The stability of ethanol in human whole blood controls: an interlaboratory evaluation.

Sterile whole human blood control materials were commercially prepared in batches containing anticoagulants and preservatives and approximately 90, 150, and 230 mg/dL ethanol with and without 0.3% (w/v) sodium azide. Aliquots in sealed vials were stored by the manufacturer at 2-8 degrees C until shipped monthly to three academic toxicology laboratories that analyzed them in duplicate by gas chromatographic headspace methods at monthly intervals for one year. The resulting data were pooled, and grand mean values were statistically analyzed to determine the respective alcohol stability in these azide-free and azide-containing blood samples. Azide-containing blood samples showed no alcohol losses during the 1-year period. Azide-free blood containing 1.0% (w/v) sodium fluoride and anticoagulants had small alcohol decreases over time, the total losses after one year being less than 5% of the original alcohol concentrations. The initial alcohol concentration of approximately 40 mg/dL also did not change during storage of additional samples of azide-free blood for one month at 4 degrees C. We concluded that addition of sodium azide to performance-test and control blood specimens for alcohol analysis is unnecessary and unwarranted and that alcohol losses in such blood samples can be minimized by simple appropriate treatments and conditions.