Presenters:

Present were members of the NGSP Steering Committee and IFCC Working Group on HbA1c Standardization, and representatives from various manufacturers, laboratories and agencies.

1. NGSP Progress Report—Randie Little, NGSP Network Coordinator

• Status of HbA1c measurement
  • The numbers of certified methods has leveled off but the number of certified laboratories continues to increase.
  • Most certified laboratories are level 1 laboratories and are located outside the U.S.  We now have certified labs in China, India and Russia.
  • There has been much improvement in the comparability of HbA1c results since 1993, but there has been little improvement over the last few years
  • CAP GH2 survey 2008A:
• Between-lab CVs ranged from 1.3 to 7.1%
• 90% of labs are using methods with between-lab CVs<5%.
• Several methods still show significant variability
• One method, the Bayer A1cNow, shows interference from EDTA.
• Efforts to reduce variability
• Beginning September 2007 the NGSP tightened the certification criteria for manufacturers (95% CI of the differences from +/-1% to +/-0.85% HbA1c)
• Beginning with the 2007A GH2 survey the CAP began using the NGSP accuracy grade as the only grading system for HbA1c.  A limit of +/-15% of the NGSP target was used for the 2007 surveys; this was lowered to +/-12% for the 2008A survey.
• The acceptable limit will be further reduced in future surveys.
• Influence from Hb variants
• 11% of laboratories are using methods with clinically significant HbAS and/or HbAC interference
• At least 21% of labs are using methods with clinically significant HbAE interference.
• At least 6% of labs are using methods with clinically significant HbAD interference.
• Once the original Roche Integra method is phased out only 5% of labs will be using methods with clinically significant HbAS and/or HbAC interference.

2. CAP Grading, future plans—David Sacks, Chair,  NGSP Steering Committee

• Prior to 2007 the CAP used peer group grading for HbA1c
• In 2007 the CAP began using accuracy grading with acceptable limits of +/-15%.  This resulted in a 99% pass rate.
• Looking at all of the CAP surveys since 2000 there has been little change in the all-method CVs which have remained around 5%.
• The PT criteria will tighten
  • In 2008 the acceptable limits have been reduced to +/-12%
  • Plan for future grading
    • 2008—Remain at 12%
    • 2009—Reduce to 10%
    • 2010—Reduce to 8%
    • 2011—Reduce to 6%
    • At this point the criteria will be re-evaluated, further reductions are possible.
    • This information will be in future CAP reports (labs will be told whether they would pass at a criterion of 6%) as well as in the minutes of this meeting and on the NGSP web site.
• Looking at pass rates for the 2007B survey, ~97% of laboratories would have passed at a criterion of +/-10% of the target at the middle HbA1c level.
• For the 2008A survey, ~95% of laboratories would have passed at a criterion of +/-8%.
• These pass rates are overall; certain individual methods are lower.
• Levels above 10% HbA1c are not as important clinically, so the CAP will reduce the high level in future surveys to ~9.5 to 10%.

Discussion: HbA1c is not a regulated analyte but if a lab fails it still has to document corrective action. The uncertainty of the NGSP network target value assignment is ~1-2% so manufacturers will have to aim for a goal that is lower than the +/-6% criterion. An ADA Expert Committee has decided to recommend HbA1c for screening/diagnosis of diabetes although a cutoff has not been established, and cardiologists will likely start adding HbA1c to risk factors for cardiovascular disease.  This makes it even more imperative that HbA1c assays improve.  Meeting the goal will be difficult for manufacturers, but the current state of HbA1c testing is simply not good enough to meet clinical requirements.  CAP is very willing to help manufacturers in this process, and will be asked to provide specific survey information to manufacturers to assist them in identifying issues with their individual methods.  CAP will also be asked to put additional questions on future surveys.  The NGSP does not plan to tighten the manufacturer certification criteria in the next year but may eventually do so.  CAP does not currently have the current pass rates at the +/-6% criterion but these will soon be made available, they will also inform labs of the planned changes to the CAP criteria and tell if they would pass/fail the 6% criterion beginning with the next survey.

3. ADAG Study Results— David Sacks, Chair, NGSP Steering Committee

• An international multicenter study to examine the correlation between HbA1c and average glucose (AG) was initiated by the ADA/EASD/IDF.
• Goal was to determine if it would be feasible to express HbA1c as AG.
• Results
  • Subjects:
    • 38% of HbA1c results were between 4 and 6.5%, 44% were between 6.6 and 8.5% HbA1c, and 18% were greater than 8.5%.
    • After exclusions there were a total of 507 subjects.
    • 422 (83%) of subjects were white, 38 (8%) were African/African-American, 39 (8%) were Hispanic and 8 (2%) were other.
    • There was an issue with sample storage at the center in India, so no Asian subjects were included in the study.
  • Data
    • CGMS-Mean of 2400 measurements/subject
    • LifeScan-300 measurements/subject
    • Total of 2700 measurements/patient over 12 weeks
    • 96% of subjects maintained HbA1c values within +/-1% over the course of the 4-month study.
    • The a priori criterion of 90% of AG results falling within +/-15% of the AG/HbA1c regression line was met.
    • The equation is: estimated average glucose (eAG) mg/dl = (28.7xHbA1c)-46.7
    • The numbers are slightly lower than the previously-utilized DCCT numbers
  • Limitations
    • Small ethnic groups
    • No data in children, subjects with renal impairment or pregnant women.
  • Implications for clinical chemistry
    • The ADA has initiated an education campaign.
    • Laboratory IT systems will use the regression equation to convert HbA1c to eAG.
    • Both HbA1c and eAG will be reported (like eGFR).
• The results of the study have been published in the August issue of Diabetes Care.

4. ADA Plans for eAG— Matt Peterson, American Diabetes Association

• It is often difficult to explain HbA1c to patients.  When hemoglobin is mentioned they tend to think of this in terms of a problem with their blood, like anemia.
• We believe that expressing the results in terms of the same blood glucose levels that they use in self-monitoring will be a more-intuitive, one-step process.
• Not all physicians agree that eAG is useful.
• If patients are very familiar with HbA1c they probably do not need eAG.  I think that for most, and especially new patients, it will help them better understand their glycemia.
• One issue is that although the new study has more accurately defined the AG/HbA1c relationship, many clinicians are still familiar with the old numbers and it will take work to “realign” their thinking.
• The ADA is already putting educational information on their web site including an online calculator to convert between HbA1c and eAG.  Thousands of handheld calculators have been ordered.
• We need manufacturers’ help in getting eAG onto lab reports.
• In 2-3 years I feel confident that the vast majority of clinicians will be using these numbers.

Discussion: Labs should be able to report eAG without manufacturers having to re-file with the FDA.  The ADA needs to make physicians aware that children were not included in the study and therefore we do not know if the HbA1c/eAG equation applies to them.  The ADA is not currently planning follow-up studies to examine the relationship in different ethnic groups.  Since the eAG numbers from the study are lower at given HbA1c levels than the numbers derived from the DCCT that are already in use by physicians, education of physicians will be very important.

5. IFCC Network Update—Cas Weykamp, IFCC HbA1c Laboratory Network Coordinator

• The ADA/EASD/IDF/IFCC Consensus Statement was signed in May of 2007 and published shortly thereafter.
• Since that time the results of the ADAG study have been made available and there have been many discussions regarding the units to report HbA1c.
• The IFCC conducted a meeting with manufacturers in Milan, Italy in December of 2007.  Outcomes:
  • All manufacturers’ assays will be required to be traceable to the IFCC Reference System.
  • The name (abbreviation) of the test in lab reports and in the clinical setting should be “HbA1c”, not “A1C”
  • The deadline for implementing traceability to the IFCC reference system will be December 31, 2009 for all instruments in current use.
  • All new instruments sold after January 1, 2011 will report both IFCC SI units (mmol/mol, no decimals) and NGSP-derived units (percentage, 1 decimal) for HbA1c tests in accordance with the consensus statement.
  • Reporting of eAG will not involve analytical systems but only the computer-based systems used to calculate ADAG.
  • The IFCC WG on HbA1c Standardization is willing to review the proposed manufacturer’s traceability chain.
• Tools provided to manufacturers to achieve traceability
  • To achieve traceability, calibrators with both IFCC (mmol/mol) and NGSP (%) numbers as well as HbA1c and total Hb in mmol/L and g/dL, with expanded uncertainties.
  • To check traceability, three levels of controls
    • Medium level is provided with low, normal and high total Hb levels.
    • IFCC (mmol/mol) and NGSP (%) numbers are provided for HbA1c
    • Mmol/L and g/dL are provided for HbA1c and total Hb
    • Expanded uncertainties are provided.
    • Traceability certificates are provided to manufacturers upon request.
  • Monitoring program—to monitor traceability
    • 24 samples/year
    • Results for a given manufacturer are known only to them.
  • Master equation to convert IFCC numbers to NGSP, JDS/JSCC, Mono-S
• Six year progress report for the IFCC Reference System will be published in Clinical Chemistry (in press)
• Almost all manufacturers now utilize IFCC services and have documented traceability
• More information about the IFCC WG and the network can be found at http://www.ifcchba1c.net

Discussion: All new instruments, not just new platforms, sold after January 1, 2011 will be required to report both numbers.  Clinicians will ultimately decide how the information is presented to the patients, it will likely be handled differently in different countries.  Even if instruments are able to report both numbers some hospital systems might not be able to handle this.  With smaller and POC systems that are not connected to a laboratory information system, physicians may have to use offline charts to convert results to eAG.

Clinical Advisory Committee | NGSP Steering Committee | Manufacturer Forum

NGSP Home