HbA1c – when should we not use it?

It can’t only be me who remembers that cringey moment when, during an exam OSCE station, or a viva, you’d confidently reel off a management plan and the investigations you’d do. 

“I’d then check the patients full blood count, renal and liver function, and thyroid function”. 

*Pause*

“Please explain why you’re doing each test and what you hope to find”

Oh dear….

It is an excellent and valid question, and there is a pragmatic balance here. We’ve talked on the course before about understanding the positive and negative predictive values of tests, and how using certain blood tests as a “screening” tool, when that’s not what they are designed for (cancer markers being a good example), can provide false reassurance. But realistically, sometimes we are adopting a broad approach. Or to quote one of my medical consultants “going on a fishing expedition”. Sometimes understanding why you shouldn’t do a test is as important as knowing why you should. 

Those of you who remember back before 2011, will have lived in the world I did that was pre-HbA1c. When this test became widely available, it made the diagnosis and monitoring of diabetes significantly easier. We have been delivering the majority of diabetic care in general practice for well over a decade now, and the HbA1c remains a key part of this. But when should we not be ticking that box?

The test measures the amount of glucose attached to the haemoglobin in the red blood cells. Any condition that affects red blood cells life span will potentially affect the result, as will any abnormality in haemoglobin structure. Compared to isolated glucose measurements, HbA1c is overall a better indicator of glucose exposure and the risk of long-term complications, with less physiological variability. It also doesn’t need the patient to have fasted or have particular timing of samples. Remembering that HbA1c levels directly correlate with glucose levels and the occurrence of complications, levels above 48mmol/mol are significantly sensitive and specific to identify those patients at risk, especially of developing retinopathy. 

A common example where a HbA1c result will not be accurate, is iron deficiency anaemia, where the body hangs on to the RBCs for longer, so their lifespan increases and exposes them to more glucose than usual. This causes a falsely elevated HbA1c. 

Hb A1c also doesn’t alter when there are acute changes in plasma glucose levels; one of the reasons why it is not suitable to diagnose T1DM. Other groups that HbA1c is not recommended in include:

• Children and young people – the unpredictable growth patterns of children affect their blood sugar levels, and the age specific cut-offs needed for accuracy don’t exist as they do in adults. 
• Pregnancy and up to 2 months post-natal – high red cell turnover and lack of reflecting recent change means oral glucose tolerance tests are preferred in pregnancy
• Patient taking medications that can cause rapid rises in glucose e.g. corticosteroids or some antipsychotic drugs e.g. olanzapine
• Patients with HIV – these patients have a shorter red cell life cycle, and the lower CD4 count if present will also impact on accuracy. A number of medications used to treat HIV also affect Hb A1c readings. 

If you’re looking at a HbA1c result, and thinking “this doesn’t quite fit”, take a pause and look at the patient’s full blood count, repeat prescriptions and medical history. Weight loss but a rising HbA1c, or a HbA1c that doesn’t correlate with home glucose readings should make us stop. If in doubt, follow the old adage, look at the patient and not the numbers. 

Diabetes UK has good patient information on understanding the HbA1c test which can be found here on their website What is HbA1c? | Blood Test | Target Levels | Diabetes UK and for more information on Abnormal Bloods, join our brand new Abnormal Bloods course live on Friday 7th March or on demand when it suits you.