May, 2014 - Can Michigan Ever Derive a Valid Uncertainty Budget for Breath Testing?

 There have been some interesting developments in the area of measurement uncertainty as it relates to breath and blood testing in Michigan.  This includes the suppression of both breath and blood test results for failure to provide a valid uncertainty budget,1 as well as the development of what is ostensibly a valid uncertainty budget for both breath and blood testing.  What does all this mean to the criminal defense practitioner?  Specifically, if the Michigan Toxicology lab has promulgated an uncertainty budget for blood tests, and the breath test unit of the Michigan State Police has promulgated one for breath testing, is uncertainty now a dead issue in Michigan?

 The “uncertainty defense” is most assuredly still alive and well in Michigan. It is probably worth mentioning though, that uncertainty is not really a defense as such, but said more properly: a given range of values or “uncertainty budget” is a necessary foundational condition precedent to the admission of a breath or blood test in Michigan.  Thus, when a report from the toxicology lab or breath test evidence ticket is offered for admission, it is up to the defense practitioner to help the court determine if an appropriate foundation relative to this range of values has been laid. The important question implicit in all of this is: was the proffered uncertainty budget properly calculated?  In other words, how did the State arrive at the values they reported?

Background and Uncertainty Basics

 While the science of measurement uncertainty – known as metrology – has been around nearly as long as the process of measurement itself, tracing the current legal issue back to its nidus inevitably leads to ISO/IEC 17025.  The ISO is the International Organization for Standardization and with the IEC (the International Electrotechnical Commission) form the specialized system for worldwide standardization.  National bodies that are members of ISO or IEC participate in the development of International Standards through technical committees established by the respective organization to deal with particular fields of technical activity.

 As it relates to criminal law, and more specifically to forensic labs providing evidence for law enforcement, there was renewed interest in the application of ISO/IEC 17025 with the publication of Strengthening Forensic Science in the United States; A Path Forward in 2009.2  Compliance with ISO/IEC 17025 is now required for laboratory accreditation.

 So what exactly is measurement uncertainty?  One definition states that uncertainty is “a parameter associated with the result of a measurement, that characterizes the dispersion of the values that could reasonably be attributed to the measurand”3 where the “measurand” is the quantity being measured.   This definition of uncertainty focuses on the range of values that the analyst believes could reasonably be attributed to the measurand.4

 Stated somewhat differently, measurement uncertainty is “an expression of the fact that, for a given measurand and a given result of the measurement of it, there is not one value but an infinite number of values dispersed about the result that are consistent with all of the observations and data and one’s knowledge of the physical world and that with varying degrees of credibility can be attributed to the measurand” 5

 While the scientific concept of measurement uncertainty and the commonly used concept of measurement error are not technically the same thing, at a recent lecture nationally renowned toxicologist Dr. A.W. Jones stated that “measurement error is inherent in all analytical methods simply because it is impossible to repeat exactly the sequence of events necessary to make the measurement.  [It is important to] allow (compensate) for uncertainty by making a deduction from the average result from several repeated measurements.”  Furthermore, that “when the results of a chemical test are used as evidence in a criminal prosecution, the degree of uncertainty in the analytical method used must be made known to the judge and jury.” 6

Breath and Blood Test Uncertainty in Michigan

 In the context of Michigan drunk driving cases, breath and blood tests are based on certain scientific principles, and the testing itself is performed using the methodologies given it by a branch of learning known as forensic science.

 Pretty much since its start forensic science has had as its goal the production of evidence for use by the state in the prosecution of crimes.  As a result, truth has never really been the goal of forensic science, at least not “truth” as that word is used in other scientific disciplines. This makes forensic science much different from “pure” science, which has had as its goal a long march, if not towards “truth”, then at least towards a greater understanding of the natural world.

 This difference in the intended end result, i.e., the courtroom vs. natural truth, is why forensic “scientists” have never embraced the science of measurement known as metrology. Measurement science understands that we can never know the absolute truth about what we are measuring, but instead, are simply “guessing” at a measurand’s true value.  The measurand is the quantity being measured, and for purposes of this article is either breath or blood alcohol.

 Unlike the true science of metrology, forensic science on the other hand states boldly that their methods derive a single true value for the thing measured.  The forensic scientist or breath test operator even more boldly testifies that the results reflect an absolute true, which is the farthest thing possible from a guess.  This is why a breath test machine had traditionally printed an evidence ticket that has one or two breath test measurements with no accompanying range of values.  These results are then utilized by the prosecutor as reflecting the absolute truth of the driver’s actual breath or blood alcohol level.

 Perhaps this is why the National Academy of Sciences found that7:

1. “Appropriate scientific standards are widely ignored in forensic laboratories” contributing “to questions about the validity of conclusions.”
2. “Few forensic science methods have developed adequate measures of the accuracy of inferences made by forensic scientists.”
3. “Much forensic evidence…is introduced in criminal trials without any meaningful scientific validation, determination of error rates, or reliability testing.”
4. “[F]orensic scientists themselves often fail to consider or appreciate measurement uncertainty.”

 In the context of breath testing, this means that when we measure someone’s breath alcohol, a true scientist would understand that the result printed on the paper is nothing more than a guess of the true value of the breath alcohol contained in the breath sample being measured.  If breath testing was done in a real laboratory by real scientists, then in order to give the measurement meaning, it would be necessary to include with the measurement the amount of uncertainty, i.e., the range of values applicable to the circumstances.

 A valid uncertainty budget requires that all sorts of variables be considered and controlled.  These include those variables (limitations) specific to the measuring instruments themselves (in Michigan the DataMaster), variables applicable to the methods used to calibrate the device, variables in the solutions used to calibrate the device, and so on. There are also a variety of variables that are applicable to the human beings operating the device (the police officer) and the individual blowing into it (the person arrested).  A good, sound, scientifically valid uncertainty budget would take into consideration all possible areas of variability/uncertainty, and then, using a complex mathematical equation, derive the amount of combined uncertainty in a given measurement derived from a particular device and operated by a particular person or set of persons.

Is it Impossible to Ever Derive an Uncertainty Budget for All of Michigan?

 As it relates to both breath and blood testing in Michigan, the respective divisions of the state police that are in charge of these units are having fits trying to figure out how to comply with ISO/IEC 17025.  So far, their attempts have failed, that is, failed when subjected to appropriate and reasonable scientific and legal scrutiny.

 There are many problems with the current status of the attempts of the Michigan State Police to derive an uncertainty budget for breath tests.  Perhaps first among them is a lack of clarity if not understanding of the quantity to being tested.  Although breath tests always probe the same physical quantity, i.e., the alcohol in a sample of human breath, their measurand is dictated by the statute or statutes applicable to breath testing, and these statutes vary between jurisdictions. Thus, identical numerical values obtained from tests in disparate jurisdictions may refer to different quantities and may not indicate the relevant statutory measurand.  This can lead to misinterpretation of results, referred to as the “measurand problem.”8

 With breath testing in the United States, there are three possible measurands, depending on the jurisdiction.  These include (I) “end respiratory air,” (II) alveolar air, and (III) breath as an indirect measurement of blood.9  The proper evaluation of uncertainty is different for each measurand.10 Michigan is a type II, or alveolar air, jurisdiction.

 To help demonstrate some of the significant limitations to the state of Michigan’s uncertainty budget determination, it may help to look at the “predicate questions and answers” as provided by the breath testing unit of the Michigan State Police to Michigan Prosecutors.  They believe that the answers to these questions will lay a proper foundation so as to allow breath test results to qualify for admission in court.  Yet, none of them address the specific nature of the measurand, that is, alveolar air.11  For example:

How is the measurement uncertainty for the result determined for breath tests?

The Toxicology Lab Division analyzed the factors that contribute to variation in the test results. From this analysis, the state toxicologist approved an algorithm for calculating measurement uncertainty for breath tests. Using this algorithm and the data collected for this breath instrument and the defendant’s breath test, we can compute a 99% confidence interval for the individual’s true mean breath alcohol concentration.

 This proposed answer actually insults the intelligence of the courts because it presupposes that admission of an uncertainty budget (UB) is essentially, res ipsa loquitur; as if the thing speaks for itself.  But, as it relates to any UB, the “thing” most assuredly does not “speak for itself.”

 Looked at a little differently, when it comes to UBs, - the devil is in the details.  Courts should not allow the prosecutors to ask a few magic questions and have the UB admitted, at least not if real science as opposed to forensic science is the guide.

 A question here is by what method did they derive their UB?  Looking at the model answer it may be supposed that the State Police assume most lawyers and judges will see or hear the word “logarithmic” and stop asking questions.

 Here is something to think about as it relates to UBs in Michigan: A scientifically valid uncertainty budget can be derived for a particular lab, including the Michigan State toxicology lab. This is not to suggest that it has been done yet for Michigan, only that it can be done.

 One could not then take this uncertainty budget from the Michigan State toxicology lab and assume or imply that it is applicable to any other forensic lab in any other state, or even another lab in Michigan.  Even a lab that used the exact same protocol!  That is, the same methodology (in testing for example blood) by headspace gas chromatography.

 If this is true, and it is, then the range of values (UB) relative to breath alcohol testing would have to be derived for each DataMaster at each police department.  And, the uncertainty budget applicable to one police department, located (for example) in New Haven, may not apply to any other department, such as one located in Grand Rapids.

 But here is something that makes it even more implausible and/or impossible: As it stands in Michigan, there is no standard protocol for breath testing.  There are administrative rules; however, as Michigan case law clearly suggests, these rules are routinely not followed; and even when it can be shown that the administrative rules for a particular test were violated, the corresponding breath test result nevertheless is almost never suppressed.  In other words, such administrative rule violations are said to “go to weight rather than admissibility.”

 What this means is that even if the state breath testing unit was able to come up with an uncertainty budget applicable to breath testing in every law enforcement jurisdiction and for every DataMaster in Michigan, this UB could really only be applicable if the administrative rules were followed precisely every time.  This would mean we would be back to strict compliance in Michigan. Fat chance of that ever happening!

 Patrick T. Barone is an adjunct professor at Cooley Law School where he teaches "Drunk Driving Law and Practice."  Mr. Barone is also the co-author of two books on DUI-related issues, including Defending Drinking Drivers (James Publishing), a well-known and highly respected multi-volume national legal treatise.  He is a frequent lecturer on trial practice and drunk driving defense tactics. He can be contacted on the web at: www.baronedefense


1. People v. Jabrocki, 79th DC # 08-5461 OD and People v. Standhardt; file number D-130733-FD (63rd District Court).
2. National Research Council, Strengthening Forensic Science in the United States: A Path Forward (Washington, DC: The National Academies Press, 2009).
3. Ellison, Williams, Quantifying Uncertainty in Analytical Measurement, EURACHEM / CITAC Guide CG 4, Third Edition, pg. 1, 2012.
4. Id. at pg 2.
5. Vosk, The Measurand Problem in Breath Alcohol Testing, J Forensic Sci. 2014 May; 59(3):811-5.
6. Mastering Scientific Evidence, New Orleans, March 20-22, 2014.
7. See, Vosk Forensic Metrology: A Primer for Lawyers, Judges and Forensic Scientists (2009), and NAS, Strengthening Forensic Science in the United States: A Path Forward, (2009) for internal citations.
8. Vosk, The Measurand Problem in Breath Alcohol Testing, J Forensic Sci. 2014 May; 59(3):811-5.
9. Id. at 812.
10. Id. at 812, et. seq.
11. Admin. R. 325.2651   “Definitions.”