EVIDENCE LIBRARY
Measurement Uncertainty Near Forensic Toxicology Cutoffs
How measurement uncertainty should be evaluated near forensic toxicology cutoffs, reporting thresholds, and quantitative decision points under ANSI/ASB 056.
Quick answer: a quantitative toxicology result is an estimate, not an exact point. When a reported concentration is close to a cutoff, reporting threshold, statutory boundary, or other decision point, the laboratory’s measurement uncertainty can become important because the plausible range around the measured value may overlap the boundary. Measurement uncertainty does not automatically invalidate the result, but it should be understood before a close numerical distinction is treated as scientifically decisive.
This issue became more concrete with ANSI/ASB Standard 056 (2025), Standard for Evaluation of Measurement Uncertainty in Forensic Toxicology. The standard establishes minimum requirements for evaluating measurement uncertainty in quantitative forensic toxicology testing and in calibration of breath-alcohol measuring instruments. It applies across postmortem toxicology, human-performance toxicology, non-regulated employment testing, court-ordered toxicology, and general forensic toxicology. It does not address uncertainty for breath-alcohol subject testing or performance measures for qualitative toxicology.
What measurement uncertainty means
Measurement uncertainty describes the dispersion of values that could reasonably be attributed to the measurand under the stated measurement conditions. It is not the same thing as a mistake, a failed quality-control event, or an admission that the laboratory does not know the answer. Instead, it is part of reporting a quantitative measurement honestly: the analytical process has finite precision, reference materials and volumetric operations carry uncertainty, calibration has uncertainty, and repeated measurements vary.
NIST emphasizes that a measurement result should be understood together with its associated uncertainty and that metrological traceability requires an unbroken calibration chain in which each link contributes to uncertainty. That is why metrological traceability and measurement uncertainty are related but not interchangeable. A result can be traceable and still have uncertainty large enough to matter for the decision being made.
Why cutoffs and decision thresholds require special care
A cutoff is a decision rule, not a physical wall in the specimen. If a method reports a concentration close to a cutoff, the scientific question is not merely whether the displayed number falls one digit above or below the boundary. The reviewer should also ask how the laboratory evaluated uncertainty, how the result was rounded, whether replicate measurements were used, and what decision rule governs results near the boundary.
The same principle applies to laboratory reporting thresholds and to interpretive concentration boundaries. A concentration near the lower end of a method’s validated range should be considered together with limits of detection and quantitation, calibration performance, and the method’s validated precision and bias. Measurement uncertainty does not convert an analytical limit into a clinical or legal limit, and it does not make concentrations on opposite sides of a boundary scientifically identical. It does, however, prevent false precision when the numerical distinction is smaller than the measurement process can support.
What belongs in an uncertainty evaluation
The relevant contributors depend on the method and workflow. They can include repeatability or reproducibility, calibrator and certified-reference-material uncertainty, volumetric operations, preparation steps, calibration-model behavior, bias, and other method-specific sources. Some uncertainty components can be estimated statistically from repeated observations; others come from certificates, calibration records, or other documented information.
A 2025 Journal of Analytical Toxicology paper by Kahl and Moore described implementation of ANSI/ASB 056 in a postmortem toxicology laboratory using the NIST eight-step approach. Their example illustrates why an uncertainty budget should be tied to the actual analyte, extraction, equipment, preparation, and analytical technique rather than treated as a generic laboratory-wide percentage. The paper also demonstrates how historical quality-control data can capture multiple real-world sources of variability while other components are evaluated from documentation such as certificates of analysis and calibration certificates.
Uncertainty is not the same as method validation
A method can have a calculated uncertainty budget and still be unsuitable for a particular use if the method was not properly validated for the analyte, matrix, range, or intended purpose. Conversely, successful method validation does not eliminate measurement uncertainty. Validation asks whether the method performs acceptably for its intended use; uncertainty characterizes the dispersion associated with a quantitative result produced by that measurement process.
For that reason, a case review should not isolate the uncertainty percentage from the underlying validation. Precision, bias, calibration, matrix effects, dilution integrity, carryover, stability, and the analytical measurement range all influence whether the result is fit for the forensic question being asked.
What uncertainty does not prove
Measurement uncertainty addresses uncertainty in the quantitative measurement. It does not by itself answer pharmacological or causation questions. It cannot establish dose, time of ingestion, route, tolerance, impairment, toxicity, intent, or cause of death. Those are separate interpretive questions requiring appropriate pharmacokinetic, clinical, behavioral, pathological, and case evidence.
It also should not be used as an automatic subtraction from every reported concentration. A laboratory’s expanded uncertainty is not a universal correction factor. Whether and how uncertainty affects a decision depends on the measurement model, the stated coverage, the applicable decision rule, and the question being answered.
Records to review when a result is near a boundary
Start with the laboratory’s measurement-uncertainty procedure and the uncertainty budget applicable to the specific analyte, matrix, extraction, and analytical method. Then examine the validation study, calibrator and quality-control records, certificates for reference materials, pipette and volumetric calibration records when relevant, analytical sequence, replicate results, dilution records, calculations, rounding rules, and the laboratory’s written decision rule for results near a reporting or legal threshold.
The case-specific question is whether the uncertainty evaluation actually corresponds to the way this specimen was measured. A generic uncertainty statement can be less informative if the laboratory used a different matrix, dilution, standard-addition workflow, preparation scheme, or instrument configuration than the assumptions built into the budget.
How to read a reported result with expanded uncertainty
When a laboratory reports a measured concentration together with expanded uncertainty, the uncertainty statement should identify or permit evaluation of the coverage level or coverage factor used. The value should not be described as an absolute interval containing the “true” concentration with certainty. It is an uncertainty statement produced under a defined measurement model and stated assumptions.
The most defensible interpretation is therefore conditional: identify the measured value, identify the uncertainty evaluation and its coverage, identify the decision boundary, and then state whether the measurement supports a robust distinction from that boundary. The closer the result is to a consequential threshold, the less appropriate it is to ignore the laboratory’s own uncertainty framework.
Relationship to calibration and reference materials
Uncertainty begins upstream of the final instrument readout. Certified reference materials carry assigned values and associated uncertainty, but purchasing a CRM does not automatically make every downstream case result exact or traceable. The laboratory must preserve the measurement chain through preparation, calibration, extraction, analysis, data processing, and quality control.
Likewise, the choice of calibration model and weighting can affect low- and high-end performance. A visually impressive calibration curve or high coefficient of determination is not a substitute for demonstrating acceptable residual behavior, bias, precision, and fit across the validated range.
Practical takeaway
Near an important quantitative boundary, the scientifically relevant question is not simply “What number did the instrument print?” It is “What quantity did the laboratory measure, how was the measurement system validated and calibrated, what uncertainty is associated with that result, and what decision rule applies at this boundary?” That approach preserves the difference between analytical measurement and legal or medical interpretation while reducing the risk of treating a close numerical result as more exact than the method can support.
Sources
NIST: Metrological Traceability — Frequently Asked Questions and Policy.