Should the Schwartz formula for estimation of GFR be replaced by cystatin C formula?

Abstract

It is common practice to estimate glomerular filtration rate (GFR) from the Schwartz formula (a height creatinine/ratio), although it has its limitations. Cystatin C was found to be a superior marker of GFR. No formula has been validated to estimate GFR from cystatin C in children. Children (aged 1.0-18 years, n=536) with various renal pathologies undergoing nuclear medicine GFR clearance studies (99mTc-DTPA single-injection technique) were tested. Cystatin C was measured with a nephelometric assay. The Schwartz GFR was calculated using enzymatically determined serum creatinine in micromoles per liter using the constant 48 for adolescent males and 38 otherwise. Using multiple stepwise regression analysis on log/log-transformed data, we derived the following relationship between the cystatin C concentration and GFR: log(GFR)=1.962+[1.123*log(1/Cystatin C)]. Using the Bland and Altman analysis to test agreement between the Schwartz formula and gold standard GFR showed considerable bias, with a mean difference of +10.8% and a trend towards overestimation of the GFR by the Schwartz formula with lower GFRs. In contrast, the Bland and Altman analysis applied on the GFR estimate derived from cystatin C showed the mean difference to be negligible at +0.3% and no trend towards overestimation of the GFR with lower GFRs. In the regression analysis of the estimate and the GFR, the Schwartz estimate showed significant deviation from linearity, whereas the cystatin C estimate did not. In conclusion, the data suggest that this novel cystatin C-based GFR estimate shows significantly less bias and serves as a better estimate for GFR in children.