Dear Mailbase.

 

We have recently identified a problem with some glycated haemoglobin analysis.

As with many other laboratories who have adopted the lean process, we operate with a single sample policy whereby the same EDTA sample is used for analysis of FBC, plasma viscosity (PV) and HbA1c. We were operating with the analysis order: FBC then PV then HbA1c. Therefore, requests which have PV and HbA1c were being centrifuged (2700 g, 10 min) before HbA1c analysis. Samples were mixed before HbA1c analysis, typically by a single inversion. We have demonstrated that centrifugation under these conditions leads to fractionation of the red blood cell population: we hypothesise that the oldest, most dense cells are forced to the bottom of the tube. As our analyser (TOSOH G8) samples 1mm from the bottom of the tube, it was sampling into the bottom of the cells, which were still packed after a single inversion of the sample, and giving an HbA1c result that reflects the population of the oldest and smallest cells in the sample i.e. artificially raised.  The effect is quite striking, and can be different from the true result by up to 10 mmol/mol.

 

We have brought this discussion to mailbase to highlight the issue with other laboratories in case any other practice is to use previously centrifuged samples for HbA1c analysis.

 

If you think that your lab (and patients) may have been affected by a similar practice, I would be extremely grateful to hear from you.

 

Many thanks in advance,

 

Mandy

 

Dr Mandy Perry

Clinical Biochemist

Blood Sciences, Template A2

Royal Devon & Exeter Foundation Trust

Barrack Rd

Exeter

EX2 5DW

Tel: 01392 402948