Reporting Estimated GFR: A Laboratory Perspective

The National Kidney Disease Education Program (NKDEP) was initiated by the National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, along with development of the Kidney Disease Outcomes Quality Initiative (KDOQI) clinical practice guidelines.1, 2 KDOQI and NKDEP recommend use of estimated glomerular filtration rate (eGFR) in addition to serum creatinine to assess kidney function.3 An eGFR value is more easily related to a patient's kidney condition than is serum creatinine alone. Indeed, a large proportion of patients with treatable kidney disease were not being identified based on creatinine results.4 Estimated GFR has been included in clinical practice guidelines to allow physicians to identify patients with chronic kidney disease (CKD) to enable earlier intervention to slow progression of the disease.

The NKDEP began in 2002 to educate physicians, high-risk patient groups, and clinical laboratories about the value of using eGFR to identify patients with CKD. In this issue of the American Journal of Kidney Diseases, Accetta et al describe the prevalence of reporting eGFR along with serum creatinine among clinical laboratories in the United States in late 2006 to early 2007.5 Their survey was conducted approximately 4 years after publication of the KDOQI guidelines and within 1 year of publication of guidelines for clinical laboratories on how to calculate and report eGFR.6 They found that eGFR was reported along with serum creatinine by 26% of physician offices, 44% of hospitals, 39% of independent laboratories, 47% of community clinics, 46% of health fair/insurance/public health settings, and 48% of other laboratories. This study also examined which estimating equation was used, whether eGFR was reported with all or selected serum creatinine results, and if laboratories not reporting eGFR were planning to do so. Figure 1 shows that reporting frequency for eGFR has continued to increase, as 70% of participants reported eGFR along with serum creatinine in the June 2008 Chemistry Survey of the College of American Pathologists (CAP), which represents predominantly larger hospitals and independent laboratories.7 Figure 2 shows that 77% of those laboratories reported eGFR with all creatinine results.

View Large Image Download to PowerPoint

Figure 1. Trend in reporting estimated glomerular filtration rate. Reprinted with permission from7.

View Large Image Download to PowerPoint

Figure 2. Reporting practices for estimated glomerular filtration rate. Reprinted with permission from7.

The NKDEP established a Laboratory Working Group (LWG) to address issues associated with laboratory measurements used in kidney disease. The first laboratory challenge in reporting eGFR was that serum creatinine results were not standardized among different measurement procedures.8 Consequently, there would be less accuracy (more variability) in the eGFR among laboratories than was observed in the clinical investigations used to develop the recommended estimating equation from the Modification of Diet in Renal Disease (MDRD) Study.

This problem was addressed by developing the NKDEP creatinine standardization program to enable the manufacturers of laboratory methods to establish calibration traceability to an isotope-dilution mass spectrometry (IDMS) reference measurement procedure that was calibrated using a pure crystalline creatinine primary standard from the National Institute for Standards and Technology (NIST). Such a primary standard is not suitable for use with routine methods that directly assay a blood, serum, or plasma sample. The LWG collaborated with NIST to develop a serum-based secondary reference material (SRM 967) that is suitable for use with routine methods to establish calibration traceability to an IDMS reference method. SRM 967 became available from the NIST in early 2007. The LWG also collaborated with CAP to make available an accuracy-based proficiency testing program (designated LN 24) that is suitable for surveillance of the accuracy of routine methods versus an IDMS reference measurement procedure at the NIST.

The manufacturers of laboratory methods are currently introducing revisions for both alkaline picrate (Jaffé) and enzymatic methods to make their calibrations traceable to an IDMS reference. The LWG anticipates that most whole blood, serum, and plasma creatinine methods will have calibration traceable to IDMS in 2009.

The original MDRD Study GFR estimating equation was based on creatinine results from a central laboratory using a routine creatinine method that, like all routine methods at that time, was not standardized to an IDMS reference measurement procedure. Following the initiative to standardize creatinine methods, the investigators established correction factors for the original MDRD Study data and published a revised estimating equation that was suitable for use with methods with calibration traceable to IDMS.9 At the present time, while methods are transitioning to have calibration traceable to IDMS, it is very important to use either the original or the IDMS-traceable equation as appropriate.

One potentially significant impact of standardized creatinine results is on assessment of kidney function for drug dose adjustment. Standardizing creatinine calibration to an IDMS reference produces a lowering of creatinine values by 10% to 20% for most routine methods. Consequently, physicians need to be aware that interpretation of serum creatinine related to kidney function is altered. Pharmaceutical manufacturers have used the Cockcroft-Gault equation to estimate creatinine clearance as the basis for drug dose adjustment recommendations, and there is no modified equation available for use with IDMS-traceable creatinine results. Consequently, creatinine clearance estimated from the Cockcroft-Gault equation will be erroneously high and potential kidney impairment will be underestimated. This effect will be more important for drugs with narrow therapeutic or toxic ranges. The MDRD Study eGFR equation will give a more appropriate estimate of kidney function, but the problem arises because the drug labeling is based on estimates derived from the Cockcroft-Gault equation. Physicians and pharmacists need to be aware of this situation to avoid inappropriate decisions on drug dosages. The NKDEP, US Federal Drug Administration, other professional kidney societies, and pharmaceutical companies need to collaborate to develop an appropriate contemporary practice for estimating kidney function for the purpose of drug dose recommendations.

Another significant impact of standardized creatinine results is on assessment of kidney function in infants and children. Analogous to the situation with the original MDRD Study equation, the original Schwartz equation is not appropriate for use with methods that have calibration traceable to IDMS because the lower creatinine values will cause the eGFR to be as much as 20% to 40% too high. A modified Schwartz equation is being developed and will be published in the near future.

The LWG is addressing issues related to the specificity of creatinine methods. For samples without interfering substances, the alkaline picrate and enzymatic methods will give comparable creatinine and eGFR values when each method has its calibration traceable to IDMS. However, it is well known that alkaline picrate methods are subject to interference from a number of drugs and endogenous substances.6, 10, 11 In particular, serum proteins cause an increase in apparent creatinine concentration that is typically corrected for an average serum protein concentration and/or minimized by measurement techniques. However, for patients with abnormal protein concentrations, an erroneous creatinine result may be produced. The protein effect can be quite significant for children, who typically have both lower serum protein and creatinine concentrations than adults, causing a proportionally greater positive error in the eGFR. Enzymatic creatinine methods are not affected by protein concentrations, have fewer drug and endogenous substance interferences, but are still subject to interferences notably from bilirubin and hemoglobin.6, 11, 12, 13 A systematic investigation of interfering substances with the current generation of routine methods has been initiated to support development of recommendations for specificity requirements.

Detection of albuminuria, a marker of kidney damage, also allows earlier diagnosis of CKD. The LWG in collaboration with the International Federation of Clinical Chemistry and Laboratory Medicine (IFCC) has begun to address standardization of urine albumin measurement. A conference was held in 2007 to assess the current state of pre-analytical, analytical, and postanalytical issues affecting urine albumin measurements. Although both urine albumin and creatinine results can be affected by patient preparation (posture, exercise, meals) and time of day of sample collection, neither is standardized and clinical guidelines vary in recommendations. Differences in results among methods have been reported for both albumin and creatinine, but accuracy is unknown because there are no reference methods for albumin in serum or urine and no reference materials for albumin or creatinine in urine. The recommended decision intervals for the albumin-creatinine ratio do not take into account the intergroup differences in albumin or creatinine excretion related to age, sex, and ethnicity. The relationship of albuminuria to kidney damage appears to be a continuous function, which is not addressed by the current thresholds for interpretation of albuminuria (“microalbuminuria” and “macroalbuminuria”). Clinical needs have been identified for standardization of urine collection methods, urine albumin and creatinine measurements, reporting of test results, and reference intervals for the albumin-creatinine ratio. A report from this conference will be published in the near future, and working groups are being formed to address the issues raised.

The Laboratory Professionals section of the NKDEP web site provides regular updates on the status of laboratory issues related to kidney disease.14