Summer 2008 Meetings
Clinical Advisory Committee | NGSP Steering Committee | Manufacturer Forum

2008 NGSP Clinical Advisory Committee Meeting
American Diabetes Association 68th Annual Scientific Sessions, June 2008

The CAC meeting was attended by representatives of various clinical diabetes organizations (e.g. IDF, ISPAD, ADA, EASD). The major topics of discussion were hemoglobin variants, variability among laboratories and methods, the results of the ADA/EASD/IDF study of the relationship between HbA1c and average glucose, possible racial differences in HbA1c values and the use of HbA1c for screening/diagnosis of diabetes.

1.  Hemoglobin Variants--Randie Little, NGSP Network Coordinator

• Based on CAP survey data 11% of labs are currently using methods with clinically significant HbAS and/or HbAC interference (these are the two most common variants in the U.S.).
• The Roche Cobas Integra is being phased out and replaced by the Roche Integra Gen. 2 which does not show interference from these variants, once this process is complete only 5% of labs will be using methods that show significant HbAS and/or HbAC interference.
• Twenty-one percent of labs are using methods that show clinically significant interference from HbAE (the second most common variant worldwide) and 6% are using methods that show interference from HbAD (fourth most common).
• The NIDDK and NGSP have information regarding the influence of these variants on HbA1c results posted on their respective web sites.

• The NGSP has resulted in considerable improvement in the comparability of HbA1c results since 1993.
• In 2007 the NGSP tightened the manufacturer certification criteria for the 95% CI of the differences from +/-1% HbA1c to +/-0.85% HbA1c.
• The CAP began using accuracy grading based on NGSP targets for the GH2 survey beginning in 2007.  The original acceptable limits were +/-15% of the NGSP target, this has been reduced to +/-12% for 2008.
• 2008 GH2A CAP survey data:
• Most method-specific between-lab CVs were <5%.
• Method-specific medians were within 0.3, 0.6 and 0.9% at the low, middle and high levels, respectively.
• Two recent articles examined the effects of variability on the clinical interpretation of HbA1c results
• Holmes et. al, Amer J Clin Path 2008;129:540-7
• Compared the differences between two assay methods in the same location, a DCA2000+ and a Variant II, over time.
• Also compared 7 methods from the CAP GH2 survey over time
• Concluded that between-method variability is a potentially significant source of inaccuracy when HbA1c results are interpreted based on clinical decision thresholds.
• Westgard (A War of Words in Clinical Medicine Part IV: A Sentinel Example, www.westgard.com)
• Reviewed CAP GH2 survey data and evaluated accuracy, bias, total error and reference change values (RCVs)
• RCV is the change in a patient’s serial test results that can be considered significant and not just due to analytical or biological variation.
• Concluded that overall agreement in terms of bias is very good but some methods still have problems with accuracty and/or imprecision.
• CAP GH2 data show that agreement between methods and laboratories has not improved between 2006 and 2008.
• All-method CVs on the CAP GH2 surveys have remained between 4 and 5% corresponding to RCVs of 0.73-0.88% HbA1c at a level of 7% HbA1c (90% probability).
• The RCVs can be reduced by either lowering the probability (e.g. from 90% to 80%, is 80% clinically good enough?) or by reducing analytical CVs.
• The question is: what is a reasonable goal for variability?

Discussion:  Some of the less precise assay methods are run on large analyzers that run many tests.  Large reference laboratories prefer these systems because they represent cost savings.  Some methods show small imprecision but significant biases vs. NGSP targets that seem to persist over time.  Manufacturers sometimes have issues with calibrator value assignments; commutability across different platforms can be a problem.  When labs fail the accuracy-based CAP criteria they will put pressure on the manufacturers; next year the CAP criteria will be reduced to +/-10% of the NGSP target.  The NGSP is trying to get large reference laboratories to certify as level 1 labs so that their performance can be monitored.

3.  IFCC Working Group on HbA1c Standardization Update--David Sacks, Chair, NGSP Steering Committee

• The IFCC held a meeting with manufacturer representatives in December, 2007 in Milan to discuss the implementation of traceability and the implications of the ADA/EASD/IDF/IFCC Consensus Statement.
• All instrumentation sold after January 1, 2011 must report HbA1c results in both IFCC (mmol/mol) and NGSP (%) units
• Quality assurance programs should use commutable materials with target values assigned by the IFCC reference system.

Discussion:  Reporting of results is likely to be country-specific.  Sweden has its own system now but will likely move to IFCC and/or NGSP numbers.  The UK has decided to report both IFCC and NGSP numbers but not estimated average glucose (eAG).  The NGSP is already considered traceable to the IFCC reference system and most manufacturers are already traceable.

4.  ADA/EASD/IDF Study of the Relationship Between HbA1c and Average Blood Glucose: Future Plans—David Nathan, NGSP Steering Committee

• The study was designed to determine if there is a relationship between HbA1c and average blood glucose (AG) such that HbA1c results can be expressed as AG.
• Unfortunately different ethnic groups were underrepresented in the study.
• The results have been published in the August 2008 issue of Diabetes Care.
• Results show that HbA1c results can be expressed as estimated average glucose (eAG) for most patients with Type 1 and Type 2 diabetes.

Discussion:  This study is more definitive than previous studies that examined the HbA1c/AG relationship due to both a large number of patients and a much larger number of glucose observations per patient.  The size of the subgroups was small, and children and pregnant women were not included in the study.  The numbers for AG were slightly lower at given HbA1c levels than those derived from the DCCT that have been published in the ADA Clinical Practice Recommendations and are therefore used by physicians.

5.  Is HbA1c a Valid Measure of Glycemia across Racial and Ethnic Groups?  William Herman, University of Michigan

• Factors that have been associated with variation in HbA1c include gender, sex hormones, visceral fat distribution, genetics, intraerythrocyte 2,3-diphosphoglycerate, and intraerythrocyte fructosamine 3-kinase.
• Data from the third National Health and Nutrition Examination Survey (NHANES III) showed racial and ethnic differences in HbA1c in non-diabetic children and young adults ages 5-24.  HbA1c in non-Hispanic whites is lower than in Mexican-Americans which is in turn lower than in non-Hispanic blacks.
• The same pattern has been observed in other trials including the DPP, ADOPT and DURABLE.
• Whether these differences reflect difference susceptibilities to complications is not known.

Discussion:  Without CGMS results one cannot know the day-to-day variability swings in glycemia.  Could these differences be related to differences in erythrocyte lifespan?  African-Americans are a very diverse group, there may be differences within that group alone.

6.  HbA1c for Screening/Diagnosis—Michael Steffes, NGSP Steering Committee

• The issue of using HbA1c for screening/diagnosis of diabetes has repeatedly come up over the years.
• HbA1c is technically a better laboratory test than glucose.
• An ADA Expert Committee has now agreed to recommend HbA1c for screening/diagnosis but no cutoffs have been established.

Discussion:  Nothing has been decided regarding pre-diabetes, the data requires further examination.  At some point cutoffs will need to be established; is sensitivity more important than specificity?  Many physicians are already using HbA1c for screening/diagnosis.  If HbA1c is to be used for this purpose, having accurate assays  is extremely important.  Medicare is very strict about what tests are used for screening/diagnosis; if all groups agree to the recommendation it will still be 4-5 years before CMS adopts this policy.

Clinical Advisory Committee | NGSP Steering Committee | Manufacturer Forum

NGSP Home