Effects of various sample pretreatment procedures on ethyl glucuronide quantification in hair samples: Comparison of positivity rates and appraisal of cut-off values.

The quantification of ethylglucuronide (EtG) in hair is nowadays recognized as the approach with the highest diagnostic performance to evaluate harmful drinking. A widely accepted cut-off of 30pg/mg has been selected after several accurate compared studies. While most of the studies that were used to establish the appropriate cut-off value prescribed to cut hair into small segments before their extraction, hair milling has subsequently been identified as the most efficient pretreatment procedure and was therefore recommended in the last Consensus document issued by the Society of Hair Testing. In this study, we initially compared the results obtained with the two sample preparations, namely cutting and milling, both being applied to the same specimens (n=781). Among these, 205 samples produced measurable EtG values with both methods, with differences ranging from -41.7% up to +415% (the mean increase in EtG concentration, switching from cutting to milling, was +62.1% and the median was +42.3%). Among the aforementioned 205 samples, 29 specimens (3.7% of the total 781 samples) produced significantly different outcome, being classified as negative (i.e., below 30pg/mg) if the cutting procedure is used, but largely positive (above 40pg/mg) when milling is used. Subsequently, the positivity rates obtained on a large population dataset (>27,000 samples) with the two procedures, were retrospectively compared using variable cut-offs values. The percentage of head hair samples with EtG concentration exceeding 30pg/mg upon application of the milling procedure shows a 45% increase (from 10.9% to 15.8%) with respect to cutting procedure, whereas the fraction of hair samples with EtG exceeding 40pg/mg (10.5%) overlaps the percentage of positive samples obtained after cutting pretreatment and applying a cut-off of 30pg/mg. On the basis of these results, it would be worth considering the application of cut-off values linked with the pretreatment procedure, taking into account the results of forthcoming inter-laboratory calibrations.