One uncertainty of the resultant BAC lies in the fact that an experimental measurement is multiplied by a number with an unknown relationship to the individual whose blood was analyzed. Harger et al (3, 4) conducted numerous studies for the estimation of the partition ratios of alcohol between air and water, urine, and blood which indicated that variable ratios were reported. Although the experimental determination of the actual alveolar BAC/BrAC ratio has not been possible, numerous studies have shown that considerable variation and disagreement exists concerning the alcohol partition ratio of "deep lung air", the breath sample obtained after the usual procedure during which the officer is encouraging the driver to blow longer and harder. Frajola (5) in Table 2 page 43 lists 10 investigations from 1941 to 1974 with ratios varying from as low as 1004/1 to a high of 7289/1 and average ratios from 1307/1 to 3478/1. Mason and Dubowski (6) report correlation studies between test results in which the devices using the 2100/1 as the ratio were compared to actual blood alcohol test results. Their results showed that for the Breathalyzer instrument only 38% of the results agreed within ± 15%. Jones (7) in 1976 reported that the ratio not only varied between individuals but, also, within the same individual from time to time; i.e. it was less than 2100/1 during the alcohol absorption phase and higher than 2100/1 during the elimination phase. When Harte (8) in 1971 introduced the Intoxilyzer which measured the alcohol in one's breath by infra-red spectroscopy, he used the 2100/1 ratio and claimed that the conversion of the breath alcohol test result to a BAC was based upon the established principle that alcohol obeys Henry's law. 

Careful examination of his so-called "established principle" reveals that some of the conditions required by Henry's law are not met. In addition to the uncertainty described above a second uncertainty involves the constant temperature requirement of Henry's law. As mentioned above the 2100/1 ratio was determined from experiments with water and alcohol solutions at a constant temperature (34°C.), but blood differs from water. The temperature of the blood in the alveoli may vary according to Mason and Dubowski (6) from 35.8°C. to 37.2°C. The range may be widened by hypothermia or hyperthermia due to fevers, infections, exercise, and the use of various medications. A third uncertainty producing condition involves the variable hematocrit. Because blood contains red blood cells which resist penetration by alcohol, variations in the hematocrit of the drivers' blood essentially means that variations in the volume of water will affect the calculation of the BAC. Mason and Dubowski (6) indicate that the potential error due to hematocrit variables is not likely to be a large one varying from 0.105 to 0.120% W/V for hematocrit ranges from 24% to 76%. A fourth uncertainty involves the presumption that the subject providing the breath sample was in the post-absorptive phase of alcohol absorption. Although at least a 15 minute interval between the apprehension and breath test of the suspect is required to wash out any mouth alcohol remaining from a recent drinking experience, the status of absorption is uncertain. A plot of the blood alcohol concentration after completion of a drinking experience will show an initial rapid rise in the BAC (the absorption phase), a peak or slight plateau, followed by the elimination phase (a continuous decline to the zero concentration). Because the breath alcohol results from the alcohol in the pulmonary arteries to the alveoli, the measurement is presumed to be representative of arterial blood. When blood is taken for the alcohol test it is venous blood that is obtained and measured. Large differences up to 0.06 to 0.09 %W/V may occur between the arterial and venous BAC's during the absorptive phase (6). Only when the alcohol has been completely absorbed and distributed throughout the body will the BAC and BrAC be equivalent. The defendant who has the venous blood test taken soon after his last drink (before absorption and distribution is complete) will have a lower BAC than if he had taken a breath test. 

A fifth uncertainty, and probably the most important one, is the result of recent new knowledge of lung physiology and its application to breath alcohol testing by Michael Hlastala (9). His studies show that the alcohol exchange process that occurs in the alveoli continues throughout the passage of the breath sample from the lungs to the mouth, and that the alcohol exchanges in the airways result in great variability of the BrAC. Before delving into that recent data concerning BrAC it is of interest to return to the concepts of the 1970's. 

In 1972 an ad hoc Committee on Alcohol Blood/Breath Ratio, twelve expert scientists with Borkenstein (10) as chairman met at Indiana University Law School to review the problem of various breath testing instruments producing varying results unless all agreed to use one specific ratio. They agreed that "available information indicated that 2.1 liters of expired alveolar air contained the same amount of alcohol as one milliliter of blood (2100/1). Four of the twelve committee members had special interest in an acceptable ratio for all instruments. Harger, Borkenstein, Forrester, and Goldberg each were associated with a specific breath test instrument. Dubowski, a member of that committee, later (6) wrote that although competent laboratories reported values ranging from 1900/1 to 2400/1, "the 2100/1 ratio was reaffirmed, but essentially by fiat." 

Mason and Dubowski (6) in their resume of alcohol testing wrote: "Various factors involved in discrepancies between results of analyses of near-simultaneous venous blood and breath specimens from the same subject are examined. Because the causes of these discrepancies cannot adequately be controlled in law-enforcement practice, we suggest that calculation of a blood-alcohol concentration based on the result of a breath analysis be abandoned. We recommend that when breath analysis is performed for law-enforcement purposes, the interpretation of the result should be statutorily base on the amount of alcohol found per unit volume of alveolar ("deep lung") air." This was the first call for a per se status for the BAC. In 1976, two years later, Mason and Dubowski (6) wrote: "In a previous communication dealing in part with the imponderables encountered in the conversion of a breath quantity to presumed blood concentration a portion of the summarizing statement was "that in actual law-enforcement practice, when a breath test is analyzed for alcohol, the quantity found cannot be used to calculate the simultaneously existing actual blood concentration without making assumptions having uncertain validities in any given case because they have not been assessed."

Informal discussions with a number of scientists who have had lengthy experiences with the technical, administrative, and forensic aspects of breath-testing did not bring forth any significant disagreement with the factual information on which that statement was based and which has been reviewed and extended here." Shortly after this publication most of the states changed their DUI statutes to make it a crime to drive with a proscribed concentration of alcohol in your breath (either 0.08 g/2100 ml. or 0.10 g/2100 ml.) This is an illustration of the legal system over-ruling science. What happened to the uncertain validities in the breath test? Have all the devices made changes to measure the temperature of the breath samples? Has any one found a way to determine the status of alcohol absorption when the breath test is given? Is there a device that can determine the blood to breath ratio of the individual being tested or do all the devices continue their use of the 2100/1 ratio? 

After their suggestion to make the BrAC a per se violation Mason and Dubowski (6 )wrote the following: "It would appear to the advantage of all concerned to remedy these accumulated defects in the application of breath testing to the problems of alcohol and traffic to do this by whatever means are found necessary. To attempt remedy "to whatever extent is practicable" is to invite the usual resistance to the slightest change in human affairs, such resistance having in the past been evident in the case of matters having to do with alcohol and traffic." It is obvious that "whatever means necessary" was successful in making scientific discrepancies legally acceptable!

Before closing this report on the blood/breath ratio, it is imperative to review the comments of Michael Hlastala, Professor in the Departments of Physiology, Biophysics, and Medicine at the University of Washington, Seattle, WA. He has conducted research and published his results in lung physiology for the last 30 years. His invited review article (9) on the alcohol breath test for the Journal of Applied Physiology in 1998 describes his studies as follows: "Alcohol excretion by the lungs is via diffusion from the bronchial circulation through the airway tissue, where it is absorbed by the inspired air. By the time air reaches the alveoli, the airstream is in equilibrium with airway tissue and BAC. Therefore, no additional alcohol can be absorbed or desorbed in the alveoli. During inhalation, the alcohol absorbed by the air from the mucosa is partially replaced by alcohol from the bronchial capillaries. On exhalation, some of the alcohol is desorbed back to the airway surfaces. All of the alcohol exhaled at the mouth comes from the airway surface via the bronchial circulation. No exhaled alcohol originates from the pulmonary circulation in the alveoli. The fact that alcohol is derived from the airways explains why the BrAC can be so easily altered by the breathing pattern. This contributes to the very large variation in the ABT (alcohol breath test) readings obtained from actual subjects." It is long past the time when the theories of the 1930s through the 80s should be discarded and the new recognition and research on breath tests accepted to improve the accuracy of BrAC analyses.

References

1. M. F. Mason and K.M. Dubowski, "Alcohol, Traffic, and Chemical Testing in the United States: A Resume and Some Remaining Problems" Clin. Chem. 20/2, 126-140 (1974). 

2. R. N. Harger, "Alcohol/Water Partition Ratio" Science News, 73, 1892, (1931). 

3. R. N. Harger, E. B. Lamb, and R. R. Hulpieu, "A Rapid Chemical Test for Intoxication Employing breath" J.A.M.A. 110, 779-785, (1938). 

4. R. N. Harger, E. B. Raney, E. G. Bridwell, and M.F. Kitchel, "The Partition Ratio of Alcohol between Air and Water, Urine and Blood; Estimation and Identification of Alcohol in these Liquids from Analysis of Air Equilibrated with Them" J. Biol. Chem. 183, 197-213 (1950). 

5. W. J. Frajola, "Defending Drinking Drivers" p. 43 Coral Press, Columbus, OH (1980). 

6. M. F. Mason and K. M. Dubowski, "Breath-Alcohol Analysis: Uses, Methods, and Some Remaining Problems" J. Forensic Science, 21, 9-41 (1976). 

7. A. W. Jones, "Variability of the Blood:Breath Ratio in vivo"J. Studies Alcohol, 39,1931-1939 (1978). 

8. R. A. Harte, "An Instrument for the Determination of Ethanol in Breath in Law-Enforcement Practice" J. of Forensic Sciences, 16 (4), 493-510 (1971). 

9. M. P. Hlastala, "The Alcohol Breath Test: A Review"J. of Appl. Physiol. 84, 401-408 (1998). 

10. R. F. Borkenstein et al "Proceedings of the ad hoc Committee on Alcohol Blood/Breath Ratio" Indiana Univ. Law School, Indianapolis, IN, January 1972.

Those wishing to respond, question or add relative information are urged to contact me by email.