Comparison of eGFR equations for estimating renal function in Chinese children under 2 years: a retrospective study

Accurate assessment of glomerular filtration rate (GFR) is essential for managing kidney diseases. However, the performance of estimated GFR (eGFR) equations remains poorly characterized in children under 2 years. This study aimed to evaluate the agreement and correlation of eGFR equations against measured GFR (mGFR) in this understudied population. This retrospective analysis enrolled 76 children. A total of 17 eGFR equations were compared with mGFR using Spearman’s correlation analysis, Bland-Altman analysis, Lin’s concordance correlation coefficient analysis, and latent profile analysis (LPA). The Tang (combined-creatinine (Cr)/cystatin C (CysC)/blood urea nitrogen (BUN)) equation demonstrated no significant difference from mGFR ( P = 0.533), with strong correlation ( r = 0.655) and excellent agreement (mean difference=-1.1, concordance correlation coefficient = 0.66). This equation exhibited the most favorable bias (bias=-2.7), precision (interquartile range of difference = 24.3), and greatest accuracy (P₃₀=75.0%) among all evaluated equations. LPA further clustered this equation with mGFR in the same profile. Subgroup analyses confirmed its robust performance across sex, chronic kidney disease status, age strata, and mGFR levels. The Tang (combined-Cr/CysC/BUN) equation demonstrated good performance in children under 2 years, representing a potentially reliable alternative for GFR estimation when gold-standard mGFR measurement is unavailable.