Glomerular filtration rate estimation using creatinine and cystatin C: a prospective study of test accuracy in adults with moderate chronic kidney disease

Objectives<br/>Equations used to estimate glomerular filtration rate (GFR) include those based on creatinine and cystatin C, an alternative kidney function marker. We studied the performance of two contemporary sets of GFR estimating equations, Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) and European Kidney Function Consortium (EKFC), including both single biomarker (creatinine or cystatin C only) and combined (creatinine and cystatin C) versions of these equations to assess their accuracy in a population with moderate chronic kidney disease (CKD). The use of creatinine-based GFR estimates such as CKD-EPIcreatinine represents standard clinical practice in many countries. The additional, or alternative, use of cystatin C in such equations may improve accuracy of GFR estimation.<br/><br/>Design<br/>A prospective multi-centre study evaluating accuracy of published GFR equations in people with CKD. <br/><br/>Setting<br/>Primary, secondary and tertiary care across six centres in England.<br/><br/>Participants<br/>Adults (&gt;18 years, n=1167) with moderate CKD (estimated GFR 30-59 mL/min/1.73 m² sustained over at least 3 months prior to recruitment).<br/><br/>Interventions<br/>Estimated GFR using the CKD-EPI_creatinine, CKD-EPI_cystatin, CKD-EPI_{creatinine-cystatin}, EKFC_creatinine, EKFC_cystatin and EKFC_{creatinine-cystatin} equations. Recently (2021) remodelled versions of the CKD-EPI equations (CKD-EPI(2021)_creatinine, CKD-EPI(2021)_{creatinine-cystatin}) were also studied.<br/><br/>Main outcome measures<br/>Accuracy of estimating equations compared against measured GFR (iohexol clearance). Accuracy was expressed as P30 (percentage of estimates within 30% of measured GFR).<br/><br/>Results<br/>Median age was 67.5 years; 58.3% were male; 86.9% white; 27.8% had diabetes; median measured GFR was 47.0 mL/min/1.73 m2; 57.0% had albuminuria. Test calibration critically affected cystatin C measurement. After cystatin C recalibration P30 values were 90.2% (CKD-EPI_creatinine), 89.5% (CKD-EPI_cystatin), 94.9% (CKD-EPI_{creatinine-cystatin}), 88.0% (CKD-EPI(2021)_creatinine), 94.9% (CKD-EPI(2021)_{creatinine-cystatin}), 89.4% (EKFC_creatinine), 91.0% (EKFC_cystatin) and 94.9% (EKFC_{creatinine-cystatin}). Creatinine-based equations showed varying bias depending on level of GFR: inclusion of cystatin C in equations ameliorated this effect. Differences in accuracy across age, sex and GFR level varied by equation, but equations combining creatinine and cystatin performed equally across age, sex, diabetes/no diabetes, albuminuria, and body mass index categories.<br/><br/>Conclusions<br/>The CKD-EPI_creatinine equation has acceptable accuracy in a white population in England with moderate CKD. Combined dual biomarker equations showed higher accuracy than the CKD-EPI_creatinine equation and their equivalent creatinine-only equations. Further research is warranted to determine the most accurate equation to use in persons with black or South-Asian ethnicity in England.