In Hans Christian Andersen’s childhood story The Emperor’s New Clothes, the naked Emperor is pressured to proclaim his new clothes visible when in fact even he could not see them. The two tailors who swindled him used the fear that he would be seen as unfit for his position as Emperor or simply stupid to force him to accept his new clothes, which weren’t even real. They claimed that the fabric was invisible to anyone not fit for the Emperor’s position or anyone who was just hopelessly stupid. “Beautiful,” the Emperor declared. He was a coward.
It took the bold wisdom of a child to proclaim that the Emperor was in fact naked and that the two tailors were fraudulent. Criminal defense attorneys should remember the lucid voice of the child in The Emperor’s New Clothes as they face government witnesses testifying about breath testing machines. Constrained only by the ethical obligation of zealously pursuing their clients interest the defense attorney seeks the truth while the government toxicologist seeks to defend the program that they work for. In short, the toxicologist defends their paycheck.
Free of the political pressures that control the government witnesses the criminal defense attorney may speak with a bold and lucid voice describing breath testing as “nonscience forensic science.”[1] The criminal defense attorney is the defender of justice and the only voice in the courtroom that solely speaks for the truth. Yet in DUI breath test cases many faulty assumptions still remain unchallenged by defense attorneys.
Prosecution experts from government toxicology labs have sought to build the faulty scientific foundation of breath testing and are too closely associated with the prosecution to challenge its basic assumptions.[2] Due to this failure of toxicology experts to challenge and strengthen the scientific basis of breath testing programs, beliefs that have long been discredited remain a part of the science used to convict defendants of driving under the influence.[3]
In 2009 the National Academy of Sciences and in 2010 the Journal of Forensic Sciences both shed light on the failure of government toxicologists in DUI breath test cases. These two recent publications carry the weight of the establishment and tell us that government experts have failed to remain scientifically neutral.
The National Academy of Sciences, the advisors to our nation on science, engineering, and medicine tell us that the law’s greatest dilemma in its heavy reliance on forensic science, however, concerns the question of whether – and to what extent- there is science in any given “forensic science” discipline.[4] The NAS recommends what seems obvious to some, that “scientific and medical assessment conducted in forensic investigations should be independent of law enforcement efforts either to prosecute criminal suspects or even to determine whether a criminal act has indeed been committed.”[5] The best science is conducted in a science setting as opposed to a law enforcement setting.[6]
In 2010 in the Journal of Forensic Sciences, Dr. Hlastala explains that, historically, lung physiological is to blame for the variability between breath and blood alcohol measurements. Since 1950, as breath testing machines have improved, continued variability between the breath and blood readings are only explained by the dynamic functions of the human body, specifically the lung.[7] Although this new paradigm in understanding lung physiology is now known, it will not be presented in a courtroom because it is contrary to the official program of the Washington State Toxicologist’s. As the NAS tells us, the toxicologist office is too aligned with the prosecutor’s office for forensic science to evolve. No reconsideration of the State Toxicologist’s program will be conducted.
The history of breath testing science in the United States is the history of the union of the prosecution and the government toxicology witness. Beginning in 1950, R.N. Harger Ph.D., Professor of Biochemistry and Toxicology at the University of Indiana Medical School, authored a study establishing a previously unknown variable of breath testing science – human breath temperature. Harger’s study, unsupported by published data and with questionable accuracy, proposed that human breath temperature ranged from 31-35 degrees Celsius and averaged 34 degrees Celsius.[8] However, it is now well established by experts that human breath temperature averages 35 degrees Celsius and does not remain at a stagnant temperature during respiration.
An accurate measurement of human breath temperature is necessary for an accurate breath test reading.[9] Washington toxicologists, using the BAC DataMaster and BAC DataMaster CDM technology, rely on the premise that human breath temperature averages 34 degrees Celsius. Because this premise is incorrect, reasonable doubt as to the accuracy of a breath test reading can be established in all cases.
Attacking and discrediting a breath test in a DUI case should not be an impossible task. By understanding some basic biology and principles of breath testing science, any attorney can effectively attack the breath test and create reasonable doubt in your client’s case. First, you must understand Henry’s Law and how temperature can impact it. Second, you must understand how the lungs work and how alcohol is transported between the blood and the air in the lungs.
William Henry established Henry’s law in 1803. The law holds that at a fixed temperature a volatile substance in a liquid will have a certain numerical relationship (a coefficient) to the same substance in the gas above it. Think of the air in a corked bottle of wine. The alcohol evaporates from the wine and collects in the air above it. If the temperature remains constant, the relationship between the alcohol in the air and the alcohol in the wine can be quantified by a coefficient (a value that expresses the correlation between the two). However, when the temperature varies, the quantity of alcohol in the airspace above the liquid will change. Temperature is a crucial variant in Henry’s Law as the relationship between liquid and vapor is directly affected by a change in temperature. Think of a boiling pot of water and the steam increasing as the temperature increases.
Henry’s Law is used to estimate a blood alcohol level in a subject based on a breath alcohol reading. As a volatile compound, alcohol in the human body obeys Henry’s Law. Just like in the bottle of wine, there will be a numerical relationship between the alcohol molecules in the blood and the alcohol molecules in the air above the blood (in the lungs). Rather than directly measuring the level of alcohol in blood, a breath test machine measures the level of alcohol in the airspace above the blood (lung air). Henry’s Law is used to correlate a blood alcohol level from a breath test reading.
Government toxicologists assume that the number of molecules of alcohol in the blood compared to the number of alcohol molecules in the breath can be expressed as a ratio. They have decided that when there are 2100 molecules of alcohol in the blood that there is one molecule of alcohol in the breath. The accuracy of this ratio, called a partition ratio, is dependent upon the assumption the breath temperature is 34 degrees Celsius. While many variables in human physiology may impact the accuracy of the partition ratio, temperature may have the largest impact on the accuracy of a breath test reading.
Lungs perform the work of gas exchange in our bodies. In each breath, air travels into our lungs and into the more than 300 million tiny air sacs called alveoli. The walls of the alveoli, to allow the transfer of gases between air and blood, are extremely thin and surrounded by fine mesh of small blood vessels called capillaries. In the alveoli, oxygen molecules are extracted from the air we inhale and infused into the bloodstream while carbon dioxide molecules are expelled. In addition to oxygen and carbon dioxide, other molecules transfer back and forth between the air in the alveoli and the blood – including alcohol molecules.
This process of transferring molecules between the air and the blood through the alveoli walls is called equilibration. Equilibration simply means to bring or keep in equilibrium – to be unchanging. A system that is in equilibrium shows no tendency to alter over time and is a stable or balanced system. In addition to oxygen and carbon dioxide, inert gases such as alcohol equilibrate between the blood and air very rapidly through the thin walls of the alveoli. Each breath, as it travels up and down the lungs, we now know involves many numerous transfers of fluid and substances seeking equilibration.
Law enforcement experts argue that due to equilibration, end-exhaled breath samples accurately reflect the alcohol content of the alveolar (deep lung) air which is in equilibrium with the blood in the body. This is based on the assumption that the first air expelled from the lungs would be air from the airways (bronchi and trachea) and that when a “plateau” in the breath alcohol curve is obtained only the air in the alveoli remains in the breath testing machine. It has further been assumed that as this “alveolar air” passed through the airways the alcohol concentration would remain unchanged.
Since 1950, law enforcement agencies have relied on two faulty assumptions in creating breath testing programs. First, an erroneous reliance on Harger’s assertion that human breath temperature averages 34 degrees Celsius, and second, an erroneous reliance on the belief that equilibration occurs only with the “alveolar air” and not throughout the lungs. However, contemporary studies have conclusively established that breath temperature is not fixed and that Harger’s average of 34 degrees Celsius is off by at least 1 degree Celsius. Modern studies have also established that it is virtually impossible to obtain a sample of pure alveoli air and that equilibration occurs throughout the respiratory system.[10]
By relying on outdated and faulty science, law enforcement experts create breath testing programs with intrinsic error rates. For example, the failure of law enforcement to accurately determine a driver’s breath temperature at the time of testing will result in an error rate of at least a 6.5% and possibly up to 37.7%.[11]
While earlier studies debunked the myth of an average breath temperature of 34 degree Celsius, in 1995 Schoknecht and Stock conducted the most contemporary and comprehensive experiment on breath temperature to date. Enlisting more than 700 subjects over a 15 month period, the Schoknecht study conclusively established that the average human breath temperature swings from a low of 30.9 degrees Celsius to a high of 36.7 degrees Celsius and actually averages 35 degrees Celsius. In comparison, Harger’s original 1950 study enlisted only six individual subjects.
While seemingly small, these differences in breath temperature matter. In a 1962 study, B.M. Wright pointed out that variations of up to 25% between breath alcohol reading and actual blood alcohol content were explained, in part, by variations in mouth temperature and not by errors in blood analysis.[12] Wright’s 1962 study was supported in 1987 and again in 1989 by Fox and Hayward who reported 23% breath to blood correlation error.[13] Even Harger, in his original 1950 study, concluded that the variation in breath temperature between 31 and 35 degrees Celsius could result in a significant variation in the blood alcohol level as calculated by law enforcement through a breath alcohol measurement.
Likewise, the ambient air temperature outside of the body can affect breath temperature and thus a breath test reading. In his 1995 study, Schoknecht found that the changes to average breath temperature corresponded with changes in the daily ambient air temperature. In the following graph, the upper line represents the average breath temperature and the lower line represents the average daily ambient air temperature.
In addition to the erroneous reliance on an average breath temperature of 34 degrees Celsius, law enforcement experts rely on an assumption that equilibration occurs only in the alveoli and that this equilibrium is unchanged during respiration. Not only is this assumption incorrect, but equilibration is also affected by changes that occur to breath temperature during respiration.
The lung’s numerous airways are lined with mucus which varies significantly in temperature. Experts have noted a variance from approximately 23 degrees Celsius at the mouth during inhalation to 37 degrees Celsius in the very smallest airways.[14] These temperatures also vary according to breathing pattern and will rise and fall with inhalation and exhalation. Due to equilibration and Henry’s Law, depending on the temperatures of the air and saliva, alcohol molecules will transfer between saliva and air all up and down the respiratory system. Wright’s study in 1962 established that final equilibration occurs in the mouth and that an increase in temperature from 31 degrees Celsius to 35 degrees Celsius would result in an approximately 24% increase in a reading of breath alcohol content.
The assumption that equilibration occurs only in the deep lungs and that during respiration air is simply forced out of the airways without any interaction with the surfaces of those airways has been discredited. In a 1987 study, Gaylarde found that it is virtually impossible to obtain an uncontaminated sample of alveolar air for breath analysis.[15] Gaylarde further found that rinsing the mouth with warm and cold temperatures significantly affected breath alcohol readings. Because the breath alcohol reading reductions were greater after rinsing with water at lower temperatures, Gaylarde concluded that the dilution of the ethanol in the saliva and reduction in breath temperature caused lower breath test readings.
Law enforcement’s reliance on faulty premises in breath testing programs creates opportunities for you and your client during trial. It is not necessary for you to prove what your client’s breath temperature was or what the actual breath reading should have been. It is only necessary to establish that with breath temperature an unknown variable in your client’s case, there is a scientifically supported error rate for the breath test reading. Conservatively, law enforcement experts must concede at a minimum an error rate of 6.5%.[16] We recommend, prior to trial, providing the government’s expert witness with the applicable studies and articles you plan to rely on. If the expert witness fails to read these materials, you can still use the articles to get the relevant information before the jury.
In trial, you will first need to establish that a conservative error rate in breath test readings is 6.5% based on variations in breath temperature. After presenting the jury with this admission by law enforcement experts, you can proceed to make two additional points: (1) could the error rate be greater than 6.5% in your client’s case, and (2) what is the total range of potential temperature fluctuation in human breath. Due to the potential range in breath temperature from 30.9 degrees Celsius to 36.7 degrees Celsius,
an argument could be made for a 37.7% possible error rate (5.8 degrees Celsius at 6.5% error per degree) in your client’s breath test.
In The Emporer’s New Clothes, it took a small child to challenge the system and point out the Emperor was naked. In a DUI case, it takes a defense attorney willing to learn some basic scientific and biological principles to attack the assumptions relied upon by law enforcement experts. Harger himself, in his 1950 study, acknowledged the uncertainty of breath testing, stating “while the precision is less than is required in research, the results are sufficiently accurate for the routine testing of automobile drivers and industrial workers.”[17] This dismissive attitude helps explain how generations of law enforcement officials erroneously held onto just one fact from Harger’s study and built a faulty breath testing system upon it. Now empowered by the National Academies of Science and the recent discussion by Dr. Hlastala in the Journal of Forensic Science should embolden defense attorneys to successfully challenge toxicologists.
Thank you Patricia Fulton for her contributions to this article.
[1] Saks, M., Faigman, D., Failed Forensics: How Forensic Science Lost Its Way and How It Might Yet Find It. Annu. Rev. Law. Soc. Sci. 2008. 4:149-171.
[2] Strengthing Forensic science in the United States: A Path Forward 2009, National Academies Press
[3] Strengthing Forensic science in the United States: A Path Forward 2009, National Academies Press; Saks, Michael, Faigman, David, Annu. Rev. Law. Soc. Sci. 2008. 4:149-171; Paradigm Shift for the Alcohol Breath Test, Hlastala, Michael, J Forensic Sci, March 2010, Vol. 55, No. 2.
[4] Strengthening Forensic Science in the United States: A Path Forward 2009, National Academies Press. P. 87
[5] Strengthening Forensic Science in the United States: A Path Forward 2009, National Academies Press. P. 23
[6] Strengthening Forensic Science in the United States: A Path Forward 2009, National Academies Press. P. 23
[7] Paradigm Shift for the Alcohol Breath Test, Hlastala, Michael, J Forensic Sci, March 2010, Vol. 55, No. 2. P. 452.
[8] Harger, R.N. (1950). Estimation of the Level of Blood Alcohol from Analysis of Breath. Journal of Laboratory and Clinical Medicine. 36, 306-318.
[9] Jones, A.W. (1982). Effect of Temperature and Humidity of Inhaled Air on the Concentration of Ethanol in a Man’s Exhaled Breath. Clinical Science. 63(5), 441-445; Wright, B.M. (1962). The Effect of Mouth Temperature on Breath Alcohol Concentration.
Journal of Physiology (London). 163, 21-22.; Dubowski, Kurt. (1975). Studies in Breath-Alcohol Analysis: Biological Factors. International Journal of Legal Medicine. 73(93), 93-117; Mason, M.F. and Dubowski, Kurt. (1976). Breath-Alcohol Analysis: Uses, Methods, and Some Forensic Problems–Review and Opinion. Journal of Forensic Science. 21(1), January, 9-41; Carpenter, Dale & Buttram, James. (1998). Breath Temperature: an Alabama Perspective. International Association for Chemical Testing Newsletter. 9(2), July, 16.
[10] Hlastala, Michael and Anderson, Joseph. (2006). The Impact of Breathing Patern and Lung Size on the Alcohol Breath Test. Annals of Biomedical Engineering. DOI: 10.1007/s10439-006-9216-3; Gaylarde, Stambuk, and Morgan. (1987). Reductions in Breath Ethanol Readings in Normal Male Volunteers Following Mouth Rinsing With Water at Differing Temperatures. Alcohol and Alcoholism. 22(2), 13-116
[11] Schoknecht, G. and Stock, B. (1995). The Technical Concept for Evidential Breath Testing in Germany. Proceedings of the International Council on Alcohol, Drugs and Traffic Safety, Proceedings of the ICADTS, T-95, Adelaide Australia..
[12] Wright, B.M. (1962). The Effect of Mouth Temperature on Breath Alcohol Concentration. Journal of Physiology (London). 163, 21-22.
[13] Fox, Glyn PhD and Hayward, John PhD. (1989). Effect of Hyperthermia on Breath-Alcohol Analysis. Journal of Forensic Sciences. 34(4), 836-341; Fox, Glyn PhD and Hayward, John PhD. (1987). Effect of Hypothermia on Breath-Alcohol Analysis. Journal of Forensics Sciences. 32(2), 320-325
[14] Anderson, Joseph and Hlastala, Michael. (2007). Breath Tests and Airway Gas Exchange. Pulmonary Pharmacology & Therapeutics. 20, 112-117.
[15] Gaylarde, Peter, et al. (1987). Reductions in Breath Ethanol Readings in Normal Male Volunteers Following Mouth Rinsing with Water at Differing Temperatures. Alcohol and Alcoholism. 22(2), 113-116.
[16] Jones, A.W. (1982). Effect of Temperature and Humidity of Inhaled Air on the Concentration of Ethanol in a Man’s Exhaled Breath. Clinical Science. 63(5), 441-445; Mason, M.F and Dubowski, Kurt. (1974). Alcohol, Traffic, and Chemical Testing in the United States: a Resume and Some Remaining Problems. Clinical Chemistry. 20(2), 9-41; Dubowski, Kurt. (1975). Studies in Breath-Alcohol Analysis: Biological Factors. International Journal
of Legal Medicine. 73(93), 93-117.
[17] Harger, R.N. (1950). Estimation of the Level of Blood Alcohol from Analysis of Breath. Journal of Laboratory and Clinical Medicine. 36(2), 306-318.