Background: Point-of-care (POC) analysers are crucial in acute care diagnostics. Due to easy access, they are also used in monitoring chronic diseases. However, use of POC devises in measurement of glycated haemoglobin (HbA1c) and monitoring diabetes mellitus is controversial.

Methods: We compared QuikRead go analyser’s HbA1c measurement performance to cobas c513 in a clinical laboratory with leftover venous K2-ethylenediaminetetraacetic acid (K2-EDTA) samples (n=151).

Results: The methods correlated well (r=0.984), but differed statistically significantly (P<0.0001) with an average negative bias of 2 mmol/mol for QuikRead go. The method deviation compared to cobas c513 was high, −20 to 11 mmol/mol (min–max), in HbA1c above 48 mmol/mol (n=98), and less, −5.5 to 1.4 mmol/mol, below HbA1c 48 mmol/mol (n=51).

Conclusions: This deviation in high level HbA1c should not be considered clinically acceptable. Deviating results can cause erroneous decisions regarding diabetes mellitus diagnosis and treatment. In lower HbA1c, the device showed relatively good performance. Concluding from this, the deviation is method-dependent and not related solely on preanalytical issues. These findings demonstrate the importance of understanding analytical variation when laboratory results and various methods are used in monitoring chronic diseases. Further studies with venous and finger-prick samples on QuikRead go analyser are needed to confirm the findings presented here.