-
Views
-
Cite
Cite
Ahmed Zahran, Ahmed Shoker, Reply, Nephrology Dialysis Transplantation, Volume 23, Issue 5, May 2008, Pages 1763–1764, https://doi.org/10.1093/ndt/gfn027
Close - Share Icon Share
Extract
Sir,
In the above-mentioned letter, concerns regarding the calibration of creatinine and the ELISA method used to measure cystatin C were raised as a potential drawback of the manuscript.
We thank the authors for their constructive criticism. Several points should be raised, however, in support of our conclusion. First, we tested the modified MDRD2-IDMS method, which is suggested as an additional equation for centres that use the Synchron LX20 system to measure their serum creatinine. Secondly, many of the above-tested equations were developed with methods closer to ours than the most recently modified assays. Thirdly, as we have identified in the manuscript, Hallan et al . have demonstrated elegantly that the bias between the creatinine-based methods is lower in patients with higher serum creatinine.
Stevens et al . have studied the impact of creatinine calibration on the performance of GFR estimating equations in a pooled individual patient database. They concluded that the effect of calibration was greater at higher levels of GFR. For the C–G equation, calibration worsened the median percentage of difference from −2% to −11.4%. Calibration improved median percentage of difference between measured and estimated GFR by the MDRD2 equation from 9% to 5.8%. A striking finding, however, was their conclusion that calibration could not account for variation in assay performance among individuals. After calibration, larger errors remained for GFR estimates >16 ml/min/1.73 m 2 [ 1 ].
Comments