Intended for healthcare professionals

Editorials

Standardisation of glycated haemoglobin

BMJ 2004; 329 doi: https://doi.org/10.1136/bmj.329.7476.1196 (Published 18 November 2004) Cite this as: BMJ 2004;329:1196
  1. Philip Home (philip.home{at}ncl.ac.uk), professor of diabetes medicine,
  2. Jean-Claude Mbanya, associate professor of medicine,
  3. Ed Horton, professor of medicine
  1. School of Medical Sciences, Diabetes, Newcastle upon Tyne, NE2 4HH
  2. Endocrine Unit, Department of Internal Medicine, University of Yaoundé I, BP 8046, Yaoundé 8, Cameroon
  3. Joslin Diabetes Center, One Joslin Place, Boston, MA 02215, USA

    Is a scientific advance, but it could worsen overall blood glucose control

    The detection of sugars in the urine of people with diabetes usually is attributed to Matthew Dobson. However, first attempts at quantification of urine sugars seem to have been made by Francis Home, an Edinburgh physician. Estimation of urine sugar as a measure of severity of diabetes had two major problems that, until recently, were shared with its contemporary equivalent, estimation of glycated haemoglobin. Firstly, both are surrogate measures of the average concentration of plasma glucose that is responsible for the microvascular and arterial damage that manifests itself in later years. Secondly, both also included in their measurement non-specific (nonglucose) substances when using available assays. Very recently, the glycated haemoglobin assay has been standardised by the scientific community to remove this non-specificity, but switching to this more accurate method for everyday reporting of results could lead to confusion and worsening of the control of diabetes.

    Urine glucose remains of value only where the cost of self monitored blood glucose control is out of reach, but glycated haemoglobin (as HbA1c) has become central to any level of diabetes care. In the diabetes control and complications trial (DCCT) and the UK prospective diabetes study (UKPDS), HbA1c was the measure used to define the relation between glucose control and outcomes such as retinopathy.1 2 As a result, HbA1c is the measure by which clinicians can relate blood glucose control to the microvascular and arterial risk in any patient with diabetes. Jeffcoate usefully reminds us, in a recent issue of Diabetic Medicine, of some of the many caveats that underlie interpretation of a series of HbA1c results in an individual.3 However, taken together with self monitored blood glucose profiles and assessment of hypoglycaemia (both of which also have problems related to reliability), HbA1c is likely to remain a cornerstone of management of diabetes for some decades.3

    In the 1980s, difficulties with assays in the Kroc study (the feasibility study for the diabetes control and complications trial) led to urgent and successful attempts to standardise the assay, anchored on methods in David Goldstein's laboratory in Minneapolis. The assay of the UK prospective diabetes study was subsequently aligned with this. With the success of the two studies, the clinical and biochemical communities sought to harmonise clinical laboratory assays to that standard, an exercise that remains a remarkable example of informal international cooperation.

    Such harmonisation served the needs of people with diabetes and their healthcare advisers to near perfection, but unfortunately the ultimate reference standard being used included a fixed non-specific element amounting to approximately 2.0% (HbA1c units) of total haemoglobin, and thus around one third of the measured HbA1c at the upper end of the quoted reference range (6.1% (HbA1c units)). Thanks to the commendable efforts of the International Federation of Clinical Chemists, which has developed a reference standard based on mass spectrographic analysis of a glycated peptide fragment of haemoglobin (summarised by Manley et al,4) the extent of the error is now securely described. With the cooperation of diagnostics' manufacturers (mandated in the European Union), specific results could be available to all people with diabetes, clinicians, and researchers.

    But is this desirable? Nearly every guideline for diabetes has based its intervention and target levels on the standard of the diabetes control and complications trial, and the curves describing the relation between HbA1c and complications from that trial and the UK prospective diabetes study are widely recognised. A further problem, incompletely understood, is that higher numbers of a pathological entity lead to greater clinical action independent of true clinical import—a quirk of human numeracy. Accordingly, a study from Sweden shows that changes in HbA1c standardisation leading to smaller numbers lead to a worsening of overall blood glucose control.5 As a result, adoption of the International Federation of Clinical Chemists' standard in reporting of HbA1c might not only risk losing the benefits of years of professional education on the quantitative relation between overall blood glucose control and complications, but might also risk increasing blindness and cardiovascular events through higher attained blood glucose concentrations.4

    Is there a way out? Not to make use of the advance introduced by the International Federation of Clinical Chemists would be unreasonable. Alternatives considered at a recent meeting convened by the International Diabetes Federation included continued reporting to the diabetes control and complications trial's standard (using the defined relation between the two standards), a coordinated global switch to reporting to the new, lower standard of the International Federation of Clinical Chemists, reporting in absolute units (such as g/kg rather than percentage), or reporting in average blood glucose equivalents. As people with diabetes already measure their own glucose concentrations by self monitoring, the last of these has some attraction. However, even this approach would still need a major global education initiative, and whether the globally influential North American community would be prepared to carry through any such change is unclear at present.

    Meanwhile one message seems clear. Piecemeal switching of the reporting standard by individual laboratories or even countries is likely to lead to confusion and ignorance among patients and healthcare professionals that could seriously damage individual health outcomes in people with diabetes. Therefore, individual practitioners need to remain sensitive to these risks, while requiring the international diabetes community to continue to move forward in achieving a coordinated programme, which makes use of the new standard.

    Footnotes

    • Competing interests PDH and J-CM are officers of the International Diabetes Federation, whose mission is to promote the global interests of people with diabetes.

    References

    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.