Reporting HbA1c: Which Is Better—Percent or SI Units? 
Considerable international effort has gone into standardizing HbA1c measurements and how they are reported. In the United States, most labs report HbA1c results in percentages only, whereas globally many labs report HbA1c results in Système International (SI) units (mmol/mol) only, or in both SI units and as a percentage. Respected authorities on this subject, Ian Young, MB BCh, MD, FRCPath, and David Sacks, MB ChB, FRCPath, recently shared their views on how HbA1c should be reported.

Ian S. Young

Ian S. Young, MB BCh, MD, FRCPath

Report HbA1c in SI Units
In 2007 an international consensus statement on the standardization of HbA1c measurement was published on behalf of the American Diabetes Association, European Association for the Study of Diabetes, International Federation of Clinical Chemistry and Laboratory Medicine (IFCC), and the International Diabetes Federation. This statement recommended that HbA1c results be reported worldwide in IFCC (SI) units and derived National Glycohemoglobin Standardization Program (NGSP) units (%).

The statement was the culmination of a program of work led by IFCC to develop a reference measurement system for HbA1c. Prior to this, HbA1c was measured in different ways around the world, meaning that results tended to be different between countries and laboratories. The new IFCC reference method, which specifically measured the concentration of only one molecular species of glycated A1c, provided a firm anchor for HbA1c measurement and ensured that measurements could be accurate
as well as precise.

Use of the SI unit enabled HbA1c results to be expressed in a scientifically correct way for the first time; introduction of this unit—proposed to be used alongside NGSP units (%)—provided a clear indication of the significant international change in HbA1c measurement that had taken place.

In much of the world, where laboratories report glucose in mmol/L, HbA1c expressed in percent has very similar numerical values to glucose expressed in mmol/L, tending to confuse patients about the meaning of the two. Reporting HbA1c in SI units gives numerical values significantly different from usual glucose values, helping patients to distinguish the different meanings of the two tests, and indeed, when SI units were introduced, clinicians had an opportunity to educate patients and focus their attention on the differences between HbA1c and glucose tests. Critics have suggested that a change in HbA1c units might confuse patients, and cause them to have poorer glycemic control; however, evidence suggests that with effective education for both patients and clinicians this is not the case.

Increasingly, medical information is shared globally and patients access information via websites which do not respect national boundaries. As a discipline, laboratory medicine promotes scientific excellence and therefore should produce standardized results traceable to reference measurement systems which provide a long-term guarantee of consistent results. In this context, reporting on HbA1c in SI units, alongside NGSP units (%), where appropriate, is in the interest of patients with diabetes worldwide.

Ian S. Young, MB BCh, MD, FRCPath, is a professor of medicine and director of The Centre for Public Health at Queen’s University Belfast in Northern Ireland. 


David B. Sacks

David B. Sacks, MB ChB, FRCPath; Verena Gounden, MB ChB

Report HbA1c in Percentage

HbA1c testing is essential to managing diabetes mellitus, used as it is to evaluate long-term blood glucose control, decide whether to change or intensify treatment, and to diagnose diabetes. Influential clinical organizations recommend measuring HbA1c at least every 6 months in all patients with diabetes.

In 1993 the Diabetes Control and Compli­cations Trial (DCCT) documented the clinical value of HbA1c testing. This large prospective randomized study showed that HbA1c values predict the microvascular complications of type 1 diabetes and that intensive glucose control significantly reduces these complications. The United Kingdom Prospective Diabetes Study reported essentially identical findings for type 2 diabetes. Both these landmark studies reported HbA1c in % and their values are directly comparable. This led to widespread adoption of DCCT values (%) in patient care. Both clinicians and patients are familiar with the DCCT units, and in contrast to most other laboratory tests, many patients know their HbA1c values and target—usually 7%—for good glycemic control.

The IFCC developed a higher order reference method for HbA1c. While the relationship between IFCC and NGSP values are linear, their absolute numbers are very different. For example, an HbA1c value of 7% is equivalent to 53 mmol/mol. Starting to report HbA1c in SI units rather than as a percent would be a costly endeavor with limited, if any, benefits to patient care and with potential for harm. No study has shown any improvement in glycemic control or patient outcome with HbA1c results reported in SI units rather than as a percent.

Confusion over values is highly likely to be detrimental to patient care. Therefore, an extensive education program for patients and healthcare professionals would be required to prevent confusion and keep patients from veering away from optimal glycemic control. This education program would necessarily be prolonged, extending well beyond the changeover period.

To implement such a change nationwide also would carry a substantial financial cost. In countries with limited resources, available funds would be much better directed at improving the quality of testing and access to services than spent on changing the reporting units for HbA1c. For laboratories, there would be costs associated with adapting instruments, computers, and laboratory information systems (LIS) to provide HbA1c values in SI units. However, many point-of-care devices do not allow automatic reprogramming. Results would have to be converted from percentages by the electronic medical record, LIS, or by manual conversion tables, charts, or various electronic media apps. This would increase the likelihood for reporting erroneous results.

In conclusion, considerable effort and funds would need to be invested to begin reporting HbA1c in SI units, which would not necessarily improve patient care. Thus, the rationale for change is obscure.

David B. Sacks, MB ChB, FRCPath, is chief of the clinical chemistry service in the Department of Laboratory Medicine at the National Institutes of Health (NIH) in Bethesda, Maryland. Verena Gounden, MB ChB, a visiting clinical chemistry fellow, contributed to Sacks’ position. Work in their laboratory is supported by NIH’s Intramural Research Program.