The recent guideline for the evaluation and management of Chronic Kidney Disease recommends assessing GFR employing equations based on serum creatinine; despite this, creatinine clearance 24-hour urine collection is used routinely in many settings. In this study we compared the classification assessed from CrCl (creatinine clearance 24h urine collection) and e-GFR calculated with CKD-EPI or MDRD formulas.

DESIGN AND METHODS: In this retrospective study we analyze consecutive laboratory data: creatinine clearance 24h urine collection, serum creatinine and demographic data such as sex and age from 15,777 patients >18 years of age collected from 2011 to 2013 in our laboratory at Careggi Hospital. The results were then compared to the estimated GFR calculated with the equations according to the recent treatment guidelines. Consecutive and retrospective laboratory data (creatinine clearance 24h urine collection, serum creatinine and, demographic data such as sex and age) from 15,777 patients >18 years of age seen at Careggi Hospital were collected.

RESULTS: Comparison between e-GFR calculated with CKD-EPI or MDRD formulas and GFR according CrCl determinations and bias [95% CI] were 11.34 [-47,4/70.1] and 11.4 [-50.2/73] respectively. The concordance for 18/65 years aged group when compared with e-GFR classification between MDRD vs CKDEPI, MDRD vs CrCl and CKD-EPI vs CrCl were 0.78, 0.34, and 0.41 respectively, while in the 65/110years aged group the concordance Kappas were 0.84, 0.38, and 0.36 respectively.

CONCLUSIONS: The use of CrCl provides a different classification than the estimation of GFR using a prediction equation. The CrCl is unreliable when it is necessary to identify CKD subjects with decrease of GFR of 5ml/min/1.73m(2)/year.