Improved equations to estimate GFR in Chinese children with chronic kidney disease

Background There is currently no specific equation for estimating glomerular filtration rate (GFR) in Chinese children with chronic kidney disease (CKD). The commonly used equations are less robust than expected; we therefore sought to derive more appropriate equations for GFR estimation. Methods A total of 751 Chinese children with CKD were divided into 2 groups, training group (n = 501) and validation group (n = 250). In the training group, a univariate linear regression model was used to calculate predictability of variables associated with GFR. Residuals were compared to determine multivariate predictability of GFR in the equation. Standard regression techniques for Gaussian data were used to determine coefficients of GFR-estimating equations after logarithmic transformation of measured GFR (iGFR), height/serum creatinine (height/Scr), cystatin C, blood urea nitrogen (BUN), and height. These were compared with other well-known equations using the validation group. Results Median Tc-99m-DTPA GFR was 90.1 (interquartile range: 67.3-108.6) mL/min/1.73 m(2) in training dataset. Our CKD equation, eGFR (mL/min/1.73 m(2)) = 91.021 [height(m)/Scr(mg/dL)/2.7](0.443) [1.2/Cystatin C(mg/L)](0.335) [13.7/BUN (mg/dL)](-0.095) [ 0.991(male)] [height(m)/1.4](0.275), was derived. This was further tested in the validation group, with percentages of eGFR values within 30% and 15% of iGFR (P30 and P15) of 76.00% and 48.40%, respectively. For centres with no access to cystatin C, a creatinine-based equation, eGFR (mL/min/1.73 m(2)) = 89.674 [height(m)/Scr(mg/dL)/2.7](0.579) [ 1.007(male)] [height(m)/1.4](0.187), was derived, with P30 and P15 73.60% and 49.20%, respectively. These were significantly higher compared to other well-known equations (p < 0.05). Conclusion We developed equations for GFR estimation in Chinese children with CKD based on Scr, BUN and cystatin C. These are more accurate than commonly used equations in this population.