Accurate estimates of kidney function are essential for appropriate treatment decisions. The best measure of kidney function is the glomerular filtration rate (GFR). To measure GFR is complicated in clinical practice, requiring substantial time and resources. Alternatively, GFR can be estimated from a blood sample by using equations (eGFR) based on the plasma concentration of creatinine or cystatin C. Creatinine is formed in muscles as a break-down product of creatine, while cystatin C is a small protein produced by nearly all cells in the body.
Is it uncertain which biomarkers and equations yield the most accurate estimates of GFR, and opinions and practices vary. The cost-effectiveness of various methods has also been questioned. This report aims to review how accurately the different equations based on creatinine, cystatin C, or a combination of both markers, can estimate GFR in different patient groups. We have not evaluated either the direct benefits to patients’ health or the levels of accuracy required in different clinical situations.
- Equations based on the plasma concentration of creatinine or cystatin C generally estimate kidney function (GFR) with sufficient and equal accuracy. This option is currently underutilized in clinical practice.* Factors such as muscle mass and meat intake (for creatinine) and corticosteroid medication (for cystatin C) should be considered when evaluating estimated GFR. Equations based on cystatin C alone are generally accurate, while creatinine-based equations, to be equally accurate, must include certain demographic information (eg age and sex) and, at times, anthropometric measurements (weight and height).
* In usual practice, an eGFR equation is defined as having sufficient accuracy when at least 75% of the estimates fall within ±30% of the measured GFR.
- The mean value of estimated GFR based on both creatinine and cystatin C is more accurate than equations based on either. This is not widely known. This applies especially to adult patients with low GFR (<30 mL/min/1.73 m2) and children.
- Creatinine-based equations are not sufficiently accurate in children with low BMI (<20 kg/m2). In the elderly (>80 years) just a few, eg the revised Lund-Malmo equation (LM-rev), are sufficiently accurate. In these patient groups the accuracy of cystatin C-based equations and equations based on the mean value of creatinine and cystatin C has not been adequately studied.
- When impaired kidney function is suspected, using both creatinine and cystatin C in estimating GFR is probably more cost-effective than using only one of the methods.
- Laboratories should report estimated GFR, thereby giving the healthcare provider a measure of kidney function instead of reporting just the creatinine value, as done previously. GFR can be estimated with sufficient accuracy from both creatinine and cystatin C, at least up to 90 mL/min/1.73 m2.
- Swedish laboratories currently use several analytical methods and equations to estimate GFR. Greater uniformity is desirable. Analyses of cystatin C should be traceable to the international cystatin C calibrator. Equations based on IDMS-traceable creatinine analyses** (MDRD, CKD-EPI, and LM-rev) should be used. LM-rev is developed in Sweden and is at least as accurate as the above-mentioned equations. The Cockcroft-Gault creatinine-based equation is substantially less accurate and should not be used.
** IDMS = Isotope dilution mass spectrometry, an internationally certified standard for plasma creatinine analysis.
- Endogenous creatinine clearance is still used to measure GFR. This method overestimates GFR and should be discontinued.
How to cite this report: SBU. Methods to estimate and measure renal function (glomerular filtration rate). Stockholm: Swedish Council on Health Technology Assessment (SBU); 2012. SBU report no 214 (in Swedish).