Measurement error in estimated GFR slopes across transplant chronic kidney disease stages.

BACKGROUND This study examines if transplant glomerular filtration rate (GFR) slope prediction is affected by the degree of transplant chronic kidney disease (CKDT) stage. METHODS Serial changes in estimated GFR (DeltaeGFR) by Cockcroft-Gault (CG) and Modified Diet in Renal Disease-Isotope Dilution Mass Spectrometry (MDRD-IDMS) equations were compared to simultaneous changes in isotope GFR (DeltaiGFR) in renal transplant patients who had at least four scans. RESULTS Total number of patients (iGFR scans) was 99 (772) while the corresponding numbers in CKDT stages 1-4 were 33 (103), 69 (239), 75 (316) and 37 (96), respectively. Measurement error [(DeltaeGFR - DeltaiGFR) x 100/DeltaiGFR] (median +/- IQR, interquartile range) estimated from CG and MDRD-IDMS slopes were -414.29 +/- 276.16% and -342.86 +/- 210.18% (stage 1); -350.00 +/- 301.22% and -300.00 +/- 525.00% (stage 2); -26.02 +/- 404.38% and -26.58 +/- 423.13% (stage 3); 10.26 +/- 142.18% and -76.92 +/- 145.64% (stage 4), respectively. The proportion of patients with CG measurement error < or =1-fold in stages 1 and 2 of 12 and 14.5% was significantly (p < 0.05) lower than that of 36.3 and 52.8% at stages 3 and 4, respectively. Similar measurement errors were observed for MDRD-IDMS. CONCLUSIONS Transplant GFR slope prediction is affected by the degree of renal dysfunction. Errors in slope prediction are much higher in those with better function and thus add another limitation for eGFR use in longitudinal studies on progressive graft dysfunction.