Yet another potential source of error in breath-alcohol analysis involves the variability in the composition of the blood.
Whole blood is made up of a mixture of solid particles suspended in a liquid. The solid particles consist of red blood cells, white blood cells, and clotting platelets; the liquid portion is called plasma. The percentage by volume of the solid particles is called the hematocrit of the blood. Thus, for example, a hematocrit of .47 would indicate that the individual's blood consists of 47 percent solid particles (cells and platelets) and 53 percent plasma.
Since plasma is a liquid and contains water, and alcohol is more soluble in water, it will contain a higher concentration of alcohol than will the solid particles in the blood. And the less plasma in the blood (i.e., the higher the hematocrit), the more concentrated the alcohol will be in the existing plasma. Thus the higher the hematocrit, the higher the alcohol concentration in the plasma. To put it another way, if two subjects have the same blood-alcohol concentration in their bodies but one has a higher hematocrit, that person's plasma will have a higher concentration of alcohol.
Applying this to breath analysis, the air in the lung is absorbing alcohol by diffusion from the blood washing the alveolar sacs in the lung tissue. This process follows Henry's Law, which applied to breath testing can be stated as: The concentration of alcohol in the deep lung air is directly proportional to the concentration of alcohol in the blood surrounding the alveolar sacs. However, Henry's Law applies to liquids, not solids, and so the breath is going to reflect the alcohol concentration in the plasma (a liquid) more than it will the alcohol concentration in the solid particles of the blood.
Since blood with a smaller percentage of plasma (higher he matocrit) will have a higher concentration of alcohol, the lung air will reflect this - that is, there will be a higher concentration of alcohol in the air above the alveolar sacs. In other words, the plasma in blood with a higher hematocrit (less plasma) will have a higher BAG and this will cause the breath to have a higher BAG.
Bottom line: Breath machines report falsely high bloodalcohol readings for persons who have blood with a high hematocrit.
What is a normal hematocrit? And what is the range of variation? The average hematocrit for men is 47 percent, with a range of 42 to 52; women's hematocrit averages 42 percent, and ranges from 37 to 47. But studies have shown that an individual's hematocrit can vary in time by as much as 15 percent.
The hematocrit of an individual can easily be determined by an appropriate blood test. And if it is high, counsel should consider introducing evidence of that figure - along with expert evidence of its significance (if necessary, during cross-examination of the state's expert).
The effect of an individual's hematocrit on breath analysis can be mathematically computed. The partition ratio of 2100:1 uniformly used in breath testing presumably uses a male-female average hematocrit of 45 percent. If the client's hematocrit is, say, 54 percent, the breath test result could be computed by multiplying it by 45/54. Assuming a breath test result of .11 percent, for example, the true BAG could be determined by the formula. 11 X 45/54 = .09. In other words, a person with a true BAG of .09 percent and a hematocrit of 54 percent would test on an otherwise accurate breath device as .11 percent.
Finally, there should be inquiry as to whether the subject was at the time of the arrest suffering from anemia. The percentage of plasma in an anemic subject's blood will be higher than normal. In other words, his blood will have a higher percentage of water. As alcohol is attracted to water, there will be a higher percentage of alcohol in the blood. Technically, this may not be a defense: The offense consists of blood-alcohol concentration and, as with sexual and racial variation in alcohol metabolism, there is no allowance for physiological differences. Nevertheless, juries are human, and evidence that the BAG reading does not accurately reflect the amount of alcohol consumed - that the defendant is being prosecuted largely because of his anemic condition - will go a long way toward an acquittal.