Evaluation of eGFR methods in a sub-Saharan African community-based pediatric population

Accurate assessment of the estimated glomerular filtration rate (eGFR) plays a pivotal role in the early detection, management, and optimal medication dosing for chronic kidney disease (CKD). However, validation of eGFR, utilizing cystatin C-based equations, is limited in African children and adolescents with CKD. We evaluate the agreement of eGFR equations incorporating both cystatin C and creatinine in this specific population.

This community-based study assessed CKD in children (2-15 years) using cystatin C and serum creatinine. eGFR agreement with the reference was evaluated with Bland-Altman plots, ROC curves, and Lin's CCC, using the Under-25 serum creatinine-cystatin C equation as the reference standard. Pairwise ROC comparisons assess the statistical differences in estimation equation agreement.

Among 666 children (mean age, 7.8 ± 3.8 years; 48.6% male), CKD prevalence was 11.6% (95% CI, 9.2-14.2%). Notably, the Chehade equation, using combined biomarkers, aligned best with the reference, displaying the lowest mean deviation (- 0.59; 95% CI, - 1.19 to 0.01), superior agreement (P10, 91.0%; P30, 96.70%), and highest discriminatory power (0.989). In contrast, CKD-EPI 2012 cystatin C had the highest mean deviation (- 35.90) and lowest discriminatory power (0.79). Equations combining creatinine and cystatin C (Schwartz, Chehade, Full Age Spectrum) demonstrated strong positive Lin's CCC with CKiD U25 creatinine-cystatin C, while Bouvet showed a notably weak correlation (Lin's CCC, 0.22).

In African children with CKD, the Chehade, CKiD Under 25 creatinine-based equations, and the Full Age Spectrum equations show promise for CKD diagnosis. However, a measured GFR is essential to identifying the most accurate eGFR equation in this population.