Carl Ward is one of the DWI lawyers in Missouri trying to make the system work properly and he had this interesting perspective:
In any given year, around 35,000 Missourian’s are arrested and charged with DWI. Nationally, the number is 1.5 million. While drunk driving continues to be a serious problem, Missouri needs to take immediate steps to improve its breath alcohol testing program to insure that only the guilty are convicted of what has become a serious crime with very serious consequences. The most respected experts in the field will tell you that Missouri’s breath alcohol testing program lacks scientific reliability and is one of the very worst in the Country. And why is that, you ask?
First, when you are arrested for DWI in Missouri, you will be given a single breath test. However, good scientific protocol absolutely REQUIRES that a driver be administered two breath alcohol tests, not just one. Since 1986, the National Safety Council Committee on Alcohol & Other Drugs Subcommittee on Technology has recommended that at least two “separate breath tests should be collected and analyzed individually in performing any evidential breath-alcohol analysis.” They further recommended that the “breath samples should be collected at intervals of not less than 2 nor more than 10 minutes, and the test results must agree within 0.02 g/210 L.”
Currently, thirty-two states and the District of Columbia require two tests. I predict we will be the last to do so…at least so long as the Governor continues to appoint prosecuting attorneys instead of doctors or research scientists to head Missouri’s Department of Health. [Bet you didn’t know that, did you?]
The reason for requiring two breath tests is that our breathalyzers are susceptible to interference from mouth alcohol contamination and other random errors. While the infrared breath alcohol testing devices used in Missouri are supposed to be able to detect the presence of mouth alcohol and other interfering substances, one study has shown that this mechanism failed to detect mouth alcohol as much as 48 percent of the time in a controlled laboratory setting.
A leading expert on breath alcohol testing has said that “(b)reath testing for alcohol using a single test method should not be used for scientific, medical or legal purposes where accuracy is important.” Of course, in Missouri, the jurors are instructed that if the driver’s BAC is .08% or more, it is prima facie evidence that the driver was intoxicated. In short, the test result is critically important if you are the one facing losing your driver’s license or prison based on a single number.
Second, good science also requires that the breathalyzer being used to test an unknown (in this case, a driver’s breath alcohol level) be checked at the time of the driver’s test to verify that the instrument is measuring accurately. This is easily accomplished by hooking up a simulator to test the breathalyzer against a known .04, .08, or .10 alcohol solution or gas mixture. Thirty-seven states and the District of Columbia require this. Missouri, along with ten other states, only require that a controlled or known sample be used when the breathalyzer is maintained. In Missouri, breathalyzers are maintained only every thirty-five days, another problem which needs to be addressed immediately.
Even Brian Lutmer, the only scientist for Missouri’s breath alcohol testing program, agrees that running two tests and a control sample at the time of testing are essential to a scientifically reliable breath testing program. These recommendations were made by him years ago and have fallen on the deaf ears of every Director of Health since then. Mr. Lutmer wrote:
“The Missouri Department of Health and Senior Services Alcohol Program, as the unit with the statutory authority for the rules and regulations used within the state for the evidential testing of individuals for the purpose of DWI prosecution, needs to be committed to the highest standards of personal and professional integrity. It is the Breath Alcohol Program’s ethical responsibility to the people of this state to guarantee that all tests performed are of the highest standard as to ensure that persons charged with drinking and driving crimes, especially as they related to per se laws, are both not being charged when testing is not adequate to make sure they are indeed over the legal limit, and that those individuals are able to be easily and successfully prosecuted when a nationally and internationally recognized standard for quality assurance is applied to our forensic breath alcohol analysis protocol.
Drunk driving convictions, especially if determined from breath alcohol results, such as in the case of per se statutes, carry significant financial and personal consequences. The possible consequences therefore compel us to ensure that the highest standards of quality control and quality assurance are utilized in an effort to minimize the potential for false positive results. In order to meet these high standards of scientific quality control and professional integrity, and to meet our duties to the people of the state of Missouri, I am suggesting the following changes within our rules and regulations: 1) duplicate breath samples for each subject tested with either a single control standard run before breath samples are provided or control standards run before and after the breath samples are provided; 2) annual re-certification of all simulator thermometers and/or barometric pressure sensors (if we decide to start utilizing dry gas controls rather than or in addition to wet-bath simulator controls), and 3) a single training regiment to be implemented throughout the entire state for certification for all officers wishing to obtain a permit for the purpose of evidential breath testing of DWI suspects.”
The Missouri Court of Appeals has also recognized that there are serious issues with the scientific reliability of Missouri’s breathalyzer program.
“In order to avoid inaccurate results caused by mouth alcohol, most states have adopted protocols that require — after the mandatory observation period — that a suspected intoxicated driver be offered two separate breathalyzer tests. If the results of those two tests vary outside of a recognized parameter, a third confirmatory test is given to determine which of the first two test results is the more reliable and which may have been tainted by mouth alcohol…
…Missouri has failed to adopt many of the protocols that assist in ensuring the scientific reliability or credibility of the test results. The Director argues in this case that the test results obtained following the Missouri testing protocols should be subject to a presumption of validity. The Majority accurately points out that this argument fails based on the statute, but the State’s argument also fails based on the science. I write separately to suggest that it may be time for the Missouri breath alcohol testing program protocols to be updated to comply with the standards in the industry. As the National Safety Council stated, ‘The significant weight assigned to breath alcohol results, along with the serious consequences arising from conviction on an impaired driving offense require evidential breath alcohol testing programs to implement appropriate quality assurance measures.’ Id. In the meantime, the reliability of and weight to be given to breathalyzer test results in Missouri clearly remains an issue for the finder of fact.”
Even with improvements to the breath alcohol testing protocols, “Breath testing, as currently used, is a very inaccurate method for measuring BAC. Even if the breath testing instrument is working perfectly, physiological variables prevent any reasonable accuracy.”
For example, breath test instruments erroneously assume that 2100mL of breath contains the same amount of alcohol as 1 mL of blood. This is referred to as the 2100:1 partition ratio. However, studies have shown that the partition ratio in fact varies from person to person and even within the same individual from hour to hour, and this variance can produce erroneous test results. In Intoxication Test Evidence 2d (Clark, Boardman, Callaghan, 1995), it was noted that Kurt M. Dubowski and other respected scientists have found the normal range in test populations vary from 1100:1 to 3000:1. “All breath test devices, for example, will report a 0.10% for a subject who has a true BAC of 0.07% if he or she has a partition ratio of 1500:1 (instead of 2100:1), and conversely, a 0.10% for a subject who has a true 0.14%, if he or she has a partition ratio of 3000:1.” Id.
I have personally witnessed dramatic differences between breath alcohol test results and simultaneously drawn blood samples. In particular, in a lab setting, a client of mine consumed four beers on an empty stomach over a 15 minute period and then blew into a BAC Datamaster every five minutes until his breath alcohol test results stopped rising, or “peaked” at 0.088. A blood sample was then taken from the client within a minute of that time. The blood sample came back from the lab a week later at .063. Pretty scary, huh?
My suggestion is that if your breath test result comes back close to, but over, the legal limit, and you don’t think you are drunk, have a friend or family member drive you directly to the hospital emergency room after your release and ask for a blood test for alcohol. If the police won’t release you, tell them you want to exercise your right to an independent test of your blood (something they are not required to tell you about), and that they are interfering with that right. You should also ask to contact your lawyer when you are asked to take a breath test. You are entitled by law to have twenty minutes to do so. Once again, the police are not required to tell you about that right.
In addition to the problem with using an assumed partition ratio of 2100:1, body temperature varies, and this variation can adversely affect the accuracy of your breath alcohol test results. Once again, breath testing instruments work on the assumption that everyone’s breath temperature is 34.0ºC. Several studies have shown this is an erroneous assumption, and that the average breath temperature of individuals is actually higher. As a result, breath alcohol test results are constantly being overestimated.
For example, in 1995, a report on the field testing of drinking drivers in Germany showed that the average breath temperature, as measured on the Draeger 7110, equipped with breath temperature monitoring, was 35ºC. A second field study conducted in Alabama showed an average breath temperature of 34.9ºC using the Draeger 7110 technology.
The result is that for every 1ºC your breath temperature exceeds 34.0ºC, your breath alcohol test result is being inflated by as much as 6.5 percent. Yet, only the State of Alabama uses breath testing equipment that corrects for breath temperature. And God forbid you are running a fever when you are arrested. If you are, you had better ask for a nurse at the jail to take and record your temperature.
Perhaps the next time the Governor nominates a prosecuting attorney (i.e., assistant attorney general) as Director for the Department of Health instead of a medical doctor or a scientist, the Missouri Senate should say enough is enough, and vote “No.” More to follow. Happy motoring.
 Alabama, Arizona, Arkansas, California, Colorado, Connecticut, Florida, Georgia, Idaho, Maine, Maryland, Massachusetts, Minnesota, Mississippi, Montana, Nevada, New Hampshire, New Jersey, New Mexico, North Carolina, North Dakota, Ohio (use of second sample depends on instrument), Oklahoma, Oregon, Pennsylvania, Rhode Island, Texas, Tennessee, Virginia, Washington, Washington DC, Wisconsin, and Wyoming require two tests.
 Simpson, C.D., Kerby, J.A. and Kerby, S.E. Effects of Mouth Alcohol on Breath Alcohol Results. INT’L J. OF DRUG TESTING: Vol. 3(1) (2007)
 Hlastala, Michael P. Physiological Errors Associated with Alcohol Breath Testing. The Champion: 16-19 (July 1985)
 Alabama, Alaska, Arizona, Arkansas, California, Colorado, Connecticut, Florida, Idaho, Iowa, Kansas, Kentucky, Maine, Maryland, Massachusetts, Minnesota, Mississippi, Montana, Nevada, New Hampshire, New Jersey, New Mexico, New York, North Carolina, North Dakota, Ohio (instrument dependent), Oklahoma, Oregon, Pennsylvania, South Carolina, Texas, Vermont, Virginia, Washington, Washington DC, West Virginia, Wisconsin, and Wyoming.
 Lutmer, Brian. Recommendations for Improvements Within the Rules and Regulations By the Breath Alcohol Program for the Purpose of Increasing the Quality Assurance of Evidential Breath Testing Within Missouri.
 Currently, thirty-two states and the District of Columbia have adopted duplicative confirmatory testing: Alabama, Arizona, Arkansas, California, Colorado, Connecticut, Florida, Georgia, Idaho, Maine, Maryland, Massachusetts, Minnesota, Mississippi, Montana, Nevada, New Hampshire, New Jersey, New Mexico, North Carolina, North Dakota, Ohio (use of second sample depends on instrument), Oklahoma, Oregon, Pennsylvania, Rhode Island, Texas, Tennessee, Virginia, Washington, Washington DC, Wisconsin, and Wyoming.
 Collins v. Director of Revenue, 399 S.W.3d 95, 104-106 (Mo. App 2013) (omissions)
 Hlastala, Michael P. Physiological Errors Associated with Alcohol Breath Testing. The Champion: 16-19 (July 1985)
 G. Schocknecht and B. Stock. The Technical Concept for Evidential Breath Testing in Germany. Proceedings of the ICADTS T-95. Adelaide Australia, 1995
 Dale A. Carpenter, PhD and James M. Butram, PhD. Breath Temperature: An Alabama Perspective. IACT Newsletter: 9:2:16-17 (1998)
 AW Jones. Physiological Aspects of Breath-Alcohol Measurement. Alcohol, Drugs, and Driving: 6:1-25 (1990)