The Clinician and Estimation of Glomerular ... - Semantic Scholar

Special Feature

The Clinician and Estimation of Glomerular Filtration Rate by Creatinine-based Formulas: Current Limitations and Quo Vadis Rossini Botev,* Jean-Pierre Mallie´,† Jack F.M. Wetzels,‡ Ce´cile Couchoud,§ and Otto Schu¨ck

Summary The GFR has a paramount diagnostic and staging role in the Kidney Disease Outcome Quality Initiative Clinical Practice Guidelines for Chronic Kidney Disease (K/DOQI-CKD). The most widely used serum creatininebased formulas in adults for estimated GFR (eGFR) are the Cockcroft-Gault (CG) and Modification of Diet in Renal Disease Study (MDRD). Recently, a new Chronic Kidney Disease Epidemiology Collaboration equation has been developed. Review of the literature revealed that CG and MDRD formulas correctly assigned overall only 64% and 62%, respectively, of the subjects to their actual K/DOQI-CKD classification’s GFR groups as determined by measured GFR (mGFR). This suggests that approximately 10 million (38%) subjects may have been misclassified on the basis of estimated CKD prevalence of 26.3 million adults in the United States. The purpose of this review is to help the clinician understand the limitations of using eGFR in daily practice. We also elaborate upon issues such as the differences among markers of mGFR, the validity of adjusting GFR for body surface area in certain populations, the limited data on boundaries for normal mGFR according to age, gender, and race, the need for calibration of a wide spectrum of serum creatinine measurements, the lack of actual eGFR value above 60 ml/min per 1.73 m2 and reference for normal mGFR in the clinical laboratories’ reports, and the performance evaluation of the eGFR formulas. Several pitfalls have to be overcome before we can reliably determine health and disease in daily nephrology practice to preserve the first rule of practicing medicine: primum non nocere. Clin J Am Soc Nephrol 6: 937–950, 2011. doi: 10.2215/CJN.09241010

Introduction The best overall index of renal function is considered to be the glomerular filtration rate (GFR) (1,2). In the current Kidney Disease Outcome Quality Initiative Clinical Practice Guidelines for Chronic Kidney Disease (K/DOQI-CKD), all individuals with GFR ⬍60 ml/min per 1.73 m2 for ⱖ3 months are classified with chronic kidney disease (CKD), and different GFR levels are used for CKD staging, where “inulin clearance is widely regarded as the gold standard for measuring GFR” (2). As an alternative to the cumbersome and costly process of GFR measurement, several serum creatinine (SCr)-based formulas for estimating the renal function in adults were created in the past (3), and the most widely used ones are the CockcroftGault (CG) formula (4), which predicts urinary creatinine clearance, and the Modification of Diet in Renal Disease Study (MDRD) prediction equation (5) for GFR. We previously described in detail the differences between these two formulas and their performance in comparison with urinary inulin clearance (Cin) (6). The K/DOQI-CKD endorsed the MDRD equation as a clinically useful estimate of GFR and on the basis of this equation, the estimated CKD prevalence in multiple countries ranges from 10% to 15% (7–10). In the United States, the estimated CKD (stages 1 to 4) www.cjasn.org Vol 6 April, 2011

prevalence according to both albuminuria and decreased estimated GFR (eGFR) was 13.1%, representing 26.3 million people on the basis of the 2000 United States population census (7). The estimated prevalence of CKD in any given population depends on the formula used for eGFR because applying the recently created Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation reduced the estimated prevalence from 13.1% to 11.5% (11), representing a decrease of 3.2 million people. Nevertheless, CKD still remains a significant public health issue. Therefore, a proper estimation of the GFR has a pivotal role for the diagnosis and staging of CKD and is essential in nephrology practice. Furthermore, the eGFR has a very important role in large epidemiologic studies whenever GFR measurement is not feasible for practical and cost-containing reasons. The K/DOQI-CKD acknowledged that eGFR methods “are not completely sensitive or specific in detecting decreased GFR” and “thus, misclassification is possible, and clinicians should carefully consider all aspects of the patient’s clinical presentation in interpreting test results and determining evaluation and management.” Usually, an eGFR formula performance is assessed by the bias, precision, and accuracy of the eGFR values in comparison with the respective measured GFR (mGFR) ones. The role of mGFR as a confirmatory

*Nephrology Department, Hawaii Permanente Medical Group, Honolulu, Hawaii; †NancyUniversite´, Nancy, France; ‡Department of Nephrology, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands; § Coordination of the French ESRD registry REIN, Agence de la Biome´decine, Saint Denis la Plaine Cedex, France; and 储University Hospital Motol, Internal Clinic, Prague, Czech Republic Correspondence: Dr. Rossini Botev, Nephrology Department, Hawaii Permanente Medical Group, 3288 Moanalua Road, Honolulu, HI 96819. Phone: 808432-8050; Fax: 808432-8051; E-mail: Rossini.C. Botev@ kp.org

Copyright © 2011 by the American Society of Nephrology

937

938

Clinical Journal of the American Society of Nephrology

test for eGFR was very articulately discussed by Stevens and Levey (12). Bias, as a mean or median difference between eGFRs and mGFRs, reveals clinically whether overall the eGFR formula under- or over-estimates the mGFR. Precision, as a SD of mean bias or as an interquartile range of 25th to 75th percentile of median bias, represents the spread of eGFRs around the mGFRs. Accuracy, as a percentage of the distribution of eGFRs within particular range of mGFRs, combines bias (systemic deviation) and precision (dispersion). The purpose of this article is to help the clinician better understand the current limitations of measurement and estimation of GFR and their implications in daily practice in view of the current K/DOQI-CKD classification. Measurement of GFR Measurement of GFR is based on clearance of a substance that is not bound to proteins, freely filtered by the glomeruli, and it is not synthesized, transported, or metabolized by the kidneys. Creatinine meets most, but not all, of the requirements for an ideal endogenous GFR marker. It is secreted and reabsorbed by the renal tubules and extrarenally eliminated in severe CKD (13). These non-GFR determinants of urinary and serum concentrations introduce significant errors when calculating the creatinine clearance as a GFR measurement (2). Furthermore, the tubular creatinine secretion varies at different levels of mGFR (14) and in patients with severe proteinuria (15). Cin is accepted as the gold standard for mGFR (2,16,17). Several experimental studies by direct nephron micropunctures have confirmed that inulin fulfills the requirements for an ideal filtration marker (18 –21). Such studies cannot be practically conducted in humans for ethical reasons. In steady-state conditions, inulin is administered by continuous intravenous infusion to achieve constant serum concentration, and several timed urine collections are used to calculate the GFR. Unfortunately, the Cin method cannot be used routinely in daily practice because of its complexity. Other methods for mGFR using urinary or plasma clearance of exogenous agents such as iothalamate, ethylene diaminetetraaceticacid (EDTA), diethylene-triamine pentaacetic acid (DTPA), or iohexol, some of which are also cited to be “gold standards,” may have advantages in that they are less complex and time consuming. There is no standardized protocol for measuring GFR, and different clearance methods, timing periods for establishing steady-state concentrations, and assays of filtration markers are used among different clinical centers. Urinary 125I-iothalamate clearance (Cio), after a single bolus subcutaneous injection, was used for mGFR in the MDRD and CKD-EPI equation populations. Israelit et al. (22) found an excellent correlation coefficient (r) of 0.97 between Cio and Cin in 20 patients with CKD and two normal subjects. Furthermore, Perrone et al. (23), after studying 16 patients with CKD and four normal volunteers, concluded that Cio by single bolus subcutaneous injection can accurately measure Cin, but the authors acknowledged that Cio might overestimate Cin in normal subjects. In the latter study, data also show significant relative Cio-Cin bias of ⫺2.7% to 23.3% at low Cin levels

(range 17.6 to 24.3 ml/min per 1.73 m2) and a range of coefficient of variation (CV) between Cin and Cio of 13.2% to 27.9%, favorably comparing with CV of 19.9% in another study (24). Notghi et al. (25) reported for 76 mGFRs after subcutaneous injection of 125I-iothalamate a variance of 484 between Cio and Cin, which represents a SD of 22 ml/min per 1.73 m2. Ott et al. (26) studied in 84 individuals the intravenous and subcutaneous methods of 125I-iothalamate administration, and the Cio had residual SD of 10.7 and 6.37 ml/min, respectively, when compared with Cin. Figures 1 and 2 depict data from studies reporting simultaneous Cio and Cin measurements (23,27–30). Data with Cin ⬍5 ml/min per 1.73 m2 (29) were omitted because of large CV for Cin and Cio, most likely related to small urine volumes. Despite excellent correlation (r) and variability (r2) coefficients, the average Cio-Cin difference (bias) was 4.6 (range ⫺33.0 to 34.8) ml/min per 1.73 m2 (8.1% [range ⫺37.1% to 61.3%]). Most importantly, the bias was not uniformly distributed across the wide Cin spectrum. The precision (one SD of bias) was 11.1 ml/min per 1.73 m2 (17.6%), and accuracy of Cio within ⫾ 10% of Cin was only 44%. Odlind et al. (31) suggested that 125I-iothalamate is excreted not only by glomerular filtration but also by tubular secretion, and Petri et al. (32) found that Cio overestimates Cin in subjects with lupus nephropathy. Exploring differences between Cin and other GFR markers as EDTA, DTPA, or iohexol is beyond the scope of this review and was recently discussed (12). In summary, from these small studies, it is evident that Cio may differ significantly from Cin at different Cin levels, and thus a eGFR formula based on Cio as mGFR might have different discriminatory ability to properly estimate the GFR. Previous studies used mainly the correlation coefficient to assess the differences between Cio and Cin, but this does not reveal the dispersion of Cio values, and the best method to apply is accuracy. Obviously, there is a need to standardize the methodology for GFR measure-

Figure 1. | Relation between simultaneously measured GFRs by urinary clearances of intravenous infusion of inulin (Cin) and subcutaneous injection or intravenous infusion of 125I-iothalamate (Cio). The trend line is in bold. y, regression equation; r, correlation coefficient; r2, variability coefficient. The data were compiled from five studies including 94 subjects (23,27–30).

Clin J Am Soc Nephrol 6: 937–950, April, 2011

The Clinician and Estimation of GFR, Botev et al. 939

Figure 2. | Difference (bias) between simultaneously measured GFRs by urinary clearances of intravenous infusion of inulin (Cin) and subcutaneous injection or intravenous infusion of 125I-iothalamate (Cio) at different Cin levels. Total bias by Bland-Altman plot (A), and relative bias at different Cin levels (B). The dashed lines represent mean difference and respective ⫹2 and ⫺2 SD limits. The data were derived from five studies including 94 subjects (23,27–30).

ment and establish calibration coefficients for the currently used GFR markers by simultaneous measurements in much larger and diverse populations. GFR Adjustment for Body Surface Area (BSA) In the 19th century, it was accepted that metabolic rate is proportional to BSA, and indexing physiologic variables, including GFR, to BSA was introduced to adjust for differences in body size. Delanaye et al. (33,34) and Geddes et al. (35) elaborated significant relevant arguments against the BSA adjustment of GFR, most importantly that BSA indexing fails to equalize the GFR values of two subjects with different BSA. In a review article, Hsu et al. (36) concluded that there is no consensus as to whether or not to normalize GFR for BSA. Indexing the CG formula values for BSA is often referred to as eGFR, but this practice is questionable because weight is already included as a variable. Furthermore, Cockcroft and Gault stated that for patients with marked obesity or ascites, the absolute weight should be substituted with lean or ideal body weight (4). Recently, in a very large data set, lean body mass improved the prediction of eGFR (37,38). Macdonald et al. (39,40) demonstrated that muscle mass, compared with the MDRD equation variables, explained more variance in the mGFR and improved its estimation. If the change of weight is adding fat, correction for BSA lowers the GFR value artificially. Obesity is associated with hyperglycemia, microalbuminuria, and hypertension, all of which may initially cause glomerular hyperfiltration. After excluding subjects with these confounding factors, Anastasio et al. (41) observed no significant difference in absolute GFR between obese (body mass index [BMI] ⫽ 46.8 ⫾ 1.78 kg/m2, BSA ⫽ 2.18 ⫾ 0.04 m2) and nonobese (BMI ⫽ 25.6 ⫾ 0.93 kg/m2, BSA ⫽ 1.76 ⫾ 0.04 m2) subjects, but the BSA-adjusted GFR significantly decreased from 106.6 ⫾ 3.0 ml/min to 84.1 ⫾ 2.32 ml/min per 1.73 m2 in the obese group. These findings have significant impli-

cations in view of the estimated 26.7% prevalence of obesity in the United States (42). In clinical practice, for purposes of drug dosage, the absolute GFR should be used instead of the BSAadjusted one, and this is recommended by the National Kidney Foundation (43). In the kidney transplant evaluation process, indexing for BSA may eliminate potential donors by artificially lowering their GFR. Conversely, adjusting the GFR of a small-sized donor will spuriously increase its actual value, but this would be irrelevant for a normal-sized recipient. Therefore, in these case scenarios, the BSA indexing of GFR is not advisable. The MDRD and CKD-EPI equations are BSA-adjusted by intrinsic design, therefore “eliminating” the need for BSA indexing of their eGFR values. There are no data for mean BSA in the development group for CG formula. The MDRD and CKD-EPI equations included populations with mean BSA of 1.91 ⫾ 0.23 and 1.93 ⫾ 0.20 m2 and BMI of 28 (calculated) and 28 ⫾ 6 kg/m2, respectively, revealing overall overweight populations (5,11). Reversing the BSA indexing process shows mean mGFR increase from 39.8 ml/min per 1.73 m2 to 43.9 ml/min and 68.0 ml/min per 1.73 m2 to 75.9 ml/min for the MDRD and CKD-EPI equation populations, respectively. Therefore the “BSA-adjusted” MDRD and CKDEPI equations decreased the absolute mGFR of their included populations on average by ⫺4.1 ml/min (⫺9.3%) and ⫺7.9 ml/min (⫺10.4%), respectively. The BSA bias could be a reason for the need to refit the MDRD equation for Chinese and Japanese populations in view of their lower mean BSA of 1.7 ⫾ 0.18 and 1.65 ⫾ 0.19 m2 and BMI of 23.6 ⫾ 3.6 and 23 kg/m2 (calculated), respectively (44,45). Therefore, large population samples with diverse body size need to be studied to determine what the best method for GFR indexing might be.

940


Normal GFR According to Age, Gender, and Race Normal mGFR values in whites are approximately 130 ml/min per 1.73 m2 for young men and 120 ml/min/1.73 m2 for young women and decline as persons age (16,17). Renal senescence was reviewed in details by Glassock et al. (46). Rule et al. (47), in an elegant study of kidney donor biopsies, revealed that nephrosclerosis increased with age, but it was not associated with CKD risk factors after adjustment for age and gender. Surprisingly, the observed mGFR decline with normal aging was not explained by nephrosclerosis. The K/DOQI-CKD defined GFR ⬍60 ml/ min per 1.73 m2 as a disease, regardless of age, gender, or race. However, it was stated: “the cut-off levels between stages are inherently arbitrary.” Extrapolating individual data for 70 men without CKD, age 24 to 89 years (48), K/DOQI-CKD reported normal mean mGFR for age groups by decades on the basis of linear regression analysis, assuming that the values for women were 8% lower at all ages (2). For men age groups ⬎50 and women age groups ⬎40 years, the two-SD range is below 60 ml/min per 1.73 m2; therefore the K/DOQI-CKD classification will misdiagnose some of these subjects with CKD stage 3. After Davies and Shock’s study (48), published 60 years ago, there is still limited data for Cin in normal subjects by age groups (49), particularly for the elderly (age ⬎75 years). The significant limitations of this study are the inclusion of men only and the small number of elderly participants: nine in the age group 70 to 79 years and 12 in the age group 80 to 89 years. A large study of a healthy Chinese population with mGFR by 99mTc-DTPA included a slightly higher number of elderly subjects: 49 in the age group of 70 to 80 years and 12 in the age group of ⬎80 years (50). Three other large studies with Cio reported mGFR by age groups in kidney donors (presumably normal individuals) from the United States, but all of the individuals were aged ⬍75 years (51–53), and only one study reported data for African Americans (53). Obviously, large studies to establish the boundaries of normal mGFR according to a wide range of ages in different genders and races are needed to resolve recent debates (54 –57) about overdiagnosing CKD in the elderly, a population of 16.6 million of ⬎75 years according to the 2000 United States population census, and estimated to be 18.8 million as of 2009 (58). Hopefully, the ongoing Berlin Initiative Study in the elderly would provide some answers for proper CKD diagnosing in this population (59).

Reporting the eGFR The K/DOQI-CKD panel should be commended for introducing the eGFR reporting by clinical laboratories. This has significant clinical implications because it reminded the practitioners that the SCr concentration becomes “abnormal” after up to 50% mGFR loss (14). Unfortunately, by requesting automatic laboratory eGFR reporting whenever SCr is ordered (60), the current K/DOQI-CKD and National Kidney Disease Education Program guidelines created several controversies in clinical practice that were extensively debated elsewhere (54 –57,61– 64). In 2007, the Australasian Creatinine Consensus Working Group revised their recommendation to extend the upper reporting limit of eGFR from 60 to 90 ml/min per 1.73 m2 and appropriately ad-

vised the practitioners that eGFR 45 to 59 ml/min per 1.73 m2 in people ⱖ70 years of age may not be associated with CKD, providing the eGFR was stable over time and no other CKD markers, such as proteinuria, were present (65). For mGFR ⬎60 ml/min per 1.73 m2, the new CKD-EPI equation, compared with the MDRD one, had a significantly lower median bias (⫺3.5 versus ⫺10.6 ml/min per 1.73 m2, respectively) in a large validation data set. If this result is confirmed in other large populations, it will be helpful for the practitioner to change the current practice by many laboratories to report simply “eGFR ⬎60 ml/min per 1.73 m2.” For example, for a 30-year-old white woman with SCr of 1.1 mg/dl, the eGFR (CKD-EPI) is 68 ml/min per 1.73 m2, representing approximately 40% loss of the normal mGFR. Even after accounting for the formulas’ underestimations at this level of mGFR, this is a significant eGFR loss, something of which the clinician would not be aware by looking at the eGFR ⬎60 ml/min per 1.73 m2 report. All combined, these new recommendations would prevent missing younger subjects with decreased eGFR in the range of 60 to 90 ml/min per 1.73 m2 who might have early asymptomatic CKD and minimize inappropriate CKD diagnosing in the elderly. Furthermore, a recent costbenefit analysis of the automatic eGFR reporting suggested “that reporting eGFR may be beneficial, but this limited benefit was reversed with virtually any reduction in quality of life caused by incorrect diagnosis of CKD” (66). Therefore, the eGFR report should be requested only by practitioners who are familiar with the pitfalls of its interpretation.

SCr Calibration: The Need for International Normalized Ratio for SCr Measurement The National Kidney Disease Education Program Laboratory Working Group launched a global SCr standardization effort to reduce interlaboratory variation. Creatinine assays using the Jaffe reaction may have higher SCr values because of interference from noncreatinine chromogens, thus creating mostly underestimation of the high levels of eGFR (67). One study (68) revealed that the mean eGFR bias of the simplified (four variables) original MDRD equation (69) with Beckman CX3 kinetic Jaffe assay was reduced from 6.7% to 0.4% for the isotope dilution mass spectrometry (IDMS) assay, 8.8% to 3.4% for the Roche enzymatic assay, and 13.7% to 7.1% for the Roche compensated kinetic Jaffe assay when applying the re-expressed IDMS-traceable MDRD formula (70). Figure 3 shows the eGFR bias, range 0% to 24.2%, between the original and IDMS-traceable MDRD equations at different IDMS-traceable SCr levels. Interestingly, this eGFR calibration bias range is much lower than the Cio-Cin one (⫺37.1% to 61.3%), thus reinforcing the need of further research to “calibrate” any differences between various GFR markers. The National Institute for Standards and Technology in the United States released only two levels of Standard Reference Material for SCr (SRM 967), 0.75 and 3.92 mg/dl, for calibrating the SCr measurements to a uniform standard, the IDMS. The actual process of standardization is at the discretion of each laboratory. Using only two point references to calibrate the entire spectrum of SCr values presumes that all calculated “calibrated” values would not



Figure 3. | Difference (bias) between eGFRs by original, (constant ⴝ 186) and serum creatinine (SCr) by Beckman Astra CX3 assay, and re-expressed, (constant ⴝ 175) and SCr by isotope dilution mass spectometry (IDMS) assay, simplified (4 variables) Modification of Diet in Renal Disease Study (MDRD) equation at different SCr measurements by IDMS assay for a hypothetical 60-year-old Caucasian male. Bias for eGFR was calculated as re-expressed MDRD minus original MDRD values as ml/min per 1.73 m2 (A) and as a percentage (B). SCr was concentration converted to mg/dL by dividing ␮mol/L by 88.4. The data were extrapolated from Table 2 in the publication of Vickery et al. (68).

deviate from the identity line. An analytic CV of SCr values among laboratories as low as 2% translates into a critical difference of 12.4% and has significant clinical implications for estimating the GFR, especially in the low range of SCr levels (71). Furthermore, when three previously reported linear equations (72–74) and a proposed correction value (75) for SCr calibration to the MDRD Study laboratory were tested in 10,108 subjects, those methods of calibration did not show valid reproducible eGFR values for purposes of K/DOQI-CKD classification (76). Extensive calibration methods of SCr assays from several laboratories had very minimal or no effect on the MDRD equation’s accuracy in six of the nine studied populations (excluding the MDRD one) (77). Therefore, there is a need to establish an international normalized ratio for SCr measurement by including reference IDMS values for a broad SCr spectrum to improve the current standardization process. Evaluating the Performance of eGFR Formulas Evaluating the performance of prediction equations by mainly emphasizing the bias could be misleading, because a bias of zero does not assure perfect performance if the positive and negative eGFR differences are equally spread around the identity line of mGFR. Precision is a very helpful additional test to clinically interpret the bias. Total (ml/min per 1.73 m2) and relative (%) values for bias and precision have different implications at different GFR levels, because they should be interpreted in a particular clinical context. For example, small eGFR total bias and precision suggest a good performance; however, this could simply reflect low mGFR values, and high relative bias and precision may unmask poor performance. Accuracy, expressed as a percentage of eGFRs within a percentage range of mGFRs, combines precision and bias by expressing how many GFR estimates are dispersed, i.e. the density, within a given range of their respective GFR measurements. Therefore, it is the most useful statistical analysis for the clinician. Because day-to-day mGFR variation was

reported to be as high as 17% (3,23,78), the accuracy within ⫾ 15% of mGFR provides the clinician with the percentage of eGFRs not deviating from the mGFRs. PubMed database research (last accessed July 2010) for data sets reporting accuracy for the CG and MDRD formulas in adults with a minimum of 100 mGFRs, because estimates from small studies can be unreliable (2), recovered 61 publications. Studies with subsets of elsewhere-reported data sets were excluded. Increasing the minimal number of mGFRs to 300 narrowed the field to 25 publications (33 data sets), which are presented in Table 1 (6,11,37,38,44,45,72,77,79 –95). The most frequently reported accuracy range was within ⫾ 30%, followed by within ⫾ 50%, which has very little clinical application. Less than half of the studies provided accuracy within ⫾ 15%, median bias, and precision by SD or interquartile range; therefore no analysis was performed for these results. The range of mean bias (eGFR–mGFR, calculated whenever not provided) was ⫺9.0 to 13.6 and ⫺29.2 to 12 ml/min per 1.73 m2, and accuracy within ⫾ 30% ranged from 49% to 88% and 53% to 89%, for CG and MDRD equations, respectively. The number of mGFRs in the data sets varied from 322 to 5504. Thus, for comparison of both equations’ performance, after excluding three data sets with significant difference in the number of CG and MDRD estimates (80), a weighted average bias and accuracy were calculated as follows,

x៮ ⫽

冘冘

n i⫽1wi xi n i⫽1wi

where w is the number of mGFRs, and x is the bias or accuracy of eGFRs in a particular data set. The weighted average bias was 0.02 versus ⫺5.34 ml/min per 1.73 m 2 (6,45,72,80 – 87,89,92,93,95), and accuracy within ⫾ 30% was 72% versus 78% (6,37,38,45,72,77,79 – 87,89,91–93,95), for the CG and MDRD formulas, respectively. The weighted average

125

125

125

I- iothalamate

I-iothalamate I-iothalamate Inulin 51 Cr-EDTA 10% plasma 125 I-iothalamate

125

GFR Method

Mariat et al. (92)

Hojs et al. (89) Ma et al. (90)b Burke et al. (91)

Saleem et al. (86) Srinivas et al. (87) Rule et al. (88)a

Bosma et al. (84) Ma et al. (44) Gera et al. (85)a

CKD

No Pre-liver transplant 3months after liver transplant 1 year after liver transplant 5 years after liver transplant Yes (ClClin) DM type 1 No Mixed

NA

Mixed Mixed Mixed CKD and KT donors

Yes (IDMS) Yes (IDMS) No Yes (ClClin)

Population

No SCr ⬍ 1.5 mg/dl Yes (ClClin) CKD KT donors 125 No KT recipients I-iothalamate 99m Yes (ClClin) CKD Tc-DTPA plasma 125 No 3-week KT recipients I-iothalamate 1-year KT recipients 2-year KT recipients 3-year KT recipients 99m No Oncology Tc-DTPA plasma 99m KT donors Tc-DTPA renogram No Iothalamate No Healthy CKD 51 No CKD Cr-EDTA plasma 99m NA CKD Tc-DTPA plasma DTPA, iohexol, EDTA, NA KT recipients iothalamate, inulin Inulin No KT recipients

Ibrahim et al. (81) I-iothalamate Nyman et al. (37) Iohexol plasma and Bjork et al. (38) 99m Verhave et al. (82) Tc-DTPA 125 Poggio et al. (83) I-iothalamate

K/DOQI-CKD (2) for Lewis et al. (79) Gonwa et al. (80)

Stevens et al. (77) Levey et al. (11) Botev et al. (6) Froissart et al. (72)

Study

SCr Calibration

97

122 ⫾ 23 55 (9 to 121)

476

850 828 457 798 684 683 684 506 346 601 599 580 320 592 567 546

1286 850

77 73 71 66 92

85

87

55 ⫾ 26

521

47 ⫾ 16

90

63 ⫾ 28

1297

82

84 79

90.7 ⫾ 40.5 NA

1447 887

98 77 97

94

57 ⫾ 33

1775

99.3 ⫾ 20.2 32 ⫾ 28 106 ⫾ 18 55 ⫾ 18 55.1 ⫾ 35.1 55.8 ⫾ 18 54.7 ⫾ 18 NA NA 83 (4 to 214) 95.5 ⫾ 11.6 101 ⫾ 17 48 ⫾ 25 47 ⫾ 34 57.9 ⫾ 36.4 NA

89 91

5504 3896 2208 2095

50%

30%

Accuracy 15%

91

91 89

94 85 91 83 83 91

99 89 98

98

92

92

86 85

98

94 96

59

64

63

56 55 52 49 73 84

87 60 85 76

88 79

64

64

61 57

80

71 79

69

73

80 89 71 86 88 66 62 72 65 65 75 80 54 75 53 76 73

78

66

68

67 62

88

83 81 69 87

66

38 49

43

54

30

34 58

CG MDRD CG MDRD CG MDRD 68 ⫾ 40 68 ⫾ 36 72.4 ⫾ 39 61.1 ⫾ 32.7

Mean n mGFR ⫾ SD mGFRs (Range)

Table 1. Performance of CG and MDRD formulas in studies with minimum of 300 measured GFRs

6.6

ⴚ4.9

8.7 10.2 11.6 13.6 ⴚ5.1 ⴚ9.0

ⴚ4.9 5.0 3.0 6.3

ⴚ6.0

ⴚ8.5

ⴚ3.5 1.9

CG

3.8

ⴚ4.3 ⴚ5.0 0.8 2.5 5.1 7.5 ⫺29.0 ⫺5.5 ⴚ29.2

ⴚ12.4 ⴚ1.0 ⴚ9.0 ⴚ3.2

ⴚ12.0

⫺9.4

⫺9.9

ⴚ6.9 NA

ⴚ9.8 ⴚ1.0

MDRD

Mean

3.5 1.9

⫺5.5

⫺1.2

⫺7.8

⫺0.5 ⫺9.0

3.4

2.7 5.5

MDRD

Median CG

Bias

25.8

21.5 15.4

CG

17.2

20.0 13.7

MDRD

Precision SD

942 Clinical Journal of the American Society of Nephrology

I-iothalamate Inulin

Inulin Iohexol plasma

Horio et al. (94)d Tidman et al. (95)

Population

Yes (IDMS) KT donors Yes (ClClin) CKD CKD, KT donors Yes (IDMS) CKD Yes (IDMS) NA

SCr Calibration

423 413 350 350 322

106 ⫾ 16 59.1 ⫾ 35.4 57.2 ⫾ 34.7 45 ⫾ 25 50.4 ⫾ 28.1

Mean n mGFR ⫾ SD mGFRs (Range) 30%

15%

85

92 69

88 76 75

86 59 59 73 80 44 45

36 39

CG MDRD CG MDRD CG MDRD

50%

Accuracy

5.9

ⴚ1.7

2.0

CG

12.0 ⫺1.3 ⴚ1.0

ⴚ16.0

MDRD

Mean

⫺0.5 ⫺16.3

MDRD

Median CG

Bias

13.2

⫺19.6

CG

22.2 19.4 12.9

MDRD

Precision SD

The data in bold type were included for average weighted analysis (see text). CG, Cockcroft-Gault formula: MDRD, Simplified (4 variables) Modification of Diet in Renal Disease Study equation; mGFR, measured GFR in ml/min per 1.73m2, urinary clearance unless otherwise described; 51Cr-EDTA, 51chromium-ethylene diaminic tetraacetic acid; 99mTc-DTPA, 99m technetium-diethylenetriamine penta-acetate; SCr, serum creatinine, calibrated to ClClin (Cleveland Clinic Laboratory) or IDMS (isotope dilution mass spectroscopy); CKD, chronic kidney disease; KT, kidney transplant; DM, diabetes mellitus; mGFR, measured GFR in ml/min per 1.73 m2; Accuracy, percentage of GFR estimates within ⫾ 50%, ⫾ 30%, and ⫾ 15% of mGFRs; Bias, estimated minus measured GFR in ml/min per 1.73 m2 (mean bias calculated if possible whenever not provided); Precision, one standard deviation (SD) of mean bias in ml/min per 1.73 m2; estimated GFR by CG in ml/min per 1.73 m2 (ml/min in references 37,80,84,85), data in italic font for corrected CG by adjusted (37) or lean body weight (86), or 0.84 for creatinine tubular secretion (87); estimated GFR by MDRD in ml/min per 1.73 m2, data in italic for IDMS re-expressed MDRD equation. a Study design included SCr measurements not exactly on the day of GFR measurement. b MDRD modified for Chinese. c CG modified for Japanese. d MDRD modified for Japanese.

125

GFR Method

Issa et al. (93) Matsuo et al. (45)c

Study

Table 1. (Continued)

Clin J Am Soc Nephrol 6: 937–950, April, 2011 The Clinician and Estimation of GFR, Botev et al. 943

944


accuracy within ⫾ 30% in data sets with mean mGFR ⬎60 ml/min per 1.73 m2 was 74% versus 80% and in data sets with mean mGFR ⬍60 ml/min per 1.73 m2 was 70% versus 75%, for the CG and MDRD formulas, respectively. The weighted average accuracy within ⫾ 30% in data sets with calibrated SCr was 74% versus 81%, and in data sets with noncalibrated SCr was 70% versus 74%, for the CG and MDRD formulas, respectively. Because the data sets had different markers for mGFR, a confounding factor for the MDRD equation, further analysis showed that the MDRD weighted accuracy within ⫾ 30% in studies with mGFR by Cio was 78%, similar to the performance of 76% in all 33 data sets. The CKD-EPI equation, compared with the MDRD

one, had 3.6% better overall accuracy in a large validation data set (11). Only six studies (two of those with ⬎300 mGFRs) were recovered to report accuracy within ⫾ 30% for the new CKD-EPI equation, range 59% to 89% (11,94,96 –99). Unfortunately, there is no agreement in the literature on what constitutes excellent accuracy. Applying the widely accepted statistical analysis of 95% confidence interval (CI) as a criterion would imply that a formula has an excellent accuracy when 95% of its GFR estimates are within a ⫾15% range of their respective GFR measurements. Recently, for graphic presentation of the mGFR and eGFR relation, some authors proposed plotting the mGFR

Figure 4. | Different trend line (in bold) appearances when measured (by inulin) and estimated (by Cockcroft-Gault [CG] formula and Modification Diet in Renal Disease Study [MDRD] equation) GFR values are plotted on X and Y axes (A and B) and Y and X axes (C and D), respectively. Only the trend lines in A and B illustrate correctly the reported results for difference (bias) between estimated and measured GFR in the study (6). The MDRD equation was simplified to four variables. r2, variability coefficient.

51

Inulin Cr-EDTA 10% plasma 99m Ma et al. (90)b Tc-DTPA plasma Mariat et al. (92) Inulin Tidman et al. (95) Iohexol plasma Chudleigh et al. (103) 51Cr-EDTA plasma 99m Ma et al. (44) Tc-DTPA plasma a 125 Michels et al. (99) I-iothalamate

Urbaniak et al. (107) All studies MDRD equation Botev et al. (6) Froissart et al. (72)

White et al. (106)

No No

Inulin 51 Cr-EDTA plasma 99m Tc-DTPA plasma Iohexol plasma

Rostoker et al. (104) Rigalleau et al. (105)

65

Yes (IDMS)

KT donors and others

29 176

84

298 200 155

370

65

KT recipients 63 N/A 62 DM type 2 53

1261 1483

64 4129

57 71

64 64

Yes (ClClin) CKD

No Yes (IDMS) No

No Mixed Yes (ClClin) CKD and KT donors NA CKD

Mixed

136

Yes (ClClin) KT recipients 69

No

193 110

72 55

170

63

Yes (IDMS)

KT donors and others CKD DM

270 206 202

271

286

476 322 292

567

2208 2095

100 6431

198

269 200

271

476 322 292

2208 2095

eGFR mGFR

57 64 69

% 1360 1416

Population

62 68

SCr Calibration

No Mixed Yes (ClClin) CKD and KT donors No KT recipients Yes (IDMS) N/A No DM type 2

51

mGFR Method

Inulin Cr-EDTA 10% plasma Mariat et al. (92) Inulin Tidman et al. (95) Iohexol plasma Chudleigh et al. (103) 51Cr-EDTA plasma 125 Michels et al. (99)a I-iothalamate

CG formula Botev et al. (6) Froissart et al. (72)

Study

Overall

65

26

36 49

45

49 67

36 69

25

78 75

85

74 69

64 72

%

53

33

14 126

54

396 322

4 1253

4

61 39

70

29 178

520 348

eGFR

⬎90

82

125

39 255

121

807 482

11 1822

16

78 52

82

39 255

807 482

mGFR

63

32

48 57 78

68

56 64

82 56

63

60 38

54

49 63 65

53 59

%

59

51

50 43 29

85

286 367

9 905

50

55 11

50

51 47 24

271 338

eGFR

60–90

93

161

104 75 37

125

512 576

11 1607

79

91 29

93

104 75 37

512 576

mGFR

68

72 76

81

73 78

71 72

83

78 66

57

64 75

71 78

%

47

215 87

134

353 466

15 1317

73

62 44

39

191 85

343 465

eGFR

30–59

69

298 114

166

484 597

21 1818

88

80 67

69

298 114

484 597

mGFR

63

45 58

62

54 79

64 59

60

68 44

41

38 46

62 68

%

17

33 47

54

147 246

9 505

9

14 16

11

28 37

170 211

eGFR

15–29

27

74 81

87

275 312

14 854

15

20 36

27

74 81

275 312

mGFR

69

63

60 65

63 45

0

62

44 43

%

mGFR Groups in ml/min per 1.73 m2 according to K/DOQI-CKD Stages

Table 2. Classification performance of CG and MDRD formulas in mGFR groups according to K/DOQI-CKD Stages

9

43

78 83

27 147

0

8

57 55

eGFR

⬍15

13

68

130 128

43 330

16

13

130 128

mGFR

Clin J Am Soc Nephrol 6: 937–950, April, 2011 The Clinician and Estimation of GFR, Botev et al. 945

86 65 14 941 8 604 57 64 21 1984 19 1520 11 1893 70 4536 Mixed

70 62 No

100 7284

9 52

1 1075

11 2068

45 57

5 1080

90 77

15 14 93 88 74 84 79 36 46 16 4 25 198 128 Yes (ClClin) KT recipients 65

The data for the percentage or number of eGFRs were extrapolated if not provided. CG, Cockcroft-Gault formula; MDRD, Simplified (4 variables) Modification of Diet in Renal Disease Study equation; mGFR, measured GFR in ml/min per 1.73m2, urinary clearance unless otherwise described; K/DOQI-CKD, Kidney Disease Outcome Quality Initiative Clinical Practice Guidelines for Chronic Kidney Disease; eGFR, estimated GFR in ml/min per 1.73m2 by CG and MDRD, data in italic font for isotope dilution mass spectroscopy (IDMS) re-expressed MDRD equation; 51Cr-EDTA, 51chromium-ethylene diaminic tetraacetic acid; 99mTc-DTPA, 99mtechnetium-diethylenetriamine penta-acetate; SCr, serum creatinine, calibrated to ClClin (Cleveland Clinic Laboratory) or IDMS; CKD, chronic kidney disease; KT, kidney transplant; DM, diabetes mellitus. a Study design included SCr measurements not exactly on the day of GFR measurement. b MDRD modified for Chinese.

43 398 37 257

16 7 44 20 36 16 22 78 61 80 67 66 59 91 29 78 52 41 31 53 60 269 200 180 131 67 65 CKD DM No No

eGFR

62 41

56 12

82 88

eGFR eGFR eGFR mGFR eGFR

Rostoker et al. (104) Inulin Rigalleau et al. (105) 51Cr-EDTA plasma 99m White et al. (106) Tc-DTPA plasma Urbaniak et al. (107) Iohexol plasma All studies

Study

mGFR Method

SCr Calibration

Population

%

Overall

eGFR mGFR

%

⬎90

mGFR

%

60–90

%

30–59

mGFR

%

15–29

mGFR

%

mGFR Groups in ml/min per 1.73 m2 according to K/DOQI-CKD Stages

⬍15

mGFR


Table 2. (Continued)

946

values on the Y axis (100,101). The different graphic appearances by interposing mGFR and eGFR values are shown in Figure 4. The bias results, underestimating the high and overestimating the low mGFR levels in the presented study (6), are properly illustrated with mGFR on the X axis but distorted when mGFR is on the Y axis. Particularly, the trend line for MDRD equation implies erroneously an underestimation of mGFR at all levels. The SCr concentration is determined by the level of GFR when the measurement is done in a steady state and under standardized conditions to account for endogenous and exogenous non-GFR creatinine determinants. In the models for eGFR prediction, besides the SCr concentration, SCr-based equations include age, gender, race, or weight, as surrogates for differences in creatinine generation from muscle mass (4,5,11), on the basis of statistically significant regression coefficients for those variables in relation to the mGFR. These models may not account for variations in the tubular handling of creatinine (15). Hence, mGFR is the independent variable, and eGFR is the dependent one. Therefore, mGFR values should be plotted on the X axis, and eGFR values should be plotted on the Y axis, as illustrated in a similar model (102). Effect of eGFR on K/DOQI-CKD Diagnosing and Staging For the practicing clinician, the ultimate evaluation of any formula’s performance is its proper eGFR assignment into the five GFR groups of the K/DOQI-CKD classification where the GFR groups are defined by mGFR. PubMed research (last accessed July 2010) recovered only 12 studies with mGFRs ⬎100 reporting this performance of the CG and MDRD formulas, and their results are presented in Table 2 (6,44,72,90,92,95,99,103– 107). Comparing CG versus MDRD formula in studies with data for both formulas (6431 subjects) showed similar overall proper classification of the subjects, 64% versus 63%, range 55% to 72% versus 53% to 71%, respectively (6,72,92,95,99,103–107). The best performance of CG and MDRD equations was in mGFR group 30 to 59 ml/min per 1.73 m2, whereas the worst one was in mGFR groups ⬍15 and ⬎90 ml/min per 1.73 m2, respectively. Proper overall eGFR assignment was 67% and 65% in studies with SCr calibration and 62% and 59% without SCr calibration, for the CG and MDRD formulas, respectively. Only one study was found to evaluate the CKD-EPI equation in GFR groups by mGFR, and its proper eGFR overall assignment was 69% (99). These results suggest that eGFR prediction equations should be used with great caution for CKD diagnosing or staging given the clinical and therapeutic consequences for millions of people, i.e. approximately 10 million (38%) subjects might have been misclassified by the MDRD equation on the basis of estimated CKD prevalence of 26.3 million adults in the United States. Some authors proposed that the CKD staging should be done by eGFR, as a category of evaluation, instead of mGFR for two main reasons: mGFR is not identical to true GFR because of errors in the measurement process, and the objective in development of eGFR equations is to determine an eGFR that is unbiased for the mGFR for populations with a given level of eGFR (101). First, although the GFR measurement by calculated clearance of a given substance may vary from the actual GFR


because of errors in the measurement process, there are currently no other practical means to measure the glomerular filtration volume. Similarly, the GFR estimation is also affected by errors in the SCr measurement because if the same sample were measured repeatedly, a CV range from 1% to 11% has been reported (36). Any repeated measurements have some errors related to the process of measurement, and the solution is to minimize the errors by standardizing the process instead of substituting the measurement with an estimate, especially when the eGFR deviates ⬎30% from the mGFR in approximately 10% to 50% of the estimates. Second, the purpose of calculating eGFR is to avoid the cumbersome measurement of GFR in daily practice. For diagnosis and classification of CKD, the clinician is primarily interested in the reliability of estimating the mGFR. Therefore, CKD staging (GFR stratification) by eGFR and then evaluating the performance of an eGFR equation in the thuscreated stages is not clinically helpful, because it omits the mGFR confirmatory test. Conclusion The introduction of the K/DOQI-CKD classification was a landmark event allowing CKD to be recognized as a global public health problem. Estimating the GFR is now an integral part of the daily clinical practice, and it is used routinely for evaluation and monitoring of the renal function. The clinician should be aware of the significant limitations of the CG and MDRD formulas for purposes of diagnosing and staging CKD in the individual patient because of limited accuracy and significant misclassification. Because current SCr-based formulas are derived from a particular population with a specific marker of mGFR, reference IDMS values for a broad SCr spectrum need to be established, and further research is required to define calibrating coefficients among different GFR markers and appropriate GFR indexing for subjects with different body sizes. This would improve the formulas’ performance in daily practice and allow eGFR comparisons between different populations in research studies. After calibration of the mGFR data from various studies of healthy subjects, boundaries of age-, gender-, and race-related normal ranges of mGFR should be established and incorporated in the laboratories’ eGFR report. Then what constitutes CKD by decreased levels of mGFR without kidney pathology markers (such as proteinuria) can be defined. All combined, this should prevent underdiagnosing CKD by revealing early decreased GFR in young individuals, those who will benefit the most from an early nephrology referral, and avoid over-diagnosing CKD in elderly individuals by medicalization of renal senescence. Great caution should be exercised when embracing new definitions of disease in certain populations to preserve the first rule of practicing medicine: primum non nocere. Disclosures None.

References 1. Smith HW: Comparative physiology of the kidney. In: The Kidney: Structure and Function in Health and Disease, New York, Oxford University Press, 1951, pp 520 –574 2. K/DOQI clinical practice guidelines for chronic kidney dis-


3. 4. 5.

6.

7. 8.

9.

10. 11.

12. 13. 14. 15. 16.

17. 18.

19. 20. 21.

22. 23.

ease: Evaluation, classification, and stratification. Am J Kidney Dis 39[Suppl 1]: S1–S266, 2002 Kemperman FA, Krediet RT, Arisz L: Formula-derived prediction of the glomerular filtration rate from plasma creatinine concentration. Nephron 91: 547–558, 2002 Cockcroft DW, Gault MH: Prediction of creatinine clearance from serum creatinine. Nephron 16: 31– 41, 1976 Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D: A more accurate method to estimate glomerular filtration rate from serum creatinine: A new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med 130: 461– 470, 1999 Botev R, Mallie´ JP, Couchoud C, Schu¨ck O, Fauvel JP, Wetzels JF, Lee N, De Santo NG, Cirillo M: Estimating glomerular filtration rate: Cockcroft-Gault and Modification of Diet in Renal Disease formulas compared to renal inulin clearance. Clin J Am Soc Nephrol 4: 899 –906, 2009 Coresh J, Selvin E, Stevens LA, Manzi J, Kusek JW, Eggers P, Van Lente F, Levey AS: Prevalence of chronic kidney disease in the United States. JAMA 298: 2038 –2047, 2007 Hallan SI, Coresh J, Astor BC, Asberg A, Powe NR, Romundstad S, Hallan HA, Lydersen S, Holmen J: International comparison of the relationship of chronic kidney disease prevalence and ESRD risk. J Am Soc Nephrol 17: 2275–2284, 2006 Chadban SJ, Briganti EM, Kerr PG, Dunstan DW, Welborn TA, Zimmet PZ, Atkins RC: Prevalence of kidney damage in Australian adults: The AusDiab kidney study. J Am Soc Nephrol 14[Suppl 2]: S131–S138, 2003 Imai E, Matsuo S: Chronic kidney disease in Asia. Lancet 371: 2173–2182, 2008 Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF 3rd, Feldman HI, Kusek JW, Eggers P, Van Lente F, Greene T, Coresh J: A new equation to estimate glomerular filtration rate. Ann Intern Med 150: 604 – 612, 2009 Stevens LA, Levey AS: Measured GFR as a confirmatory test for estimated GFR. J Am Soc Nephrol 20: 2305–2313, 2009 Perrone RD, Madias NE, Levey AS: Serum creatinine as an index of renal function: new insights into old concepts. Clin Chem 38: 1933–1953, 1992 Shemesh O, Golbetz H, Kriss JP, Myers BD: Limitations of creatinine as a filtration marker in glomerulopathic patients. Kidney Int 28: 830 – 838, 1985 Branten AJ, Vervoort G, Wetzels JF: Serum creatinine is a poor marker of GFR in nephrotic syndrome. Nephrol Dial Transplant 20: 707–711, 2005 Smith HW: The reliability of inulin as a measure of glomerular filtration. In: The Kidney: Structure and Function in Health and Disease, New York, Oxford University Press, 1951, pp 231–238 Wesson L: Renal hemodynamics in physiological states. In: Physiology of the human kidney, New York: Grune & Stratton, 1969, pp 96 –108 Hendrix JP, Westfall BB, Richards AN: Quantitative studies of the composition of glomerular urine: XIV. The glomerular excretion of inulin in frogs and necturi. J Biol Chem 116: 735–747, 1936 Richards AN, Bott PA, Westfall BB: Experiments concerning the possibility that inulin is secreted by renal tubules. Am J Physiol 123: 281–298, 1938 Gutman Y, Gottschalk CW, Lassiter WE: Micropuncture study of inulin absorption in the rat kidney. Science 147: 753–754, 1965 Chang RL, Ueki IF, Troy JL, Deen WM, Robertson CR, Brenner BM: Permselectivity of the glomerular capillary wall to macromolecules. II. Experimental studies in rats using neutral dextran. Biophys J 15: 887–906, 1975 Israelit AH, Long DL, White MG, Hull AR: Measurement of glomerular filtration rate utilizing a single subcutaneous injection of 125I-iothalamate. Kidney Int 4: 346 –349, 1973 Perrone RD, Steinman TI, Beck GJ, Skibinski CI, Royal HD, Lawlor M, Hunsicker LG: Utility of radioisotopic filtration markers in chronic renal insufficiency: Simultaneous comparison of 125I-iothalamate, 169Yb-DTPA, 99mTc-DTPA, and inulin. Am J Kidney Dis 16: 224 –235, 1990

948


24. Agarwal R: Ambulatory GFR measurement with cold iothalamate in adults with chronic kidney disease. Am J Kidney Dis 41: 752–759, 2003 25. Notghi A, Merrick MV, Ferrington C, Anderton JL: A comparison of simplified and standard methods for the measurement of glomerular filtration rate and renal tubular function. Br J Radiol 59: 35–39, 1986 26. Ott NT, Wilson DM: A simple technique for estimating glomerular filtration rate with subcutaneous injection of (125I)lothalamate. Mayo Clin Proc 50: 664 – 668, 1975 27. Elwood CM, Sigman EM, Treger C: The measurement of glomerular filtration rate with 125I-sodium iothalamate (Conray). Br J Radiol 40: 581–583, 1967 28. Rosenbaum RW, Hruska KA, Anderson C, Robson AM, Slatopolsky E, Klahr S: Inulin: An inadequate marker of glomerular filtration rate in kidney donors and transplant recipients? Kidney Int 16: 179 –186, 1979 29. Skov PE: Glomerular filtration rate in patients with severe and very severe renal insufficiency: Determined by simultaneous inulin, creatinine and 125 iothalamate clearance. Acta Med Scand 190: 395–399, 1971 30. Anderson CF, Sawyer TK, Cutler RE: Iothalamate sodium I 125 vs cyanocobalamin Co 57 as a measure of glomerular filtration rate in man. JAMA 204: 653– 656, 1968 31. Odlind B, Ha¨llgren R, Sohtell M, Lindstro¨m B: Is 125I iothalamate an ideal marker for glomerular filtration? Kidney Int 27: 9 –16, 1985 32. Petri M, Bockenstedt L, Colman J, Whiting-O’Keefe Q, Fitz G, Sebastian A, Hellmann D: Serial assessment of glomerular filtration rate in lupus nephropathy. Kidney Int 34: 832– 839, 1988 33. Delanaye P, Radermecker RP, Rorive M, Depas G, Krzesinski JM: Indexing glomerular filtration rate for body surface area in obese patients is misleading: Concept and example. Nephrol Dial Transplant 20: 2024 –2028, 2005 34. Delanaye P, Mariat C, Cavalier E, Krzesinski JM: Errors induced by indexing glomerular filtration rate for body surface area: reductio ad absurdum. Nephrol Dial Transplant 24: 3593–3596, 2009 35. Geddes CC, Woo YM, Brady S: Glomerular filtration rate: What is the rationale and justification of normalizing GFR for body surface area? Nephrol Dial Transplant 23: 4 – 6, 2008 36. Hsu CY, Chertow GM, Curhan GC: Methodological issues in studying the epidemiology of mild to moderate chronic renal insufficiency. Kidney Int 61: 1567–1576, 2002 37. Nyman U, Bjo¨rk J, Sterner G, Ba¨ck SE, Carlson J, Lindstro¨m V, Bakoush O, Grubb A: Standardization of p-creatinine assays and use of lean body mass allow improved prediction of calculated glomerular filtration rate in adults: a new equation. Scand J Clin Lab Invest 66: 451– 468, 2006 38. Bjo¨rk J, Ba¨ck SE, Sterner G, Carlson J, Lindstrom V, Bakoush O, Simonsson P, Grubb A, Nyman U: Prediction of relative glomerular filtration rate in adults: New improved equations based on Swedish Caucasians and standardized plasma-creatinine assays. Scand J Clin Lab Invest 67: 678 – 695, 2007 39. Macdonald JH, Marcora SM, Kumwenda MJ, Jibani M, Roberts G, Glover R, Barron J, Lemmey AB: The relationship between estimated glomerular filtration rate, demographic and anthropometric variables is mediated by muscle mass in nondiabetic patients with chronic kidney disease. Nephrol Dial Transplant 21: 3488 –3494, 2006 40. Macdonald JH, Marcora SM, Jibani M, Roberts G, Kumwenda MJ, Glover R, Barron J, Lemmey AB: Bioelectrical impedance can be used to predict muscle mass and hence improve estimation of glomerular filtration rate in non-diabetic patients with chronic kidney disease. Nephrol Dial Transplant 21: 3481–3487, 2006 41. Anastasio P, Spitali L, Frangiosa A, Molino D, Stellato D, Cirillo E, Pollastro RM, Capodicasa L, Sepe J, Federico P, Gaspare De Santo N: Glomerular filtration rate in severely overweight normotensive humans. Am J Kidney Dis 35: 1144 –1148, 2000 42. US Department of Health and Human Services: Centers for Disease Control and Prevention. State-specific prevalence of obesity among adults: United States, 2009. MMWR 59: 1–5,

43.

44.

45.

46. 47.

48. 49. 50. 51.

52. 53.

54. 55. 56. 57. 58.

59.

60.

61. 62.

2010. Available at: http://www.cdc.gov. Accessed August 6, 2010 National kidney foundation. Kidney learning solutions. GFR. Interpretation of GFR Estimates. Available at: http://www.kidney.org/professionals/KLS/gfr.cfm#38. Accessed August 5, 2010 Ma YC, Zuo L, Chen JH, Luo Q, Yu XQ, Li Y, Xu JS, Huang SM, Wang LN, Huang W, Wang M, Xu GB, Wang HY: Modified glomerular filtration rate estimating equation for Chinese patients with chronic kidney disease. J Am Soc Nephrol 17: 2937–2944, 2006 Matsuo S, Imai E, Horio M, Yasuda Y, Tomita K, Nitta K, Yamagata K, Tomino Y, Yokoyama H, Hishida A: Collaborators developing the Japanese equation for estimated GFR. Revised equations for estimated GFR from serum creatinine in Japan. Am J Kidney Dis 53: 982–992, 2009 Glassock RJ, Winearls C: Ageing and the glomerular filtration rate: Truths and consequences. Trans Am Clin Climatol Assoc 120: 419 – 428, 2009 Rule AD, Am H, Cornell LD, Taler SJ, Cosio FG, Kremers WK, Textor SC, Stegall MD: The association between age and nephrosclerosis on renal biopsy among healthy adults. Ann Intern Med 152: 561–567, 2010 Davies DF, Shock NW: Age changes in glomerular filtration rate, effective renal plasma flow, and tubular excretory capacity in adult males: J Clin Invest 129: 496 –507, 1950 Slack TK, Wilson DM: Normal renal function: CIN and CPAH in healthy donors before and after nephrectomy. Mayo Clin Proc 51: 296 –300, 1976 Sun X, Chen Y, Chen X, Wang J, Xi C, Lin S, Liu X: Change of glomerular filtration rate in healthy adults with aging. Nephrology 14: 506 –513, 2009 Gonwa TA, Atkins C, Zhang YA, Parker TF, Hunt JM, Lu CY, White MG: Glomerular filtration rates in persons evaluated as living-related donors: Are our standards too high? Transplantation 55: 983–985, 1993 Rule AD, Gussak HM, Pond GR, Bergstralh EJ, Stegall MD, Cosio FG, Larson TS: Measured and estimated GFR in healthy potential kidney donors. Am J Kidney Dis 43: 112–119, 2004 Poggio ED, Rule AD, Tanchanco R, Arrigain S, Butler RS, Srinivas T, Stephany BR, Meyer KH, Nurko S, Fatica RA, Shoskes DA, Krishnamurthi V, Goldfarb DA, Gill I, Schreiber MJ Jr: Demographic and clinical characteristics associated with glomerular filtration rates in living kidney donors. Kidney Int 75: 1079 –1087, 2009 Glassock RJ, Winearls C: An epidemic of chronic kidney disease: Fact or fiction? Nephrol Dial Transplant 23: 1117– 1121, 2008 Coresh J, Stevens LA, Levey AS: Chronic kidney disease is common: What do we do next? Nephrol Dial Transplant 23: 1122–1125, 2008 Glassock RJ, Winearls C: Screening for CKD with eGFR: Doubts and dangers. Clin J Am Soc Nephrol 3: 1563–1568, 2008 Melamed ML, Bauer C, Hostetter TH: eGFR: Is it ready for early identification of CKD? Clin J Am Soc Nephrol 3: 1569 – 1572, 2008 US Census Bureau. United States Census 2000: Profile of General Demographic Characteristics. 2008 Population Estimates: Age, sex, race, and Hispanic origin estimates in United States. Available at: http://factfinder. census.gov. Accessed August 6, 2010 Schaeffner ES, van der Giet M, Gaedeke J, To¨lle M, Ebert N, Kuhlmann MK, Martus P: The Berlin initiative study: The methodology of exploring kidney function in the elderly by combining a longitudinal and cross-sectional approach. Eur J Epidemiol 25: 203–210, 2010 National Kidney Disease Education Program: Laboratory Professionals: Reporting eGFR. Available at: http://www.nkdep.nih.gov/labprofessionals/reporting_eGFR. htm. Accessed August 5, 2010 Rainey PM: Automatic reporting of estimated glomerular filtration rate: Jumping the gun? Clin Chem 52: 2184 –2187, 2006 Levey AS, Stevens LA, Hostetter T: Automatic reporting of


63. 64.

65.

66. 67.

68.

69. 70.

71. 72.

73.

74. 75.

76.

77.

78.

79.

80.

81.

estimated glomerular filtration rate: Just what the doctor ordered. Clin Chem 52: 2188 –2193, 2006 Botev R, Mallie´ JP: Reporting the eGFR and its implication for CKD diagnosis. Clin J Am Soc Nephrol 3: 1606 –1607, 2008 Poggio ED, Rule AD: A critical evaluation of chronic kidney disease: Should isolated reduced estimated glomerular filtration rate be considered a ‘disease’? Nephrol Dial Transplant 24: 698 –700, 2009 Mathew TH, Johnson DW, Jones GR: Chronic kidney disease and automatic reporting of estimated glomerular filtration rate: Revised recommendations. Australasian Creatinine Consensus Working Group. Med J Aust 187: 459 – 463, 2007 den Hartog JR, Reese PP, Cizman B, Feldman HI: The costs and benefits of automatic estimated glomerular filtration rate reporting. Clin J Am Soc Nephrol 4: 419 – 427, 2009 Murthy K, Stevens LA, Stark PC, Levey AS: Variation in the serum creatinine assay calibration: A practical application to glomerular filtration rate estimation. Kidney Int 68: 1884 – 1887, 2005 Vickery S, Stevens PE, Dalton RN, van Lente F, Lamb EJ: Does the ID-MS traceable MDRD equation work and is it suitable for use with compensated Jaffe and enzymatic creatinine assays? Nephrol Dial Transplant 21: 2439 –2445, 2006 Levey AS, Greene T, Kusek JW, Beck GJ: A simplified equation to predict glomerular filtration rate from serum creatinine [Abstract]. J Am Soc Nephrol 11: 155A, 2000 Levey AS, Coresh J, Greene T, Stevens LA, Zhang YL, Hendriksen S, Kusek JW, Van Lente F: Using standardized serum creatinine values in the modification of diet in renal disease study equation for estimating glomerular filtration rate. Ann Intern Med 145: 247–254, 2006 Delanaye P, Cavalier E, Krzesinski JM: Estimated glomerular filtration rate. Ann Intern Med 146: 7, 2007 Froissart M, Rossert J, Jacquot C, Paillard M, Houillier P: Predictive performance of the modification of diet in renal disease and Cockcroft-Gault equations for estimating renal function. J Am Soc Nephrol 16: 763–773, 2005 Hallan S, Asberg A, Lindberg M, Johnsen H: Validation of the Modification of Diet in Renal Disease formula for estimating GFR with special emphasis on calibration of the serum creatinine assay. Am J Kidney Dis 44: 84 –93, 2004 Toffaletti J, Blosser N, Hall T, Smith S, Tompkins D: An automated dry-slide enzymatic method evaluated for measurement of creatinine in serum. Clin Chem 29: 684 – 687, 1983 Coresh J, Astor BC, McQuillan G, Kusek J, Greene T, Van Lente F, Levey AS: Calibration and random variation of the serum creatinine assay as critical elements of using equations to estimate glomerular filtration rate. Am J Kidney Dis 39: 920 –929, 2002 Van Biesen W, Vanholder R, Veys N, Verbeke F, Delanghe J, De Bacquer D, Lameire N: The importance of standardization of creatinine in the implementation of guidelines and recommendations for CKD: Implications for CKD management programmes. Nephrol Dial Transplant 21: 77– 83, 2006 Stevens LA, Manzi J, Levey AS, Chen J, Deysher AE, Greene T, Poggio ED, Schmid CH, Steffes MW, Zhang YL, Van Lente F, Coresh J: Impact of creatinine calibration on performance of GFR estimating equations in a pooled individual patient database. Am J Kidney Dis 50: 21–35, 2007 Levey AS, Greene T, Schluchter MD, Cleary PA, Teschan PE, Lorenz RA, Molitch ME, Mitch WE, Siebert C, Hall PM, Steffes MW: Glomerular filtration rate measurements in clinical trials. J Am Soc Nephrol 4: 1159 –1171, 1993 Lewis J, Agodoa L, Cheek D, Greene T, Middleton J, O’Connor D, Ojo A, Phillips R, Sika M, Wright J Jr: Comparison of cross-sectional renal function measurements in African Americans with hypertensive nephrosclerosis and of primary formulas to estimate glomerular filtration rate. Am J Kidney Dis 38: 744 –753, 2001 Gonwa TA, Jennings L, Mai ML, Stark PC, Levey AS, Klintmalm GB: Estimation of glomerular filtration rates before and after orthotopic liver transplantation: evaluation of current equations. Liver Transpl 10: 301–309, 2004 Ibrahim H, Mondress M, Tello A, Fan Y, Koopmeiners J, Thomas W: An alternative formula to the Cockcroft-Gault


82.

83.

84.

85.

86.

87. 88.

89.

90.

91.

92.

93.

94.

95.

96.

97.

98.

and the modification of diet in renal diseases formulas in predicting GFR in individuals with type 1 diabetes. J Am Soc Nephrol 16: 1051–1060, 2005 Verhave JC, Fesler P, Ribstein J, du Cailar G, Mimran A: Estimation of renal function in subjects with normal serum creatinine levels: influence of age and body mass index. Am J Kidney Dis 46: 233–241, 2005 Poggio ED, Wang X, Greene T, Van Lente F, Hall PM: Performance of the modification of diet in renal disease and Cockcroft-Gault equations in the estimation of GFR in health and in chronic kidney disease. J Am Soc Nephrol 16: 459 – 466, 2005 Bosma RJ, Doorenbos CR, Stegeman CA, van der Heide JJ, Navis G: Predictive performance of renal function equations in renal transplant recipients: An analysis of patient factors in bias. Am J Transplant 5: 2193–2203, 2005 Gera M, Slezak JM, Rule AD, Larson TS, Stegall MD, Cosio FG: Assessment of changes in kidney allograft function using creatinine-based estimates of glomerular filtration rate. Am J Transplant 7: 880 – 887, 2007 Saleem M, Florkowski CM, George PM, Woltersdorf WW: Comparison of two prediction equations with radionuclide glomerular filtration rate: Validation in routine use. Ann Clin Biochem 43: 309 –313, 2006 Srinivas S, Annigeri RA, Mani MK, Rao BS, Kowdle PC, Seshadri R: Estimation of glomerular filtration rate in South Asian healthy adult kidney donors. Nephrology 13: 440 – 446, 2008 Rule AD, Larson TS, Bergstralh EJ, Slezak JM, Jacobsen SJ, Cosio FG: Using serum creatinine to estimate glomerular filtration rate: Accuracy in good health and in chronic kidney disease. Ann Intern Med 141: 929 –937, 2004 Hojs R, Bevc S, Ekart R, Gorenjak M, Puklavec L: Serum cystatin C-based equation compared to serum creatinine-based equations for estimation of glomerular filtration rate in patients with chronic kidney disease. Clin Nephrol 70: 10 –17, 2008 Ma YC, Zuo L, Zhang CL, Wang M, Wang RF, Wang HY: Comparison of 99mTc-DTPA renal dynamic imaging with modified MDRD equation for glomerular filtration rate estimation in Chinese patients in different stages of chronic kidney disease. Nephrol Dial Transplant 22: 417– 423, 2007 Burke JT, Brault Y, Kahan BD, Hricik DE, Grinyo´ JM, Chapman JR, Polinsky M, Neylan JF: Accuracy and variability of equations to estimate glomerular filtration rates in renal transplant patients receiving sirolimus and/or calcineurin inhibitor immunosuppression. Transpl Int 21: 434 – 440, 2008 Mariat C, Alamartine E, Afiani A, Thibaudin L, Laurent B, Berthoux P, De Filippis JP, Thibaudin D, Mayor B, Elessawy AB, Berthoux F: Predicting glomerular filtration rate in kidney transplantation: Are the K/DOQI guidelines applicable? Am J Transplant 5: 2698 –2703, 2005 Issa N, Meyer KH, Arrigain S, Choure G, Fatica RA, Nurko S, Stephany BR, Poggio ED: Evaluation of creatinine-based estimates of glomerular filtration rate in a large cohort of living kidney donors. Transplantation 86: 223–230, 2008 Horio M, Imai E, Yasuda Y, Watanabe T, Matsuo S: Modification of the CKD epidemiology collaboration (CKD-EPI) equation for Japanese: Accuracy and use for population estimates. Am J Kidney Dis 56: 32–38, 2010 Tidman M, Sjo¨stro¨m P, Jones I: A Comparison of GFR estimating formulae based upon s-cystatin C and s-creatinine and a combination of the two. Nephrol Dial Transplant 23: 154 –160, 2008 White CA, Akbari A, Doucette S, Fergusson D, Knoll GA: Estimating glomerular filtration rate in kidney transplantation: Is the new chronic kidney disease epidemiology collaboration equation any better? Clin Chem 56: 474 – 477, 2010 Kukla A, El-Shahawi Y, Leister E, Kasiske B, Mauer M, Matas A, Ibrahim HN: GFR-estimating models in kidney transplant recipients on a steroid-free regimen. Nephrol Dial Transplant 25: 1653–1661, 2010 Tent H, Rook M, Stevens LA, Son WJ, Pelt LJ, Hofker HS, Ploeg RJ, Heide JJ, Navis G: Renal Function Equations before and after Living Kidney Donation: A Within-Individual Com-

950

99.

100. 101. 102. 103.

104.


parison of Performance at Different Levels of Renal Function. Clin J Am Soc Nephrol 5: 1960 –1968, 2010 Michels WM, Grootendorst DC, Verduijn M, Elliott EG, Dekker FW, Krediet RT: Performance of the Cockcroft-Gault, MDRD, and new CKD-EPI formulas in relation to GFR, age, and body size. Clin J Am Soc Nephrol 5: 1003–1009, 2010 Rule AD: Understanding estimated glomerular filtration rate: Implications for identifying chronic kidney disease. Curr Opin Nephrol Hypertens 16: 242–249, 2007 Stevens LA, Zhang Y, Schmid CH: Evaluating the performance of equations for estimating glomerular filtration rate. J Nephrol 21: 797– 807, 2008 Motulsky H: Multiple, logistic, and proportional hazards regression. In: Intuitive Biostatistics, New York, Oxford University Press, 2010, pp 296 –314 Chudleigh RA, Dunseath G, Evans W, Harvey JN, Evans P, Ollerton R, Owens DR: How reliable is estimation of glomerular filtration rate at diagnosis of type 2 diabetes? Diabetes Care 30: 300 –305, 2007 Rostoker G, Andrivet P, Pham I, Griuncelli M, Adnot S: A modified Cockcroft-Gault formula taking into account the

body surface area gives a more accurate estimation of the glomerular filtration rate. J Nephrol 20: 576 –585, 2007 105. Rigalleau V, Lasseur C, Raffaitin C, Perlemoine C, Barthe N, Chauveau P, Combe C, Gin H: The Mayo Clinic quadratic equation improves the prediction of glomerular filtration rate in diabetic subjects. Nephrol Dial Transplant 22: 813– 818, 2007 106. White C, Akbari A, Hussain N, Dinh L, Filler G, Lepage N, Knoll GA: Chronic kidney disease stage in renal transplantation classification using cystatin C and creatinine-based equations. Nephrol Dial Transplant 22: 3013–3020, 2007 107. Urbaniak J, Weyde W, Smolska D, Zagocka E, Klak R, Kusztal M, Krajewska M, Wozniak M, Klinger M: S-cystatin C formulae or combination of s-cystatin C and s-creatinine formulae do not improve prediction of GFR. Nephrol Dial Transplant 23: 2425–2426, 2008 Published online ahead of print. Publication date available at www.cjasn.org.