Performance of point-of-care creatinine testing in the estimation of glomerular filtration rate: comparison with laboratory creatinine and cystatin C-based measurements

BACKGROUND: Chronic kidney disease (CKD) is a major global public health issue, especially in low- and middle-income countries where access to laboratory testing is limited. Point-of-care tests for creatinine may offer an alternative for detection of CKD with kidney function impairment in such settings. However, its analytical and clinical validity compared to standard laboratory methods remains uncertain. AIM: To assess the performance of a point-of-care test device for estimating glomerular filtration rate (GFR) compared to laboratory-based creatinine and cystatin C-based estimated GFR (eGFR). METHODS: This cross-sectional analytical study was conducted at Ouakam Military Hospital and included 405 adults undergoing routine kidney function testing. Each participant underwent point-of-care testing using the Nova StatSensor® device and standard laboratory tests for serum creatinine and cystatin C. The eGFR was calculated using the 2021 Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equations. Performance was evaluated through correlation analyses, Bland-Altman plots, Deming regression, percentage of values within ±30% of the reference value (P30) accuracy, and diagnostic performance using receiver operating characteristic (ROC) curves. RESULTS: Point-of-care tests for creatinine showed moderate correlation with laboratory serum creatinine (r = 0.48) and respective eGFR values (r = 0.60), but consistently overestimated creatinine, leading to underestimation of eGFR (mean bias: -31.09 mL/min/1.73 m2). Correlation with cystatin C-based eGFR was weaker (r = 0.35), with a significant negative bias (-38.57 mL/min/1.73 m2). P30 accuracy was low (32.6% vs lab creatinine; 28.1% vs cystatin C). The diagnostic performance for detecting CKD stage 3 and over (eGFR < 60 mL/min/1.73 m2) demonstrated a high negative predictive value (91.3% vs 90% using cystatin C and serum creatinine, respectively, as reference). CONCLUSIONS: While point-of-care testing shows limitations for precise GFR estimation, it demonstrates utility for community-based CKD screening with high negative predictive value.