Using glomerular filtration rate estimating equations: clinical and laboratory considerations.

Glomerular filtration rate (GFR) estimating equations, infrequently used just a decade ago, are now recommended for the evaluation of kidney function for routine clinical care and are routinely reported by the vast majority of clinical laboratories (1 ). Current clinical guidelines recommend estimated GFR (eGFR) based on serum creatinine (eGFRCr) as the initial diagnostic test, and either a measured clearance or estimated GFR based on serum cystatin C or the combination of serum cystatin C and creatinine (eGFRCys and eGFRCr-Cys, respectively) as a confirmatory test (2, 3 ). These recommendations apply to all adults, irrespective of geographic region or clinical presentation. According to the guidelines, measurement of serum concentrations of creatinine and cystatin C should use assays traceable to international reference measurement procedures and materials, and estimation of GFR should use equations developed by the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI), specifically the 2009 creatinine equation and the 2012 cystatin C and creatinine–cystatin C equations, unless other equations have been found to be more accurate in the population of interest (Table 1) (4–6 ). The rationale for these recommendations is as follows. First, serum creatinine concentration is routinely measured in clinical practice, and most commercial creatinine measurement procedures are now well standardized; thus eGFRCr is available for use as an initial diagnostic test in most clinical encounters in adults. On the other hand, the international reference standard for cystatin C has been developed only recently, and not all commercial assays are sufficiently well standardized for routine clinical use (7, 8 ). Second, the CKD-EPI equations can be computed from variables that are generally available in clinical and laboratory information systems (age and sex in addition to serum concentrations of creatinine and cystatin C, with separate values reported for African-American individuals) (9 ). Third, the CKD-EPI equations were developed using standardized assays, thus avoiding analytical method–related biases in eGFR results. Fourth, the CKD-EPI equations were developed in diverse populations with a wide range of GFR values and clinical characteristics (including subjects with and without CKD and diabetes); thus, they are broadly applicable. However, application to selected populations in which the relationship of the filtration marker to measured GFR (mGFR) differs from the relationship observed in the development population will be associated with systematic bias in eGFR. Fifth, even when unbiased, eGFRCr and eGFRCys are limited by imprecision (uncertainty) compared with mGFR, owing to variation in non-GFR determinants of creatinine and cystatin C that are not accounted for by other variables in the equations. However, GFR estimates based on both filtration markers are likely to be more precise than estimates based on either marker alone, by minimizing errors due to nonGFR determinants of the serum concentration of each filtration marker. In particular, serum creatinine and cystatin C concentrations appear to be influenced by different clinical factors (Table 1). Nonetheless, there are numerous clinical conditions in which GFR estimates are less than accurate, and guidelines recommend understanding factors that lead to inaccuracy (Table 1). Rapidly changing GFR (as in development of and recovery from acute kidney injury) leads to non–steady-state conditions for serum concentrations of creatinine and cystatin C, and GFR estimates are more accurate in the steady state. Systematic differences in non-GFR determinants of the filtration markers or analytical biases for the creatinine or cystatin C measurement procedures used in the study population vs those used for equation development lead to systematic bias in eGFR. Higher GFR magnifies errors under these conditions because of the inverse relationship of GFR with filtration marker concentration. In addition, imprecision or systematic differences in GFR measurement procedures between the study populations used during the development and validation of the GFR estimating equation themselves will lead to the appearance of imprecision or bias in eGFR. In this issue of Clinical Chemistry, Meeusen et al. investigated whether some clinical presentations are associated with differences in performance of the CKD-EPI 1 Division of Nephrology, Tufts Medical Center, Boston, MA; 2 Department of Laboratory Medicine and Pathology, University of Minnesota, Minneapolis, MN. * Address correspondence to this author at: Department of Laboratory Medicine and Pathology, University of Minnesota, MMC609, Room 763-1 Mayo Bldg, 420 Delaware St SE, Minneapolis, MN 55455. Fax 612-626-2696; e-mail [email protected]. Received July 27, 2015; accepted August 3, 2015. Previously published online at DOI: 10.1373/clinchem.2015.245282 © 2015 American Association for Clinical Chemistry 3 Nonstandard abbreviations: GFR, glomerular filtration rate; eGFR, estimated GFR; CKDEPI, Chronic KidneyDisease Epidemiology Collaboration;mGFR,measuredGFR; KDIGO, Kidney Disease: Improving Global Outcomes. Clinical Chemistry 61:1