In individual patients, an estimated glomerular filtration rate (eGFR) based on serum creatinine and/or cystatin C often differs substantially from measured GFR (mGFR), investigators reported in the Annals of Internal Medicine.

The individual-level difference between mGFR and eGFR is large, according to Tariq Shafi, MBBS, MHS, of The University of Mississippi Medical Center in Jackson, Mississippi, and colleagues. “Clinicians need to recognize that the eGFR is not an mGFR replacement and consider eGFR’s inaccuracy while managing individual patients.”

In analyses of data from 3223 participants across 4 studies, the investigators found that only 37% of eGFR results based on creatinine (eGFRCR) fell within 10% of mGFR. For example, at an eGFRCR of 45 mL/min/1.73 m2, 15% of the participants had an mGFR outside of the range of 30-60 mL/min/1.73 m2, 30% had an mGFR outside the range of 35-45 mL/min/1.73 m2, and 57% had an mGFR outside the range of 40-50 mL/min/1.73 m2. Such discrepancies between measured and estimated GFR were observed regardless of an individual’s race, age, or sex.

Continue Reading

The agreement in chronic kidney disease (CKD) staging between mGFR and eGFRCR was only 58% overall, Dr Shafi’s team reported. Among the 42% of patients who were misclassified, 22% were reclassified to a lower CKD stage and 20% to a higher stage. For 39%, the misclassification was by 1 CKD stage.

The eGFR equations incorporating cystatin C did not improve the probability of large errors and CKD stage misclassification. The investigators pointed out that non-GFR factors influence the serum concentration of creatinine and/or cystatin C, such as obesity, muscle mass, meat consumption, and fasting.

Findings Poised to Change Clinical Practice

The clinical implications of these findings are far-reaching. Underestimating GFR may exclude patients from receiving optimal therapies and overestimating GFR may leave patients vulnerable to drug adverse effects such as hyperkalemia from mineralocorticoid antagonists and toxicity from chemotherapy, Dr Shafi’s team pointed out.

In an interview with Renal & Urology News, Christine A. White, MD, MSc, of Queen’s University in Kingston, Ontario, Canada, who was not involved in the study, said the research by Dr Shafi and colleagues “exposes the substantial discrepancy that often exists between estimated GFR using creatinine or cystatin C and measured GFR in individuals. This discrepancy is not captured by the traditionally reported metric of overall population bias where individual positive and negative biases negate each other, leading to an erroneous impression of unbiased GFR estimation at the individual level.

“The discrepancy,” she continued, “has not been well appreciated by clinicians assessing patients’ kidney function who then make work-up and treatment decisions according to eGFR, nor is it explicitly addressed in many society clinical practice guidelines.”

Dr White said the study authors appropriately recommend that the uncertainty of eGFR be reported by laboratories alongside eGFR results and that GFR measurement should become more widespread. Dr White will be conducting a session on GFR measurement techniques at the American Society of Nephrology’s Kidney Week in November 2022.

“There are many ways to measure GFR using either renal or plasma clearance of a variety of exogenous markers with different sampling strategies,” she said. Few comparative studies have evaluated various protocols against the gold standard renal inulin clearance, which is no longer commercially available, she noted.

“Available studies indicate that GFR measurement protocols should be tailored according to specific patient characteristics such as level of eGFR and presence of edema,” Dr White said.

Iohexol has several advantages over other tracers, she explained. It is non-radioactive, inexpensive, and already in use worldwide as a contrast media, so it is “poised to become a preferred tracer.”

According to Dr White, “Studies such as this one may galvanize the development of GFR measurement protocols that are both accurate and logistically and economically feasible across the spectrum of GFR and clinical presentations.”

The study authors noted that advances in nonradiolabeled GFR measurement techniques have made it a “highly feasible” and safe outpatient procedure. Race, sex, age, and socioeconomic factors should not cause errors, although results may vary due to normal physiology and measurement errors, such as incomplete bladder emptying. It should be a “priority” and made “widely available,” they wrote.

The 4 cohorts in this study included GENOA (Genetic Epidemiology Network of Arteriopathy), ALTOLD (Assessing Long Term Outcomes in Living Kidney Donors), ECAC (Epidemiology of Coronary Artery Calcification), and CRIC (Chronic Renal Insufficiency Cohort). Overall, 58% of participants were White, 32% Black, 7% Hispanic, and 3% another race. The GFR was directly measured using urinary clearance of nonradiolabeled iothalamate in GENOA and ECAC, radiolabeled iothalamate in CRIC, and plasma clearance of iohexol in ALTOLD.

The investigators calculated eGFRCR using the Chronic Kidney Disease Epidemiology (CKD-EPI) 2021 race-free equation, which is valid only in adults, and the European Kidney Function Consortium (EKFC) equation, which is valid in individuals aged 2 years or older. They calculated the eGFR from serum cystatin C (eGFRCYS) and from serum creatinine and serum cystatin C combined (eGFRCR-CYS) using the CKD-EPI 2012 and 2021 equations, respectively.


Shafi T, Zhu X, Lirette ST, et al. Quantifying individual-level inaccuracy in glomerular filtration rate estimation: a cross-sectional study. Ann Intern Med. Published online July 4, 2022. doi:10.7326/M22-0610