Introduction:

the Colombian guidelines for chronic kidney disease (CKD) recommend estimating glomerular filtration (GF) using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation. No studies have been performed in the Colombian population to compare the accuracy of this equation to that of others used in clinical practice.

Design and methods:

we evaluated the GF estimation performance of the Modification of Diet in Renal Disease (MDRD-4), Cockroft-Gault (CG) and body surface area adjusted Cockroft-Gault (CG-BSA) equations against the CKD-EPI equation in 757 adult patients. Performance was evaluated using bias, precision and accuracy measurements.

Results:

the mean GF by CKD-EPI was 37.32±12.71 mL/min/1.73m2; by MDRD-4 it was 39.8±13.2 mL/min/1.73m2, by CG it was 35±12.6 mL/min and by CG-BSA it was 34.52±11.34 mL/ min/1.73m2. All the equations had bias with respect to GF by CKD-EPI. The most accurate equation was GF estimated by MDRD-4 (MeGF) with 97.1 and 99.74% of measurements within 15 and 30%, respectively; and the least accurate was GF estimated by CG (CGeGF) with 59.7 and 81.77% of the measurements within 15 and 30%, respectively. The concordance correlation coefficient between GF by CKD-EPI and MDRD-4 was 0.97, with CG and CG-BSA at 0.78 and 0.85, respectively

Conclusions:

the most accurate alternative equation for estimating glomerular filtration in this Colombian population is MDRD-4, which has a high concordance with the CKD-EPI equation. Estimation of GF with the CG equation is not recommended.

        · abstract in Spanish     · text in English | Spanish     · English ( pdf ) | Spanish ( pdf )