Blind Spots in Blood Alcohol Testing: An Interview with Kevin Schug

As part of “From Sample to Verdict,” we interviewed Kevin Schug of the University of Texas at Arlington to discuss the current progress of limitations in forensic blood alcohol testing and how chromatography is being used to address these challenges (1,2).

Your LCGC Blog back in 2023 (2) highlighted widespread deficiencies in method validation and quality control in forensic blood alcohol testing. What are the most common validation oversights you've encountered across the 150+ litigation discovery packets you've reviewed?

We are now in 2025, and the number of cases our firm has reviewed has exceeded 300. We still see many of the same deficiencies in forensic blood alcohol testing that were apparent two years ago. Regarding method validation, many laboratories still rely on validations performed on an instrument many years prior, and those validations were not performed in accordance with current industry standards. We see a lack of method re-validation following major instrument maintenance. Method validations and routine quality control procedures often lack the appropriate use of blank and ethanol-fortified whole controls. We see an underassessment of uncertainty levels reported for measurements that are commonly relied upon in litigation. For many laboratories, there is a lack of rigor in following the scientific standards expected by the analytical chemistry measurement science community when it comes to performing quantitative measurements from a complex biological fluid.

You raise serious concerns about laboratories applying a fixed uncertainty value (4.3% at 99.7% confidence) across all instruments and locations. What would a more scientifically rigorous uncertainty assessment process look like in practice for blood alcohol determinations?

The application of a fixed uncertainty value across multiple instruments, of different make and model, in different locations, does not make sense.Each instrument has its operational history, and each instrument will eventually exhibit decreased performance and need to be maintained or repaired. Uncertainty would be better determined on a batch-by-batch basis on individual instruments. These methods involve the preparation and measurement of a compound from human blood, a complex matrix. A more reasonable way to assess uncertainty and bias would be to incorporate sufficient replicates of ethanol-fortified whole blood controls at different concentration levels through a batch sequence run. The variability of these matrix-matched control responses could be used to assign a more realistic uncertainty value to the results determined in that batch. Even better would be to analyze multiple prepared samples from multiple separate blood tubes collected for any individual being tested.

Many laboratories rely on calibration using neat aqueous standards and manufacturer-provided uncertainties. Why is this insufficient when measuring ethanol in whole blood, and what role do biological matrix effects play in creating additional variability?

The old saying that blood is thicker than water applies here. Blood is a much more complex sample matrix than water. The most rigorous BAC determination would use human whole blood for calibration and quality control. Matrix-matched calibration is a best practice when analyzing biological fluids. To reduce cost, laboratories are allowed to use aqueous-based calibrators if the method has been validated and is quality controlled using whole blood controls. Many laboratories only use aqueous calibrators and aqueous controls.Operating this way, these laboratory tests are blind to interferences and matrix effects that can bias a BAC measurement made from whole blood samples. Regarding uncertainty, many laboratories also do not account for the extra variability that arises from preparing and analyzing whole blood samples. Incredibly, these labs ascribe the majority of their uncertainty to measurements of aqueous controls and manufacturer-stated uncertainties on certified reference materials. The laboratories that do this are underassessing the uncertainty of their measurement, because they do not account for the variability associated with analyzing whole blood.

You mentioned that some laboratories justify not re-evaluating matrix effects because a study was conducted on different instruments nearly a decade ago. What are the dangers of extrapolating old performance data across diverse instrumentation and laboratory environments?

No instrument is the same. Even if it is the same make and model from a given manufacturer, each instrument is individually assembled and individually maintained. The performance of a method on one instrument cannot be used to prove the performance of a method on another instrument. Importantly, a method will not continue to perform perfectly in perpetuity. For example, when major maintenance is performed on an instrument, human hands have handled and changed parts. The assembly of a headspace gas chromatograph is very precise. For example, columns must be installed into the injection port with a tolerance of less than 1 mm in their placement. It is just proper science to rigorously assess whether a change in hardware has caused some matrix effect to occur, perhaps because of some small misalignment. As I said before, the failure to perform interference and accuracy tests using whole blood controls leaves a blind spot where the presence or absence of such effects cannot be proven.

Sample preparation is cited as a major source of error. In your experience, how do differences in analyst technique, pipetting (measured sample volumes), and sample viscosity contribute to variability, and how should laboratories account for this in uncertainty estimates?

I do not agree with the practice of setting a defined uncertainty across an entire set of instruments and laboratories. To claim that uncertainty is static and will not change over time is naïve.If this practice is to be continued, then it needs to explicitly involve the measurement of uncertainty from whole blood samples. Even pipetting blood samples will have greater uncertainty and vary more among different analysts than pipetting aqueous standards. It would be better if accuracy and precision were evaluated with each batch, but this would require greater replication and better coverage of the linear range using whole blood controls than is typically encountered in forensics labs.

Your findings suggest that forensic laboratories often fail to include quality control measures like regularly analyzing fortified whole blood controls. What are the consequences of omitting these checks in routine forensic workflows, especially in legal contexts?

The problem is that if the laboratory fails to include regular measurement of whole blood controls, both positive and negative, then there is limited information available to assess the potential for interferences and matrix effects biasing a BAC measurement. The laboratory can claim to not have issues, but there is no proof one way or the other. This is a situation that could easily be avoided if the proper controls are in place. However, most laboratories are not legally forced to comply with industry consensus standards and thus, their method validations and quality control often fall short of the minimum requirements for proper blood alcohol determination.

Considering the legal weight of blood alcohol results in Driving Under the Influence (DUI) cases, what reforms would you recommend to ensure forensic laboratories provide more defensible and scientifically sound data in court?

My primary recommendation would be that forensics laboratories should be legally bound to apply the procedures outlined in industry consensus guidelines for method validation and quality control, among other guidance documents, set forth by the American National Standards Institute (ANSI) and the American Academy of Forensic Sciences Standards Board (ASB).

References

1. Medusa Analytical, Scientific Consulting & Expert Witnesses. Medusa Analytical. Available at: https://www.medusaanalytical.com/ (accessed 2025-07-14).
2. Schug, K.; Hildenbrand, Z. L. The LCGC Blog: Forensics Laboratories Underassess Uncertainty in Blood Alcohol Determinations. LCGC North America. Available at: https://www.chromatographyonline.com/view/the-lcgc-blog-forensics-laboratories-underassess-uncertainty-in-blood-alcohol-determinations (accessed 2025-07-14).