Participants: Unfortunately the list of participants created during the meeting was not retrieved at the end. Consequently, we apologize if anyone is omitted from this list. If a name was omitted please notify Greg Miller and we will add it to the list.
Neil Greenberg, Chair WG_GFRA, Ortho Clinical Diagnostics, USA Greg Miller, Chair WG-SAU and LWG of NKDEP, Virginia Commonwealth University, USA Bob Rej, New York State Department of Health, USA David Bruns, University of Virginia, USA David Bunk, NIST, USA David Seccombe, University of British Columbia, Canada Gary Myers, AACC, USA Graham Jones, St Vincent’s Hospital and University of New South Wales, Australia Gunner Nordin, Equalis, Sweden Ingrid Zegers, IRMM, Belgium Jim Fleming, LabCorp, USA John Eckfeldt, University of Minnesota, USA Joris Delanghe, Ghent University, Belgium Maria Stella Graziani, Ospedale Civile Maggiore, Italy Mauro Panteghini, University of Milan, Italy Stella Raymondo, Uruguay Sverre Sandberg, Haukeland University Hospital, Norway Yoshi Itoh, Asahikawa Medical University, Japan
Next meeting: joint meeting of WG-GFRA, WG-SAU, and LWG of NKDEP on Thursday July 28, 2011, 8-12:30 at the Marriott Marquis Hotel Room A706, Atlanta, GA, USA (in conjunction with the AACC Annual Meeting).
1. Results from the creatinine specificity study were reviewed. Briefly, residual serum or plasma samples were obtained from a non-diseased control group and 19 other groups of patients with various clinical conditions likely to include substances reported in the past to influence creatinine results by either enzymatic or Jaffe methods. In addition 4 volatile substances were spiked into 2 concentrations of serum creatinine. All samples were measured by 3 Jaffe and 4 enzymatic commercial measurement procedures, and by an ID-LC-MS/MS creatinine measurement procedure. All procedures were shown to be in good agreement with the assigned values for NIST SRM 967. In addition the ID-LC-MS/MS procedure was in good agreement with reference materials used in the IFCC round robin exercise for reference measurement procedures. Overall, the influence of interfering substances on bias for Jaffe methods was greater than on enzymatic methods. However, the results showed that all procedures, whether enzymatic or Jaffe, were affected by some of the interfering substances, and that there were substantial differences in magnitude of bias among different commercial implementations of both enzymatic and Jaffe methods.
2. There was discussion about the suitability of the concentration intervals represented by the patient samples used; particularly for pediatric
3. It was suggested to investigate correlation between the concentration of the analyte parameter used to select patient samples and the observed bias for creatinine measurement. In some cases, e.g. hemolysis or bilirubin, this approach could be investigated but will have a limitation in interpretation because the actual number and concentration of all potentially interfering substances present in a sample is unknown.
4. It was suggested to provide the raw data in a supplemental file so any follow up data analysis could be undertaken by interested persons.
5. It was suggested that a separate investigation of performance at low creatinine concentrations more representative of pediatric samples would be useful since the influence of protein and other potential interfering substances is likely different at low concentrations.
6. The performance requirements for serum creatinine published in 2006 may need to be revisited. Those requirements were based on not influencing eGFR by MDRD equation more than 10% mean square root error. Current interpretation based on relative change value to follow progression may require more stringent performance. The influence of imprecision may need to be treated as a separate parameter rather than only provide a total error goal.
7. Recommendations for creatinine quality specifications associated with interfering substances were not agreed but need to be developed. It may be of interest to consider the population served by a laboratory (e.g. pediatric vs. adult) in determining the quality requirements.
1. Preliminary results from the urine albumin method harmonization assessment were reviewed. The assessment was conducted by sharing 343 non-frozen urine samples among 16 commercial quantitative urine albumin methods from 5 global manufacturers. The results became available in April, 2011 and for all samples had median values that varied 33% among procedures, and when results were restricted to 10-400 mg/dL the medians varied 20% among procedures. The measuring intervals varied among procedures and the instructions for dilutions were incomplete or missing in the IFUs for some procedures. The results will be more thoroughly evaluated in the months ahead. Diluted DA470k/IFCC from IRMM and a candidate urine albumin reference material from JSCC were included in the measurements and results will be assessed to determine commutability of the materials among the procedures examined and the effectiveness of these materials as common calibrators. The JSCC material is value assigned by traceability to diluted DA470 so will have a larger uncertainty in value assignment. Follow up of discrepant patient sample results among methods will include assessment of molecular forms by tandem mass spectrometry analysis.
2. Preliminary results from assessment of albumin adsorption to container plastics used for collection of urine and for instrument cups suggested that adsorption is less than 2% at concentrations of 20 mg/L.
3. The experimental design for the physiologic variability assessment project was briefly reviewed. A recent report that used 250 stable hypertensive patients was reviewed by Sverre Sandberg who pointed out that its experimental design was intended to evaluate different urine collection times and did not control all aspects necessary for a CVi determination. It was decided that the original study should proceed after review of the experimental design and in particular the urine collection to ensure the likelihood of successful completion.
4. David Bunk reported that NIST and John Lieske at Mayo Clinic are both developing candidate reference measurement procedures based on ID-LC-MS/MS methods. The two groups are collaborating to develop and validate the methods. The NIST method will use a larger number of ion pairs to allow investigation of molecular forms present in urine samples. Both procedures will need a certified urine albumin reference material to be used as a calibrator. Neil Greenberg suggested to investigate a highly purified albumin available from Ortho Clinical Diagnostics.
5. David Bunk reported that NIST is developing a urine matrix creatinine reference material. An action item is to develop quality specifications for this reference material. It may also be desirable to develop a urine matrix albumin reference material but its value assignment requires finalization of the candidate reference measurement procedures.
6. The need for investigation of decision thresholds for urine albumin that included age, gender and race, and considered the risk of kidney and cardiovascular damage at lower concentrations than the current general cut points was discussed. Recommendations need to be developed for the limit of quantitation for urine albumin and for measurement quality specifications especially at lower concentrations.
Page last updated: March 1, 2012