Estimation of glomerular filtration rate in liver-transplanted children: comparison of simplified procedures using 51Cr-EDTA and endogenous markers with Sapirstein's method as a reference standard.

This study evaluated simple procedures for GFR determination in 48 liver-transplanted children. After injection of (51)Cr-EDTA, blood samples were obtained up to four h, and activity retention in the body was measured for 60 min with scintillation probes. As a reference, GFR was calculated according to Sapirstein. Simplified calculations were performed according to Brochner-Mortensen, Russel, Devaux and Oberhausen. Additionally, GFR was determined using plasma creatinine and cystatin C according to Schwartz and Filler, respectively. The reference revealed mildly reduced GFR (62 +/- 20 mL/min/1.73 m(2)). Russel's method provided the highest degree of correlation (r(2) = 0.95), the smallest bias in GFR determination (-2%), and only one false exclusion plus one false diagnosis of chronic kidney disease. Oberhausen's method with blood sampling at one h post-injection performed slightly worse (r(2) = 0.67, bias: 3%). All other methods resulted in significantly different GFR estimates compared to the reference. Nevertheless, notably, the second narrowest 95% limits of agreement (-31% to 45%) was observed using cystatin C. In conclusion, this data implies to prefer Russel's method as a simplified procedure, and if patients cannot be available long enough (four h) for measurements, Oberhausen's method instead. If radiotracer methods are not available at all or for screening GFR, cystatin C appears to be the procedure of choice.