Tube under-filling and artefactual hyperkalaemia

Artefactual hyperkalaemia is a common finding in samples taken in general practice, and so the recent article by Huyghe et al. ¹ in which two types of blood collection tubes were compared was most welcome. No significant differences in results were noted between the two tubes studied, but the investigators did report that ‘the risk of hyperkalaemia was 4.8 times higher in serum from tubes that were incompletely filled.’ They concluded that it is essential that blood collection tubes are correctly filled to avoid erroneous potassium results.

While it is obviously sensible to fill a collection tube with the recommended amount of blood (and sometimes this is essential, for example when testing blood coagulation), this is often difficult in practice. Elderly persons, children, those with marked peripheral oedema, intravenous drug abusers and subjects with peripheral vasoconstriction often present considerable difficulty to phlebotomists and frequently only a small amount of blood can be collected. It is therefore critical to know with certainty whether tube under-filling significantly impacts on the interpretation of biochemical results.

Unfortunately, the investigators did not assess the impact of the haemolysis index of the samples in question. Given that both potassium and lactate dehydrogenase concentrations were higher in under-filled tubes, one might reasonably suggest that the degree of haemolysis was higher in these samples and it was this alone that was responsible for the findings reported. Haemolysis is known to affect several commonly measured biochemical parameters. ²

The way in which venepuncture is performed is known to impact on the likelihood of haemolysis. ³ Perhaps the under-filled tubes contained blood from subjects on whom venepuncture was technically difficult or blood collected using a less commonly used method such as via an intravenous cannula. Either problem might have increased the haemolysis of these samples.

The relation between tube under-filling and haemolysis should be assessed more thoroughly in this study before the authors’ conclusion is accepted.