Uncertainty comes to Michigan

The horseless carriage; rock and roll; DNA evidence: like many historic developments, the uncertainty budget is here to stay. Like many new developments, the uncertainty budget has its critics. The Michigan State Police (MSP) seems to be uncomfortable with it. However, it appears that MSP may be starting to embrace the sources of uncertainty in flame ionization headspace gas chromatography, the fairly simple process of using a flame to ionize a gaseous mix that contains what is suspected to be beverage alcohol in blood from a suspect in an OWI-DUI-based charge.

The MSP joined the ranks of those trying to estimate the uncertainty in its measurement just in the nick of time. In January, 2011 Dr. Michelle Glinn, acting head of the day-to-day operations of the forensic toxicology unit, authored a memorandum to inspector Greg Michaud. Her memo outlined her understanding of measurement uncertainty and a protocol for the MSP to follow in determining uncertainty in forensic analysis.

Measurement uncertainty is one overarching principle of metrology, the science of measuring things. There are three 3 basic concepts: 1) measuring something 100 different times gets 100 different results; 2) The more data collected and analyzed the more confident you are that you understand the uncertainty's source and that the result is a known range containing the true value; and 3) you won't know with 100 percent confidence the true value. Once you have tried to determine all the sources of uncertainty in your measurement and understand the impact each has on your overall measurement process then you can build an "uncertainty budget" or an estimate of the range of values within which the true value may fall.

The rift on the horizon is between the scientists who develop or study and understand measurement uncertainty and the three concepts above. This community follows the Guide to Uncertainty Measurement (GUM) developed by the International Bureau of Weights and Measures (BIPM ). BIPM GUM is the protocol on which all others are predicated including protocols developed by the National Institute for Standards and Technology (NIST) and the ASCLAD-LAB. That is important because ASCLAD-LAB is the agency, which "accredits" member crime labs in America including the Michigan State Police crime lab. MSP is scheduled for re-accreditation in 2012.

The ASCLAD-LAB accreditation is the peg on which a state toxicologist hangs her hat when testifying about the reliability of the crime lab. The state toxicologists are taking the position that the uncertainty doesn't need to be reported with measurement results such as blood alcohol concentration. The uncertainty is not currently being reported in the MSP results.

The state toxicologists also appear to be taking a "simplified" approach: using only six sources of uncertainty in their uncertainty budgets without regard to the overall impact of each on the analysis. For example, variance in the calibration solution is given as much weight as the variance in the amount of the internal standard. The two will have varying impact on the overall budget based strictly on the amount of variance in each one according to the vendor that provided them.

Judge Peter J. Wadel of the 79th judicial district court heard Michigan's first uncertainty case. Judge Wadel ruled that MSP must report the uncertainty in order for the blood alcohol concentrations to be admitted in his court. Many expected this ruling on May 6, 2011, to be overturned on appeal. However, instead of appealing, the prosecutor produced a result with an uncertainty budget, using MSP's uncertainty protocol as of January 2011.

The ASCLAD-LAB requirements now hold that every toxicology lab must have an uncertainty budget in order to achieve re-accreditation. MSP will need an uncertainty budget that is acceptable to ASCLAD-LAB's inspectors. The ASCLAD-LAB protocol does NOT require the lab to report the uncertainty with the measurement unless the "customer" requires it. Both Dr. Glinn and her predecessor, Dr. Felix Adatsi, have testified previously that the customer is the police agency that sends the measurand (the thing to be measured) to be analyzed.

The MSP unveiled its uncertainty budget to Judge Wadel asking him to allow its 2008 measurement into evidence after his ruling. On September 28th, the court issued an order ruling that the "customer" is the courts, the prosecutor and the defense attorney not the police agency. Further, the MSP lab report should include the measurement uncertainty. He rejected the argument that the MSP lab "benefits the defendant" when it simply reports the lower of two results from two analyses so long as the 1st and 2d agree within .01 grams of ethanol/100 milliliters of blood.

It is too late for uncertainty to save the criminal records of all those accused citizens, whose blood alcohol content may have been below the legal threshold. For all those whose alleged blood sits in the MSP crime lab now and may be under the legal threshold when the uncertainty is reported to a jury, it is just in the nick of time.


Michael J. Nichols, of The Nichols Law Firm PLLC, focuses his practice exclusively on complex OWI/OWID cases and other select criminal and other litigation matters. He is the author of the "Michigan OWI Handbook" published by West, chairs the Ingham County Bar Association Criminal Law section and is a member of the National College of DUI Defense, the National Association of Criminal Defense Lawyers, The Criminal Defense Attorneys of Michigan and the State Bar and Ingham County Bar Association's Criminal Law Sections. He can be reached at 517-432-9000; mnichols@nicholslaw.net; or www.michiganduidefender.com.

Published: Thu, Nov 3, 2011