Examining commonly used equations for estimating the glomerular filtration rate (GFR) in a healthy cohort of children...

In this study, we compared established equations for estimating the glomerular filtration rate (GFR) in a cohort of apparently healthy children and adolescents and aimed to evaluate the equations' validity.

Blood samples and anthropometric data from 4,776 apparently healthy participants (0.25-21 years) were analyzed. The glomerular filtration rate was estimated (eGFR) using the revised Schwartz Bedside (2009), the Cystatin C- and Serum Creatinine- based Schwartz equation, the Chronic Kidney Disease in Children (CKiD) equation, the 3 versions of the U25 (U25cys, U25scr, U25ave), the Serum Creatinine version, as well as the Cystatin-C based equation from the European Kidney Function Consortium (EKFCcr, EKFCcys), the CAPA equation, and the Cystatin C-Creatinine-based Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI). The resulting eGFR distributions were compared, and the percentages of eGFR values within the expected physiological eGFR range (90-120 ml/min/1.73 m²) were calculated stratified by age and sex. Subsequently, age ranges with implausible distributions were identified.

EKFCcr, CKiD, Schwartz, and U25scr yielded the highest proportion of eGFRs within the expected range and showed consistent values across age groups without large jumps. Bland-Altman analysis indicated extremely low biases between Schwartz and EKFCcr and very high biases especially between CKD-EPI, CAPA and Bedside in reference to the other equations. Males showed higher eGFRs than females.

We found that the EKFC equations, Schwartz, and U25 resulted in eGFR distributions within the expected physiological range for an apparently healthy pediatric cohort. Estimates from simpler equations such as Bedside and CAPA were more prone to implausible results.

Clinical trial number NCT02550236 (clinicaltrials.gov, date of registration: 2014-12-15).