Individual-level estimated glomerular filtration rate (eGFR) values differ substantially from measured GFR (mGFR) values, reports a study in the Annals of Internal Medicine.
The researchers analyzed data on 3223 participants in four US epidemiologic studies that included mGFR values. The mean age of participants was 59 years; 55% of participants were women, and 32% were Black.
The mGFR values were obtained using non-radiolabeled iothalamate in two studies, radiolabeled iothalamate in one study, and plasma clearance of iohexol in one study and were indexed to 1.73 m2 of body surface area. The eGFR values were calculated from serum creatinine alone and with serum cystatin C. The magnitude and clinical significance of any differences in the paired mGFR and eGFR values were assessed.
The mean mGFR value was 68 mL/min/1.73 m2. Nine percent of participants had an mGFR value of less than 30 mL/min/1.73 m2. Overall eGFR values were higher than mGFR values; the median difference was –0.6 mL/min/1.73 m2, with significant differences between groups. However, individual-level differences between values were large across subgroups defined by race, age, and sex.
At a creatinine-based eGFR of 60 mL/min/1.73 m2, 50% of eGFR values were between 62 and 67 mL/min/1.73 m2, 80% between 45 and 76 mL/min/1.73 m2, and 95% between 36 and 87 mL/min/1.73 m2. At an eGFR of 45 mL/min/1.73 m2, 15% of participants had mGFR values outside the range of 30–60 mL/min/1.73 m2, 30% outside the range of 35–45 mL/min/1.73 m2, and 57% outside the range of 40–50 mL/min/1.73 m2.
The discrepancies led to errors in classification of chronic kidney disease by mGFR versus eGFR. The misclassification rate was 42%, with 22% of participants placed in a lower eGFR category and 20% in a higher category. Analysis of cystatin C-based eGFR showed no meaningful improvement.
The results show “substantial individual discrepancy” between eGFR and mGFR values. “Our findings highlight the need to make direct GFR measurements available to patients who need them,” the researchers conclude, noting that non-radiolabeled techniques have made GFR measurement simpler and more feasible for clinical use [Shafi T, et al. Quantifying individual-level inaccuracy in glomerular filtration rate estimation: A cross-sectional study. Ann Intern Med, published online ahead of print July 5, 2022. doi: 10.7326/M22-0610; https://www.acpjournals.org/doi/10.7326/M22-0610].