Setting Target Measurement Uncertainty

Significance of the topic

Measurement uncertainty (MU) determines whether an analytical result is fit for its intended use. Establishing a target MU—the maximum admissible uncertainty for a specific measurement goal—is essential for making defensible compliance, commercial or regulatory decisions. An appropriately chosen target MU balances the need for protection (for example, public health or product quality) against the cost and effort of obtaining very low uncertainty. The Eurachem/CITAC guidance on setting and using target uncertainty provides a structured approach to define these limits in chemical measurement.

Objectives and overview

This summary explains the rationale for setting target MU, outlines approaches recommended by the Eurachem/CITAC guide, and illustrates the practical consequences of poor target-setting using a fictional case study. It highlights how different uncertainty magnitudes influence decisions and costs, and recommends pragmatic steps for laboratories and stakeholders to establish and use target uncertainties.

Methodology and approaches to setting target measurement uncertainty

• Definition: The target MU is the largest acceptable uncertainty for a particular measurement purpose. It can be expressed as a target standard uncertainty (utg) or an expanded uncertainty (Utg) using a coverage factor (commonly k = 2 for ≈95% confidence).
• Hierarchy of evidence: The Eurachem/CITAC guide recommends using a hierarchy of indicators to derive target MU, ranked from the most robust (statistically based performance requirements, regulatory limits tied to health/safety criteria) to less direct inputs (historical lab performance, consensus practice).
• Practical strategies: Common approaches include deriving target MU from decision limits or action levels, performance criteria from proficiency testing or interlaboratory studies, risk-based thresholds (e.g., health-based guidance values), and cost–benefit analysis that considers measurement expense versus the value of improved decision certainty.
• Implementation: When a customer or regulator does not stipulate a target, the laboratory must define one based on the intended use, documenting the rationale and checking that the lab’s MU meets that target before reporting results as fit-for-purpose.

Illustrative scenario and key findings

The guide’s impact is illustrated by a fictional example: a farmer sells oranges to a juice producer that enforces limits on thiabendazole residues (accept only < 1 mg kg-1) and pays premiums for higher Brix (sweetness) above defined thresholds. Two laboratories analysed the same batch and reported metrologically compatible results but with different uncertainties, producing distinct commercial outcomes:

• Laboratory C reported thiabendazole 0.592 ± 0.019 mg kg-1 (k = 2) and Brix 70 ± 25 °Bx (k = 2). The very small uncertainty for the pesticide is expensive to achieve but unnecessarily excessive relative to the decision threshold, while the very large uncertainty for Brix makes price-related decisions unreliable.
• The producer’s laboratory reported thiabendazole 0.51 ± 0.20 mg kg-1 and Brix 61.2 ± 1.1 °Bx (both k = 2). The larger pesticide uncertainty was still compatible with Laboratory C’s result, but the tighter Brix uncertainty supported a different pricing decision.

These contrasting outcomes show two important points: overly small MU can inflate analytical costs without improving decision quality; overly large MU can make compliance or commercial decisions ambiguous or untenable. Even when values are metrologically compatible, differing uncertainty magnitudes can change the inferred status relative to thresholds and thus affect stakeholders’ choices.

Main results and discussion

• Fit-for-purpose requirement: A measurement is only suitable if its MU does not exceed the target MU tied to the intended decision. If no external target exists, the laboratory must set and justify one.
• Decision-driven targets: Targets are most robust when derived from the decision context—e.g., regulatory action limits, health-based thresholds, or customer commercial criteria—because these anchor uncertainty tolerances directly to consequences.
• Cost versus uncertainty trade-off: The example demonstrates the need for laboratories and clients to agree expected uncertainty levels in advance to avoid wasted expense or disputed outcomes.
• Use of coverage factors: Reporting both standard and expanded target uncertainties (utg and Utg) clarifies the statistical confidence associated with the target and facilitates comparison between laboratories that may use different reporting conventions.

Benefits and practical applications

• Improved decision quality: Well-defined target MU ensures that reported results support clear and consistent compliance and commercial decisions.
• Resource optimization: Laboratories can avoid unnecessarily costly analysis by targeting the uncertainty that is sufficient for the decision rather than minimizing uncertainty at all costs.
• Transparency and traceability: Documented target setting enhances communication between labs, clients and regulators and provides a defensible audit trail for quality management.
• Performance benchmarking: Targets allow laboratories to compare their routine MU with desired levels and prioritize method development or process improvements where needed.

Methodology and instrumentation (general guidance)

Although the source scenario does not list specific instruments, establishing and demonstrating compliance with a target MU typically involves:

• Constructing an uncertainty budget that quantifies contributions from sample heterogeneity, extraction, calibration, instrument repeatability, matrix effects, and operator variability.
• Using validation data, proficiency testing, and interlaboratory comparisons to estimate realistic between-lab and within-lab components of uncertainty.
• Applying appropriate statistical models to combine uncertainty components and to convert standard uncertainties to expanded uncertainties using a chosen coverage factor.
• Where applicable, selecting or qualifying instrumentation and method parameters to meet the defined target MU cost-effectively (for example, choosing an analytical technique with adequate precision and selectivity for the decision threshold).

Future trends and potential applications

• Harmonisation: Greater alignment of industry and regulatory target-setting practices is likely, with more consistent use of decision-driven criteria and shared performance benchmarks.
• Data-driven targets: Wider availability of interlaboratory data and big-data analytics will enable evidence-based target MU derivation reflecting real-world method performance.
• Risk-based frameworks: Integration of MU targets into risk assessment and management workflows will strengthen links between analytical uncertainty and health/safety decisions.
• Automation and adaptive methods: Advances in automated uncertainty estimation and laboratory informatics will streamline routine verification that measured MU meets targets, improving efficiency for high-throughput labs.
• Decision-support tools: Software that jointly evaluates measurement results, MU and user-defined decision rules will reduce ambiguity when reported values lie near action limits.

Conclusion

Setting a target measurement uncertainty is a practical necessity to ensure analytical results are meaningful for their intended use. Targets should be derived from the decision context where possible and documented by the laboratory when not specified externally. The fictional example underscores the economic and decision-making consequences of mismatched uncertainty: too small MU can generate unnecessary cost, while too large MU undermines confident decisions. Applying the Eurachem/CITAC guidance helps laboratories and stakeholders define appropriate, evidence-based targets and align measurement practice with the needs of users.

Reference

R. Bettencourt da Silva, A. Williams (Eds.) Eurachem/CITAC Guide: Setting and Using Target Uncertainty in Chemical Measurement, 2015; First English edition 2018. ISBN 978-989-98723-7-0. Produced by the Eurachem/CITAC Measurement Uncertainty and Traceability Working Group.