Haematocrit correction of whole blood phosphatidylethanol (PEth) concentrations to estimate erythrocyte PEth concentrations: sensitivity, specificity and influence on test utility

Phosphatidylethanol (PEth) forms in erythrocyte membranes after alcohol consumption, offering a persisting biomarker, that is measurable in whole blood, washed erythrocytes, and dried blood spots. For a predominantly erythrocyte-restricted analyte, erythrocyte concentrations seem to have most validity in patients who are anaemic through alcoholism or other pathology, despite preparation increasing assay complexity. Differences in specimen preparation alters PEth concentrations from the same patient, meaning that criteria for interpreting PEth results should relate to specimen type, presenting a barrier to achieving harmonisation. We therefore tested whether erythrocyte PEth might be validly calculated by haematocrit correction of a whole blood PEth measurement.

PEth testing primarily serves to distinguish drinkers from non-drinkers. In choosing between specimen types, it is important to compare their utility in separating those two groups. We therefore processed 281 blood samples from 17 non-drinkers and 61 drinkers, to prepare matched whole blood and washed erythrocyte specimens. These were assayed by liquid chromatography-tandem mass spectrometry and compared in identifying alcohol consumption. The erythrocyte PEth concentration in whole blood specimens was also calculated by correcting whole blood concentration by the specimen’s haematocrit, as an alternative to preparing washed erythrocytes. The haematocrit-corrected erythrocyte concentrations were included in these comparisons.

Predictably, this work found that sensitivity was consistently better at the lower cut-off of 8 µg/L than at 20 µg/L. Sensitivities were also higher for washed erythrocytes than whole blood, explained by the lower erythrocyte mass in the same volume of whole blood. Haematocrit-corrected whole blood PEth concentrations correlated with erythrocyte concentrations, except for the four highest values, which did not influence comparative sensitivity. Specificity was 100% for washed erythrocytes, whole blood and haematocrit-corrected whole blood at either cut-off, because non-drinkers had undetectable PEth. We conclude that haematocrit-correction of whole blood PEth concentrations theoretically provides an alternative to preparation of washed erythrocytes.