The incidence of transplantation from living donors in

Transcription

1 Imprecision of Creatinine-Based GFR Estimates in Uninephric Kidney Donors Jane C. Tan,* Bing Ho, Stephan Busque, Kristina Blouch,* Geraldine Derby,* Bradley Efron, and Bryan D. Myers* *Division of Nephrology, and Departments of Surgery and Statistics, Stanford University School of Medicine, Stanford, California; and Division of Nephrology, Northwestern University, Chicago, Illinois Background and objectives: To ensure long-term safety of living kidney donors, it is now recommended that they be followed for at least 2 years after donation and that serum creatinine levels be monitored. Such levels are often subjected by clinical laboratories to estimating equations and are reported as estimated GFR (egfr). The accuracy of such equations in uninephric living donors has yet to be validated. This is especially important in older living donors, who often have senescence-related depression of GFR. Design, setting, participants, & measurements: We compared urinary creatinine clearance, four-variable Modification of Diet in Renal Disease estimating equation (egfr), and the recently reported CKD-EPI GFR estimating equation with true GFR measured by the urinary iothalamate clearance (igfr) in 64 subjects after kidney donation. Results: Creatinine clearance overestimated igfr. Both creatinine-based estimating equations were poorly correlated with and underestimated igfr. More than half of kidney donors had egfr <60 ml/min per 1.73 m 2 after donation, a level that categorized them as having stage 3 chronic kidney disease by our current laboratory reporting, whereas only 25% had igfr <60 ml/min per 1.73 m 2. This misclassification disproportionately affected older donors age >55 years, of whom 80% had egfr <60 ml/min per 1.73 m 2. Neither significant albuminuria nor hypertension was observed. Conclusions: The current practice of reporting egfr after donation commonly leads to a misclassification of chronic kidney disease, particularly in older donors. To ensure long-term well-being of living kidney donors, more precise estimates of GFR are required, particularly among older potential donors. Clin J Am Soc Nephrol 5: , doi: /CJN The incidence of transplantation from living donors in the United States has nearly doubled over the past decade (1). Since 2001, more than 6000 individuals have made living kidney donations annually, a number that now exceeds that for deceased kidney donors (1). The willingness of transplant teams to accept living kidney donation is largely on the basis of evidence suggesting a safe long-term outcome for the donor (2 4). To ensure such long-term safety, extensive evaluation of donors is routinely performed (5). Such evaluation includes an estimation of GFR, the best overall measure of the functional performance of the kidney. GFR, measured by the urinary clearance of a true filtration marker (6), is cumbersome and costly and is not available at most transplant centers. Instead, the clearance of endogenous creatinine is typically used as a surrogate measure of GFR. Most centers have arbitrarily selected 80 ml/min per 1.73 m 2 as a threshold for GFR, below which living donation is not permitted (7). Observations from our own laboratory reveal that for urinary inulin clearance, 80 ml/min per 1.73 m 2 represents the lower end of the normal range in a large population of Received July 29, Accepted December 6, Published online ahead of print. Publication date available at Correspondence: Dr. Jane Tan, 750 Welch Road, Suite 200, Palo Alto, CA Phone: ; Fax: ; Jane.tan@stanford.edu healthy individuals younger than 55 years (8). Because creatinine is secreted by the proximal tubule, as well as freely filtered, the creatinine clearance (C creat ) overestimates the GFR and, assuming a lower limit of 80 ml/min per 1.73 m 2 for this surrogate filtration marker, could well lead to acceptance of kidney donors whose actual GFR is below the normal range. The long-term safety of donation under these circumstances remains to be demonstrated. Equations derived from the Modification of Diet in Renal Disease (MDRD) study have been widely applied to estimate GFR (egfr) from serum creatinine and have largely supplanted the urinary C creat as a method of GFR estimation in practice. Although reporting of egfr may aid in the early detection of chronic kidney disease (CKD) in persons with elevated serum creatinine concentrations, the application of egfr to healthy kidney donors may be misleading. Recent studies reported egfr in a healthy binephric population to be inaccurate (9,10). Moreover, the precision of these equations has not been examined in uninephric kidney donors. In an attempt to refine estimates of GFR, particularly among persons with normal or near normal GFR, a new equation was recently developed (the CKD-EPI equation) and is expected to replace the MDRD equation for routine clinical laboratory reporting of egfr (11). The accuracy of this equation in the living donor population, especially in the elderly, has not been tested. Copyright 2010 by the American Society of Nephrology ISSN: /

2 498 Clinical Journal of the American Society of Nephrology Clin J Am Soc Nephrol 5: , 2010 To ensure the safety of the growing living donor population, United Network for Organ Sharing now requires follow-up and monitoring of the serum creatinine in living kidney donors for at least 2 years after donation (1). Such follow-up of living donors combined with often routine reporting of egfr by clinical laboratories warrants a careful comparison of such estimating equations against GFR, as determined by clearance of a true filtration marker. In this study, we examined the accuracy of C creat and the most commonly used abbreviated fourvariable MDRD, as well as the new CKD-EPI estimating equation in predicting in living donors after kidney donation the actual GFR, determined by the urinary clearance of the filtration marker, iothalamate (igfr). Additional analysis compared these results in younger versus older subjects. Materials and Methods Patient Population All protocols were approved by the Institutional Review Board of Stanford University School of Medicine. The study population was composed of 64 subjects who had previously undergone uninephrectomy for living kidney donation from 5 to 86 months before this study. The median time from donation was 13 months. A comparison group consisted of a population of 275 patients with various stages of glomerular disease who had undergone inulin clearance as part of a series of studies for progression of glomerular disease. Findings from this group were used to characterize the relationship between serum creatinine and igfr over the entire range of GFR. Measurement of GFR All studies were performed in the General Clinical Research Center at Stanford University. GFR was estimated by the urinary iothalamate clearance as described in detail elsewhere (12). Cold iothalamate was given by constant infusion during a state of water diuresis. Four 20- minute urine collections and bracketing plasma were analyzed to determine the igfr from the urinary clearance of iothalamate. The urinary clearance of endogenous creatinine (C creat ) and the urinary albumin/creatinine ratio were determined simultaneously. Mean arterial pressure (MAP) was also determined at the commencement of the clearance study. An Array 360 System (Beckman Instruments Inc., Brea, CA) was used to measure urinary albumin concentration by rate nephelometry (13). Serum and urinary creatinine levels were determined with a Beckman creatinine analyzer. A second determination of serum creatinine concentration was performed on each patient by the Stanford Hospital clinical laboratory using a similar kinetic assay to minimize noncreatinine chromogen and employing a Siemens RxL kinetic Jaffe assay. C creat was computed by multiplying the fractional clearance of creatinine by the simultaneous iothalamate clearance (6). The abbreviated four-variable MDRD egfr was calculated as follows (14): egfr/1.73 m SCr Age (if female) 1.21 (if black) where SCr is the serum concentration performed by the Stanford clinical laboratory. The CKD-EPI equation was calculated according to the equations published by Levey et al. (11) Statistical Analyses All data are expressed as mean SD. Linear regression analysis was used to determine the relationship between C creat or egfr and igfr. Standard assumptions for linear regression were met. The concordance coefficient was calculated to assess the degree of correlation between the various estimating equations and igfr. Bias and the interquartile ranges were calculated. The precision of the estimating equations was expressed as root mean square error, and accuracy was expressed as % estimating equation within 10% and 30% of igfr. Residual plots were determined (15,16). Nonlinear regression analysis was used to characterize the relationship between serum creatinine levels and igfr. Results Sixty-four kidney donors were studied a median time of 13 months after transplantation, with a range of 5 to 86 months. Their age at time of study ranged from 21 to70 years (median, 49 years). Forty-seven percent were women; 6% were African American, 11% Hispanic, and 19% Asian American. Age at donation ranged from 18 to 67 years (median, 48 years). At screening, all subjects had serum creatinine levels within the normal range and C creat 80 ml/min per 1.73 m 2. The exclusion criteria employed before transplantation include a history of and/or treatment for hypertension, abnormal glucose tolerance test, body mass index 30, and proteinuria. GFR in Postnephrectomy Donors C creat substantially overestimated igfr: versus ml/min per 1.73 m 2 (P ), as shown in Figure 1A and Table 1. Whereas the correlation was reasonably strong (R ; W 0.895), there was a positive bias of 19 ml/min per 1.73 m 2 (Figure 1B). The interquartile range (IQR) was 11 to 26 ml/min per 1.73 m 2. The accuracy of C creat, as judged by being within 10% of igfr, was only 11% (Table 1). C creat failed to identify 13 of 16 patients whose igfr fell below a threshold of 60 ml/min per 1.73 m 2, which has been used to classify stage 3 CKD (Figure 1A) (17). The egfr underestimated igfr in uninephric donors: versus ml/min per 1.73 m 2 (P ). As shown in Figure 2A, the correlation between the two values was relatively weak (R ; W 0.830). As shown in Figure 2B and Table 1, a negative bias was 10 ml/min with an IQR of 18 to 1 ml/min per 1.73 m 2. The egfr was within 10% of igfr in only 25% of subjects (Table 1). More than half of these subjects (35 of 64) had an egfr 60 ml/min per 1.73 m 2, although only 25% (16 of 64) had igfr 60 ml/min per 1.73 m 2 (Figure 2A). Similar to the MDRD egfr, the CKD-EPI equation also underestimated igfr. Again, the correlation between the two values was relatively weak (R ; W 0.836), although slightly better than the MDRD egfr (Figure 3A). A negative bias was 9 ml/min per 1.73 m 2 with an IQR of 16 to 1 ml/min per 1.73 m 2 (Table 1). Once again, only 25% of subjects were within 10% of igfr. More than half of the subjects (33 of 64) had an egfr 60 ml/min per 1.73 m 2 versus only 16 of 64 with igfr 60 ml/min per 1.73 m 2 (Figure 3A). Two subjects had borderline microalbuminuria, 32 and 38 mg/g creatinine, despite igfr above 60 ml/min per 1.73 per m 2 : 82 and 62 ml/min per 1.73 m 2, respectively. Each of the

3 Clin J Am Soc Nephrol 5: , 2010 GFR in Living Kidney Donors 499 Figure 1. (A) Comparison of igfr versus C creat in uninephric subjects after kidney donation. The solid diagonal line represents the identity line. The dashed lines represent 60 ml/min per 1.73 m 2, a value under which a subject would be considered to have chronic kidney disease. W, concordance coefficient. (B) Residual plot demonstrating the bias and error of C creat. The solid line represents the regression line. Table 1. Bias, accuracy, and precision of estimating equations Median Bias IQR Within 10% of igfr, % (n) Within 30% of igfr, % (n) RMSE C creat to (7) 53 (34) 10.1 MDRD egfr to 1 25 (16) 84 (54) 12.8 CKD-EPI egfr 9 16 to 1 25 (16) 84 (54) 12.4 IQR, interquartile range; RMSE, root mean square error. Figure 2. (A) Comparison of igfr versus MDRD egfr. The dashed lines represent 60 ml/min per 1.73 m 2, a value under which a subject would be considered to have chronic kidney disease. (B) Residual plot demonstrating the bias and error of MDRD egfr. The solid line represents the regression line. remaining 62 subjects had negligible or unmeasurable albuminuria ( 30 mg/g creatinine). None of the 64 subjects was hypertensive, with mean MAP of mmhg. Older Donors and egfr On average, igfr was significantly lower in older ( 55 years) than younger ( 55 years) subjects, versus ml/min per 1.73 m 2, respectively (P 0.001). The igfr was 60 ml/min per 1.73 m 2 in 52% of older donors versus only 8% of younger donors (Figures 1 through 3 and Table 2), suggesting a substantial incidence of CKD in the former, using the MDRD definition. Among older donors, 80% had egfr 60 ml/min per 1.73 m 2, compared with only 38% of the younger group (Table 2). By our current clinical labora-

4 500 Clinical Journal of the American Society of Nephrology Clin J Am Soc Nephrol 5: , 2010 Figure 3. (A) Comparison of igfr versus CKD-EPI egfr. The dashed lines represent 60 ml/min per 1.73 m 2, a value under which a subject would be considered to have chronic kidney disease. (B) Residual plot demonstrating the bias and error of CKD-EPI egfr. The solid line represents the regression line. Table 2. Prediction of CKD in older and younger donors Uninephric subjects with GFR 60 ml/min per 1.73 m 2 Method of measurement Age 55 yr, % Age 55 yr, % C creat 3 12 MDRD egfr CKD-EPI egfr igfr 8 52 C creat, creatinine clearance; igfr, iothalamate GFR. tory reporting practices, these patients are routinely reported as having CKD stage 3. None of the subjects had egfr 30 ml/min per 1.73 m 2. If the CKD-EPI egfr was applied, 76% of the older living donors and 36% of younger donors would be reported as having CKD stage 3 (Table 2). Serum Creatinine versus GFR in Kidney Donors Relative to Patients with CKD In recent years, we have measured GFR using a true filtration marker (inulin or iothalamate) in 275 patients with chronic glomerular disease. We have used a smoothing spline regression technique to fit the individual points for serum creatinine versus igfr. The fitted regression line (Figure 4, solid line) confirms the well-known hyperbolic relationship between these two quantities when the data points (gray x symbols) span the entire biologic range of GFR. In Figure 4, we have superimposed the corresponding data points after transplantation for our uninephric donors (black circles). Over the moderately depressed igfr range in uninephric posttransplantation donors, the inverse relationship is only minimal, with a slope , an intercept of 1.64, and a relatively weak correlation with R The absence of a substantial inverse relationship between serum creatinine and igfr in the uninephric Figure 4. Serum creatinine versus igfr in subjects with chronic glomerular disease (gray x ) and living donors after donor nephrectomy (black circles). living donor population explains the poor correlation between egfr and igfr in this setting. Discussion Living donor kidney transplantation now comprises 40% of all kidney transplantations. Studies of long-term donor safety have concluded that carefully screened living donors are not at increased risk of mortality or significant morbidity (2,3,18). However, the demography of living donors has been steadily changing, with an increasing trend toward the use of older and medically complex living donors (19,20). For example, the number of living donors age 50 years or older almost doubled from 853 in 1998 to 1561 in 2008, corresponding to an increase from 19% to 25% of all living donors (1). Recognizing the increased acceptance of and willingness for living donation by the general population, an effort is being made by the transplant community to better define long-term health outcomes of living donation. An example of this is the recent United Network for Organ Sharing guideline to follow-up and monitor serum creatinine

5 Clin J Am Soc Nephrol 5: , 2010 GFR in Living Kidney Donors 501 levels of living donors in their transplant center for at least two years after donation (21). The convention in most transplant programs is to select living kidney donors on the basis of a value for GFR that is above 80 ml/min per 1.73 m 2 (5), and reports of long-term safety of living donors have been on the basis of populations screened by this traditional guideline. We have determined GFR by inulin clearance in healthy individuals. Among those younger than 55 years (n 133), a value of 80 ml/min per 1.73 m 2 represents the sixth percentile (8). Among healthy individuals age 55 years (n 29), however, 80 ml/min per 1.73 m 2 represents the 49th percentile. Thus, fully one half of potential older donors are predicted to have a GFR below the conventional threshold for donor acceptance, a phenomenon that is not recognized because of an overestimation of GFR by C creat, which is used as a surrogate in most transplant centers. The significantly lowered GFR evident in older potential donors reflects the phenomenon of renal senescence, in which GFR depression is associated with a loss of filtration capacity attributable to glomerulopenia (8). We submit that the posttransplantation reduction of igfr in older donors in this study also likely reflects renal senescence. A substantial impediment to using igfr to identify the development of CKD after living kidney donation is that the performance of urinary clearance of a filtration marker is timeconsuming and not widely available. As discussed above, creatinine secretion leads to a systematic overestimate of igfr by C creat (6) and could obscure the detection of subjects whose igfr may fall to values in the CKD range after transplantation. On the other hand, using egfr, either by MDRD or CKD-EPI equations, does not provide a reasonable solution to this dilemma. Underestimation of the igfr by egfr during such follow-up of subjects after donation has led to increased categorization by many clinical laboratories of donors with values 60 ml/min per 1.73 m 2 as having stage 3 CKD (17,22). This study reveals that because of adaptive hyperfiltration, igfr is actually 60 ml/min per 1.73 m 2 after donation in most patients 55 years of age (Figures 2A and 3A), notwithstanding the finding that the vast majority (80%; Table 2) have an egfr that labels them as having CKD. Among those older posttransplantation donors with igfr 50 to 60 ml/min per 1.73 m 2, the absence of albuminuria and hypertension suggests that a GFR within this range is unlikely to indicate clinically meaningful CKD. Adaptive hyperfiltration after uninephrectomy has been shown to elevate igfr in the remaining kidney by approximately 33% (23 25). Thus, an older donor with a pretransplantation GFR of 80 to 90 ml/min per 1.73 m 2 (i.e., 40 to 45 ml/min per 1.73 m 2 per kidney) would be expected to increase GFR in the remaining kidney to a postdonation value of 50 to 60 ml/min per 1.73 m 2. We suspect that this range of igfr is likely to represent an adequate compensatory response to uninephrectomy and to portend a satisfactory long-term outcome in an older donor (26). There has been increasing debate over the validity of the widespread use of egfr in the general population (26 31). A particular concern is that egfr 60 ml/min per 1.73 m 2 after donation could mislabel perfectly healthy uninephric individuals, converting them to the worried-well (27). One possible source of the imprecision of the egfr in this study is our failure to calibrate the assay of serum creatinine to that of the Cleveland Clinic reference laboratory (32 36). The majority of the samples used in this study were obtained before the calibration of assays across the United States to the sample provided by the National Institute of Standards and Technology (NIST). Another and, we suspect, more important cause of imprecision is the relationship between serum creatinine levels and GFR in those with normal or near normal renal function. The fitted regression line of GFR versus serum creatinine in subjects with chronic glomerular disease (Figure 4, solid line) confirms the well-known hyperbolic relationship between these two quantities when the data points (gray x symbols) span the entire biologic range of GFR. Superimposition of corresponding data points for follow-up igfr after transplantation in our uninephric donors (black circles) demonstrates that over the observed GFR range there is a minimal inverse relationship between serum creatinine and GFR. The failure of the MDRD and other estimating equations to reliably predict GFR from serum creatinine in apparently healthy transplant donors is therefore not surprising. This is in contrast to GFR predictions by MDRD and other estimating equations in persons with moderate or advanced CKD, the population from which the MDRD equation was derived (igfr 25 to 55 ml/min per 1.73 m 2 ). Inspection of Figure 4 reveals that over the MDRD study range of igfr, the serum creatinine-to-gfr relationship occupies the steep part of the hyperbolic curve. Conclusions We submit that egfr cannot be expected to accurately predict igfr when the latter is in the upper part of the biologic range, and that the practice of predicting GFR from egfr in living donors should be abandoned (26). Acknowledgments This study was supported by a grant from the John M. and Abby Sobrato Foundation, and National Institutes of Health grants R01 DK and M01 RR Disclosures None. References 1. United Network for Organ Sharing: U.S. Transplantation Data. Available at: Accessed July 29, Ibrahim HN, Roley R, Tan L, Rogers T, Bailey RF, Guo H, Gross CR, Matas AJ: Long-term consequences of kidney donation. N Engl J Med 360: , Najarian JS, Chavers BM, McHugh LE, Matas AJ: 20 years or more of follow-up of living kidney donors. Lancet 340: , Narkun-Burgess DM, Nolan CR, Norman JE, Page WF, Miller PL, Meyer TW: Forty-five year follow-up after uninephrectomy. Kidney Int 43: , Davis CL, Delmonico FL: Living-donor kidney transplantation: A review of the current practices for the live donor. J Am Soc Nephrol 16: , Shemesh O, Golbetz H, Kriss JP, Myers BD: Limitations of

6 502 Clinical Journal of the American Society of Nephrology Clin J Am Soc Nephrol 5: , 2010 creatinine as a filtration marker in glomerulopathic patients. Kidney Int 28: , Kasiske BL, Ravenscraft M, Ramos EL, Gaston RS, Bia MJ, Danovitch GM: The evaluation of living renal transplant donors: Clinical practice guidelines. Ad Hoc Clinical Practice Guidelines Subcommittee of the Patient Care and Education Committee of the American Society of Transplant Physicians. J Am Soc Nephrol 7: , Hoang K, Tan JC, Derby G, Blouch KL, Masek M, Ma I, Lemley KV, Myers BD: Determinants of glomerular hypofiltration in aging humans. Kidney Int 64: , 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: , 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: , 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: , Alejandro V, Scandling JD, Jr., Sibley RK, Dafoe D, Alfrey E, Deen W, Myers BD: Mechanisms of filtration failure during postischemic injury of the human kidney. A study of the reperfused renal allograft. J Clin Invest 95: , Nakamura Y, Myers BD: Charge selectivity of proteinuria in diabetic glomerulopathy. Diabetes 37: , Miller WG: Reporting estimated GFR: A laboratory perspective. Am J Kidney Dis 52: , Rule AD: Understanding estimated glomerular filtration rate: implications for identifying chronic kidney disease. Curr Opin Nephrol Hypertens 16: , Stevens LA, Zhang Y, Schmid CH: Evaluating the performance of equations for estimating glomerular filtration rate. J Nephrol 21: , Levey AS, Coresh J, Balk E, Kausz AT, Levin A, Steffes MW, Hogg RJ, Perrone RD, Lau J, Eknoyan G: National Kidney Foundation practice guidelines for chronic kidney disease: Evaluation, classification, and stratification. Ann Intern Med 139: , Matas AJ, Bartlett ST, Leichtman AB, Delmonico FL: Morbidity and mortality after living kidney donation, : Survey of United States transplant centers. Am J Transplant 3: , Reese PP, Feldman HI, McBride MA, Anderson K, Asch DA, Bloom RD: Substantial variation in the acceptance of medically complex live kidney donors across US renal transplant centers. Am J Transplant 8: , Rook M, Bosma RJ, van Son WJ, Hofker HS, van der Heide JJ, ter Wee PM, Ploeg RJ, Navis GJ: Nephrectomy elicits impact of age and BMI on renal hemodynamics: Lower postdonation reserve capacity in older or overweight kidney donors. Am J Transplant 8: , United Network for Organ Sharing: Guidelines for the medical evaluation of living kidney donors (Living Donor Committee). Available at: pubcommentpropsub_208.pdf. Accessed July 29, K/DOQI clinical practice guidelines for chronic kidney disease: Evaluation, classification, and stratification. Am J Kidney Dis 39: S1 S266, Ogden DA: Donor and recipient function 2 to 4 years after renal homotransplantation. A paired study of 28 cases. Ann Intern Med 67: , Flanigan WJ, Burns RO, Takacs FJ, Merrill JP: Serial studies of glomerular filtration rate and renal plasma flow in kidney transplant donors, identical twins, and allograft recipients. Am J Surg 116: , Ewald J, Aurell M: Renal function studies after donor nephrectomy. Scand J Urol Nephrol 6 7: , Glassock RJ, Winearls CG: Routine reporting of estimated glomerular filtration rate: Not ready for prime time. Nat Clin Pract Nephrol 4: , Cheung CK, Bhandari S: Perspectives on egfr reporting from the interface between primary and secondary care. Clin J Am Soc Nephrol 4: , 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: , Bennett WM: Reporting egfr. Clin J Am Soc Nephrol 3: , Glassock RJ, Winearls C: Screening for CKD with egfr: Doubts and dangers. Clin J Am Soc Nephrol 3: , Melamed ML, Bauer C, Hostetter TH: egfr: Is it ready for early identification of CKD? Clin J Am Soc Nephrol 3: , Clase C: Reply from the authors: Estimating the prevalence of low glomerular filtration rate requires attention to the creatinine calibration assay. J Am Soc Nephrol 13: , Clase CM, Garg AX, Kiberd BA: Prevalence of low glomerular filtration rate in nondiabetic Americans: Third National Health and Nutrition Examination Survey (NHANES III). J Am Soc Nephrol 13: , 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: , 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: , 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 Access to UpToDate on-line is available for additional clinical information at