Accuracy of glomerular filtration rate estimation formulas for detecting moderate chronic kidney disease in elderly brazilians

OBJECTIVE: The objective of this study was to investigate the accuracy of a glomerular filtration rate (GFR) estimation equation based upon serum cystatin C (Hojs) and the abbreviated Modification of Diet in Renal Disease (MDRD) study and the Cockcroft-Gault (CG) corrected for body surface area (CG-BSA) estimation equations for detecting moderate chronic kidney disease against 51Cr-EDTA clearance (taken as the gold standard) in southern elderly Brazilians. METHOD: An Observational prospective study in which GFR estimates via CG without BSA correction were performed in 160 community patients 60 aged or more. The ones with estimated GFR < 60/ml/min underwent GFR determination using 51Cr-EDTA clearance. The GFR obtained with this method was compared with GFR estimates obtained via CG-BSA, MDRD and Hojs. RESULTS: 69,1% of the patients were female. Forty of the 160 studied elderly studied had an estimated GFR < 60/ml/min and underwent GFR determination using 51Cr-EDTA. Of the forty patients with an estimated GFR < 60 ml/min, only 18 presented GFR values < 60 ml/min/1.73 m2 using 51Cr-EDTA clearance. The accuracy within 30% was 78% for MDRD, 75% for CG-BSA and 45% for Hojs, with a trend towards superiority of the first two equations over the last one. CONCLUSIONS: Either CG-BSA or MDRD can be used for GFR estimation in southern elderly Brazilians until larger studies about the subject are available. The common practice of using CG without BSA correction for chronic kidney disease screening led to overdiagnosis of the condition in our study.