Accuracy of modification of diet for renal disease and Cockcroft-Gault equations as

Introduction: We intended to assess the accuracy of re-expressed Modification of Diet for Renal Disease (MDRD) and Cockcroft-Gault (CG) equations to estimate glomerular filtration rate (GFR) in chronic kidney disease in two different etiologies of acute renal failure (ARF): acute tubular necrosis (ATN) and acute glomerulonephritis (AGN). Methods: Patients admitted for ARF or the patients complicated with ARF during the course of their hospitalization were enrolled to the study (n=21; 14 females and 7 males; 11 ATN and 12 AGN). When the plasma creatinine reached a steady state ( < 15% change in two consecutive days), GFR was measured with double plasma sample method (GFRDPSM) using 99mTc-DTPA. GFR was also estimated by MDRD (GFRMDRD) and CG (GFRCG) equations. Results: The patients aged 44.8±19.5 years and weighted 67.8±10.7kg. GFRDPSM (32.9±14.7 ml/min) was statistically different from the GFRMDRD (11.6±8.2 ml/min; p < 0.001) and CG-GFR (16.5±10ml/min; p < 0.001). The difference between DPSM-GFR and MDRD-GFR was statistically significant in patients with either ATN (n=11; 31.9±15.0 vs. 11.7±10.3ml/min and p=0.001) or AGN (n=10; 34.1±15.1 vs. 11.4±5.6 ml/min and p=0.001); similarly the GFRCG was lower than GFRDPSM in patients with either ATN (16.5±12.5ml/min and p < 0.01) or AGN (16.3±7.1ml/min and p < 0.005). No statistically significant correlation was found between the GFRDPSM and GFRMDRD (r=0.34; p=0.13) but GFRDPSM and GFRCG values were correlated (r=0.48; p=0.03). Out of subjects with GFRDPSM > 30, 92.3% had GFRMDRD < 30ml/min and 84.6% had GFRCG < 30ml/min. Conclusion: Our results indicate that MDRD and CG equations were substantially inaccurate in patients with ARF. More precise methods of GFR evaluation is recommended in these patients.