Measurement Uncertainty in Laboratory Medicine: A Guide to Calculation and Reporting

Measurement uncertainty quantifies the range within which the true value of a laboratory measurement is expected to lie. For laboratory students, technicians, researchers, and diagnostic professionals, understanding how to calculate and report measurement uncertainty is essential for interpreting results correctly and supporting clinical decisions. This article explains the concept of measurement uncertainty, provides a step-by-step approach to calculating it using standard methods, and discusses its role in clinical decision-making. A calculation worksheet and reporting template are included as practical tools.

At a Glance

Measurement uncertainty is a quantitative expression of the doubt associated with a measurement result. It reflects the combined effect of random and systematic errors in the entire measurement process, from sample collection to final result reporting. The table below summarizes key aspects of measurement uncertainty that laboratory professionals need to understand.

The Concept of Measurement Uncertainty

Measurement uncertainty is a fundamental property of all measurements. No measurement can be performed with perfect accuracy, and every result carries some degree of doubt. In laboratory medicine, this doubt arises from multiple sources, including sample collection, storage conditions, reagent lot variations, instrument performance, and operator technique. The Guide to the Expression of Uncertainty in Measurement provides the foundational framework for quantifying this doubt, and its recommendations apply to biomedical sciences as well as industrial metrology.

The true value of an analyte in a patient sample cannot be known with certainty. What laboratories measure is an estimate of that true value, and measurement uncertainty describes how close that estimate is likely to be. This concept is distinct from error, which is the difference between a measured value and the true value. Error can be random or systematic, but uncertainty encompasses both types of error in a single quantitative expression.

Measurement uncertainty is not a criticism of laboratory performance. It is a realistic acknowledgment that all measurement processes have limitations. When a laboratory reports a creatinine result of 80 µmol/L with an expanded uncertainty of 6 µmol/L, it is stating that the true value is likely to fall between 74 and 86 µmol/L. This information allows clinicians to interpret the result appropriately, particularly when results are near clinical decision limits.

Why Measurement Uncertainty Matters in Clinical Decision-Making

Clinical laboratory results are used to diagnose diseases, monitor treatment, and assess prognosis. The Global Burden of Disease Study 2021 demonstrates that accurate disease burden estimates depend on reliable health data, and laboratory results are a critical component of that data. When measurement uncertainty is ignored, clinicians may overinterpret small changes in results that are actually within the range of analytical variation.

Consider a patient with a hemoglobin result that changes from 120 g/L to 115 g/L between two visits. Without knowing the measurement uncertainty, a clinician might conclude that the patient's condition has worsened. If the expanded uncertainty is 5 g/L, however, the two results are statistically indistinguishable. The apparent change could be due to analytical variation instead of a true physiological change.

Measurement uncertainty also affects the interpretation of results near clinical decision limits. A glucose result of 7.1 mmol/L with an expanded uncertainty of 0.3 mmol/L could have a true value anywhere from 6.8 to 7.4 mmol/L. This range crosses the diagnostic threshold for diabetes, meaning the result is indeterminate and requires confirmation. Reporting measurement uncertainty alongside the result helps clinicians understand when additional testing is needed.

The Journal of Medical Biochemistry emphasizes that adequate assessment of measurement uncertainty is one of the most important factors for reliable interpretation of results. The ultimate goal is reducing diagnostic uncertainty, which requires understanding both the analytical aspects of each test and the clinical context in which results are used.

Regulatory and Accreditation Requirements

ISO 15189 is the international standard for medical laboratory quality and competence. The standard requires laboratories to estimate measurement uncertainty for each measurement procedure and to provide this information to laboratory users upon request. The Malawi-Liverpool Wellcome Research Programme experience demonstrates that implementing a quality management system aligned with ISO 15189 involves multiple steps, including gap analysis, documentation development, and staff training.

ISO/TS 20914:2019 provides practical guidance for estimating measurement uncertainty in medical laboratories. This technical specification introduces a top-down approach that balances statistical rigor with operational feasibility. The approach uses within-laboratory reproducibility data derived from internal quality control results collected over 6 to 12 months, combined with bias estimates from external quality assessment or certified reference materials.

The hemostasis laboratory review notes that international guidance recommends against reporting measurement uncertainty directly to service users on every result. Instead, laboratories are required to provide this information upon request. This approach ensures that clinicians receive the information they need without overwhelming them with technical details on every report.

Core Principles of Uncertainty Calculation

The Measurement Equation

The measurement equation describes the relationship between the output of a measurement process and its input quantities. For most laboratory tests, the output is the analyte concentration, and the inputs include instrument readings, calibration factors, sample volume, and reagent concentrations. The Advances in Clinical Chemistry review explains that when the output depends on multiple inputs, the propagation of uncertainties requires understanding differential calculus.

For routine laboratory practice, the measurement equation is often simplified. The combined standard uncertainty is calculated by combining the standard uncertainties of each input component. This combination uses the law of propagation of uncertainty, which accounts for the sensitivity of the output to changes in each input.

Standard Uncertainty and Combined Uncertainty

Standard uncertainty is the uncertainty of a measurement component expressed as a standard deviation. Each source of uncertainty must be expressed in standard deviation units before it can be combined. The Biochemia Medica article on statistical distributions explains that some components are naturally expressed as standard deviations, while others are expressed as ranges or intervals that must be converted.

For example, the uncertainty of a calibrator value might be expressed as a range of plus or minus a certain amount. This range must be converted to a standard deviation using the appropriate statistical distribution. A rectangular distribution, where all values within the range are equally likely, is converted by dividing the half-width of the range by the square root of 3. A normal distribution, where values near the center are more likely, is converted by dividing the half-width by 2.

The combined standard uncertainty is calculated by taking the square root of the sum of the squares of each component standard uncertainty. This calculation assumes that the components are independent, which is a reasonable approximation for most laboratory measurements. The second part of the Advances in Clinical Chemistry review discusses cases where correlations between inputs must be considered, such as in method comparison studies.

Expanded Uncertainty

Expanded uncertainty provides an interval that is expected to contain the true value with a specified level of confidence. It is calculated by multiplying the combined standard uncertainty by a coverage factor. A coverage factor of 2 corresponds to approximately 95% confidence for a normal distribution. The Advances in Clinical Chemistry review explains that the Gaussian density distribution is needed for determining expanded uncertainties.

For most clinical laboratory applications, a coverage factor of 2 is used. This means that the true value is expected to fall within the expanded uncertainty interval approximately 95% of the time. Some applications may require different coverage factors, such as a factor of 3 for 99.7% confidence, but these are less common in routine practice.

Step-by-Step Guide to Calculating Measurement Uncertainty

Step 1: Define the Measurand and Measurement Procedure

The first step is to clearly define what is being measured and how. This includes the analyte, the sample type, the measurement method, and the units of measurement. The definition should be specific enough that the uncertainty estimate applies to the actual measurement procedure used in the laboratory.

For example, measuring serum creatinine by an enzymatic method is different from measuring it by the Jaffe method. Each method has its own sources of uncertainty, and the uncertainty estimate must reflect the specific method in use.

Step 2: Identify Sources of Uncertainty

The next step is to identify all sources of uncertainty that contribute to the final result. Common sources include:

• Within-laboratory imprecision, estimated from internal quality control data
• Bias, estimated from external quality assessment or certified reference materials
• Calibrator uncertainty, provided by the manufacturer or determined from calibration certificates
• Sample-related factors, such as collection, handling, and storage
• Operator-related factors, such as technique and experience

The Critical Reviews in Clinical Laboratory Sciences article emphasizes that measurement uncertainty reflects the entire laboratory measurement process, beyond the final measurement step. All factors that influence the measured value should be considered.

Step 3: Quantify Each Source of Uncertainty

Each identified source of uncertainty must be quantified as a standard uncertainty. For within-laboratory imprecision, this is typically the standard deviation of internal quality control results collected over a defined period. The pragmatic measurement uncertainty letter recommends using 6 to 12 months of internal quality control data to capture routine sources of variation.

For bias, the uncertainty is typically derived from external quality assessment results or from measurements of certified reference materials. The bias itself is the difference between the laboratory's result and the target value, and the uncertainty of the bias estimate must also be considered.

Calibrator uncertainty is usually provided by the manufacturer on the calibration certificate. This value must be converted to a standard uncertainty using the appropriate statistical distribution.

Step 4: Convert All Components to Standard Deviations

All uncertainty components must be expressed in standard deviation units before they can be combined. The Biochemia Medica article on statistical distributions provides guidance on converting different types of uncertainty expressions to standard deviations.

For a component expressed as a range with a rectangular distribution, divide the half-width by the square root of 3. For a component expressed as a range with a triangular distribution, divide the half-width by the square root of 6. For a component expressed as a confidence interval, divide by the appropriate coverage factor.

Step 5: Calculate Combined Standard Uncertainty

The combined standard uncertainty is calculated by taking the square root of the sum of the squares of each component standard uncertainty. This is expressed mathematically as:

u_c = square root of (u_1 squared + u_2 squared + u_3 squared + ...)

where u_c is the combined standard uncertainty and u_1, u_2, u_3 are the individual component standard uncertainties.

This calculation assumes that the components are independent. When correlations exist, additional terms must be included in the calculation. The Advances in Clinical Chemistry review discusses correlation considerations in detail.

Step 6: Calculate Expanded Uncertainty

The expanded uncertainty is calculated by multiplying the combined standard uncertainty by a coverage factor. For a coverage factor of 2, the expanded uncertainty is:

U = 2 x u_c

The expanded uncertainty is typically reported as the result plus or minus the expanded uncertainty, for example, 80 µmol/L plus or minus 6 µmol/L.

Step 7: Document and Review

The uncertainty estimate must be documented, including the data used, the calculations performed, and the assumptions made. The estimate should be reviewed periodically and updated when significant changes occur, such as new reagent lots, instrument maintenance, or method changes.

The hemoglobin measurement study demonstrates the application of ISO/TS 20914:2019 to estimate measurement uncertainty in a specific testing context. The study evaluated within-laboratory imprecision, calibrator uncertainty, and bias uncertainty, and confirmed that all scenarios met desirable and minimum performance targets.

Top-Down Approaches to Uncertainty Estimation

Nordtest Approach

The Nordtest approach is a practical method for estimating measurement uncertainty using routine laboratory data. It combines within-laboratory reproducibility, estimated from internal quality control data, with bias, estimated from external quality assessment or certified reference materials. The comparison of top-down approaches found that the Nordtest approach using inter-laboratory internal quality control scheme data was the most practical formula for routine use.

The Nordtest approach is relatively simple to implement because it uses data that laboratories already collect for quality assurance purposes. It does not require additional measurements or complex experimental designs.

Eurolab Approach

The Eurolab approach is similar to the Nordtest approach but requires additional measurements to obtain uncertainty data. The comparison of top-down approaches noted that the Eurolab approach requires additional measurements, which may be a barrier to implementation in busy clinical laboratories.

Cofrac Approach

The Cofrac approach tends to produce higher uncertainty estimates than the Nordtest or Eurolab approaches. The comparison of top-down approaches found that the Cofrac approach showed the highest uncertainties, followed by Eurolab and Nordtest. The choice of approach depends on the laboratory's resources, the availability of data, and the requirements of accreditation bodies.

Practical Implementation in the Laboratory

Data Collection for Uncertainty Estimation

The quality of an uncertainty estimate depends on the quality of the data used. Internal quality control data should be collected over a sufficient period to capture routine sources of variation. The pragmatic measurement uncertainty letter recommends using 6 to 12 months of internal quality control data, partitioned by reagent and quality control lot as recommended by CLSI C24 and ISO/TS 20914.

For analytes used in long-term monitoring, such as thyroglobulin, PSA, and creatinine, the relevant uncertainty spans every reagent and calibrator lot, recalibration, and maintenance event within the clinical application window. A result from 2021 may be compared with one from 2026, and the uncertainty must reflect all sources of variation within that interval.

Using Internal Quality Control Data

Internal quality control data provides the primary estimate of within-laboratory imprecision. The standard deviation of quality control results is calculated for each control level and each analyte. This standard deviation represents the random variation of the measurement process under routine conditions.

The measurement uncertainty as a universal concept article notes that measurement uncertainty reflects how different the results of multiple measurements of the same analyte in the same sample are, as well as whether the method is subject to significant random and systematic deviation.

Using External Quality Assessment Data

External quality assessment data provides an estimate of bias. The difference between the laboratory's result and the target value for each survey sample represents the systematic error of the measurement process. The mean bias across multiple surveys provides a more stable estimate than a single survey.

The comparison of top-down approaches found that bias was highest using proficiency testing data, followed by certified reference calibrators and inter-laboratory internal quality control scheme data. The choice of bias source affects the final uncertainty estimate.

Incorporating Calibrator Uncertainty

Calibrator uncertainty is a component of measurement uncertainty that is often overlooked. The uncertainty of the calibrator value propagates through the measurement process to the final result. Calibrator uncertainty is typically provided by the manufacturer on the calibration certificate.

The hemoglobin measurement study evaluated three scenarios: within-laboratory imprecision alone, addition of calibrator uncertainty, and incorporation of significant bias uncertainty. All scenarios met desirable and minimum performance targets for allowable uncertainty, demonstrating that the inclusion of additional uncertainty components does not necessarily lead to unacceptable performance.

Reporting Measurement Uncertainty

What to Report

Measurement uncertainty should be reported in a way that is understandable to the user. The Critical Reviews in Clinical Laboratory Sciences article emphasizes that the presentation of measurement uncertainty should be reduced to routine and simple presentation, understandable to both laboratory professionals and physicians.

The expanded uncertainty is typically reported as the result plus or minus the uncertainty value, with the coverage factor and confidence level stated. For example: "Creatinine 80 µmol/L, expanded uncertainty 6 µmol/L (coverage factor 2, approximately 95% confidence)."

When to Report

International guidance recommends against reporting measurement uncertainty directly to service users on every result. Instead, the laboratory is required to provide this information upon request. The hemostasis laboratory review notes that this approach balances the need for transparency with the practical needs of clinicians who may not require uncertainty information for every result.

Laboratories should be prepared to provide measurement uncertainty information when requested by clinicians, researchers, or accreditation bodies. This requires maintaining up-to-date documentation of uncertainty estimates for all measurement procedures.

Interpreting Results with Uncertainty

Measurement uncertainty should be considered when interpreting results near clinical decision limits. The free valproic acid monitoring study demonstrated that incorporating measurement uncertainty altered therapeutic classification in 4 of 9 patients, mainly near the lower decision limit. This led to more results being reported as indeterminate.

When a result is near a clinical decision limit, the uncertainty interval may cross the limit. In this case, the result should be interpreted with caution, and additional testing may be needed to confirm the classification.

Common Failure Patterns in Uncertainty Estimation

Insufficient Data Duration

Using too little internal quality control data leads to unstable uncertainty estimates. The pragmatic measurement uncertainty letter recommends using 6 to 12 months of data to capture routine sources of variation. Shorter periods may miss important sources of variation, such as reagent lot changes or seasonal effects.

Ignoring Bias

Some laboratories calculate uncertainty from imprecision data alone, ignoring bias. This leads to underestimation of uncertainty because systematic errors are not accounted for. The measurement uncertainty as a universal concept article notes that the true value of the analyte cannot be known, and the closest value is obtained by a perfect method with zero bias and inaccuracy, which is practically impossible to achieve.

Using Inappropriate Statistical Distributions

Converting uncertainty components to standard deviations requires knowledge of the appropriate statistical distribution. Using the wrong distribution leads to incorrect standard uncertainties and, consequently, incorrect combined and expanded uncertainties. The Biochemia Medica article on statistical distributions provides guidance on the main statistical distributions used in measurement uncertainty calculation.

Failing to Update Estimates

Measurement uncertainty estimates should be reviewed and updated periodically. Changes in reagents, calibrators, instruments, or procedures can affect uncertainty. The Malawi-Liverpool Wellcome Research Programme experience demonstrates the importance of ongoing quality management, including regular review of quality indicators.

Overlooking Pre-Analytical Factors

Measurement uncertainty is often considered only for the analytical phase, but pre-analytical factors can contribute significantly to the total uncertainty. Sample collection, handling, and storage can introduce variation that is not captured by internal quality control data. The anatomic pathology study highlights the importance of pre-examination processes and interface management in specimen identification and tracking.

Measurement Uncertainty in Specific Laboratory Applications

Hemostasis Testing

Hemostasis assays present unique challenges for measurement uncertainty estimation. The hemostasis laboratory review notes that measurement uncertainty describes uncertainty propagation across the entire traceability chain, from international standards to final patient results. The applicability of ISO/TS 20914:2019 to specialist areas such as hemostasis has been questioned, and laboratories must carefully consider the appropriate approach for their specific assays.

Therapeutic Drug Monitoring

Therapeutic drug monitoring requires careful consideration of measurement uncertainty because results are interpreted against narrow therapeutic intervals. The free valproic acid monitoring study demonstrated that reporting free valproic acid concentrations with measurement uncertainty may reduce the risk of misclassification and enhance transparency.

Hemoglobin Testing in Blood Establishments

Hemoglobin testing in blood establishments determines donor eligibility and informs transfusion decisions. The hemoglobin measurement study applied ISO/TS 20914:2019 to estimate measurement uncertainty in hemoglobin testing using the cyanmethemoglobin method. The study confirmed that measurement uncertainty estimation is a European and ISO-endorsed best practice in transfusion laboratory medicine.

LC-MS/MS Methods

Liquid chromatography-tandem mass spectrometry methods require validation of analytical performance, including measurement uncertainty. The creatine deficiency syndromes study developed and validated a rapid LC-MS/MS method and assessed measurement uncertainty in accordance with ISO 15189:2022 requirements. The study demonstrated robust intra- and inter-day precision and satisfactory results in external quality assessment schemes.

Records and Documentation

Uncertainty Budget

An uncertainty budget is a document that lists all sources of uncertainty, their standard uncertainties, and their contribution to the combined uncertainty. The uncertainty budget should be maintained for each measurement procedure and updated when significant changes occur.

The uncertainty budget should include:

• A description of the measurand and measurement procedure
• A list of all uncertainty components
• The standard uncertainty of each component
• The combined standard uncertainty
• The expanded uncertainty and coverage factor
• The date of the estimate and the data used

Quality Control Records

Internal quality control records provide the data for estimating within-laboratory imprecision. These records should include the control material used, the control levels, the dates of analysis, and the results obtained. The pragmatic measurement uncertainty letter recommends partitioning quality control data by reagent and quality control lot to keep long-term estimates sound.

External Quality Assessment Records

External quality assessment records provide the data for estimating bias. These records should include the survey dates, the target values, the laboratory's results, and the calculated bias for each survey. Multiple surveys should be used to obtain a stable estimate of bias.

Limitations and Professional Escalation

Limitations of Measurement Uncertainty

Measurement uncertainty is a valuable tool, but it has limitations. It does not account for all sources of variation, particularly those that are difficult to quantify, such as pre-analytical factors. It also does not address the clinical validity of a test, which depends on the biological variation of the analyte and the clinical context.

The Journal of Medical Biochemistry notes that setting diagnoses in medicine is a decision with a certain degree of uncertainty, instead of a statistically and mathematically calculated conclusion. Measurement uncertainty is one input to clinical decision-making, but it is not the only input.

When to Escalate

Laboratory professionals should escalate concerns about measurement uncertainty in the following situations:

• When uncertainty estimates exceed acceptable performance targets
• When internal quality control data show trends or shifts that suggest increased variation
• When external quality assessment results show persistent bias
• When changes in reagents, calibrators, or instruments may affect uncertainty
• When clinicians raise concerns about the interpretation of results near decision limits

Escalation should follow the laboratory's quality management procedures and may involve the quality manager, laboratory director, or other designated personnel.

Frequently Asked Questions

What is the difference between measurement uncertainty and error?

Measurement uncertainty is a quantitative expression of the doubt associated with a measurement result, while error is the difference between a measured value and the true value. Error can be random or systematic, but uncertainty encompasses both types of error in a single quantitative expression. The true value cannot be known, so error cannot be calculated exactly, but uncertainty can be estimated from the known sources of variation in the measurement process.

How is measurement uncertainty calculated in a clinical laboratory?

Measurement uncertainty is calculated by identifying all sources of uncertainty, quantifying each as a standard deviation, combining them using the law of propagation of uncertainty, and multiplying the combined standard uncertainty by a coverage factor to obtain the expanded uncertainty. The most practical approach for routine use is the top-down method using internal quality control data for imprecision and external quality assessment data for bias.

What is the difference between combined standard uncertainty and expanded uncertainty?

Combined standard uncertainty is the result of combining all component standard uncertainties using the law of propagation of uncertainty. Expanded uncertainty is calculated by multiplying the combined standard uncertainty by a coverage factor, typically 2, to provide an interval that is expected to contain the true value with approximately 95% confidence.

How much internal quality control data is needed for uncertainty estimation?

ISO/TS 20914:2019 recommends using 6 to 12 months of internal quality control data to estimate within-laboratory reproducibility. This duration captures routine sources of variation, including reagent lot changes, calibrator changes, and maintenance events. For analytes used in long-term monitoring, a longer duration may be needed to match the clinical application window of the test.

Should measurement uncertainty be reported on every laboratory result?

International guidance recommends against reporting measurement uncertainty directly to service users on every result. Instead, the laboratory is required to provide this information upon request. Laboratories should maintain up-to-date documentation of uncertainty estimates and be prepared to share them when requested by clinicians, researchers, or accreditation bodies.

How does measurement uncertainty affect the interpretation of results near clinical decision limits?

When a result is near a clinical decision limit, the uncertainty interval may cross the limit. This means the true value could be on either side of the limit, making the classification indeterminate. In this case, the result should be interpreted with caution, and additional testing may be needed to confirm the classification.

What are the common mistakes in measurement uncertainty estimation?

Common mistakes include using insufficient data, ignoring bias, using inappropriate statistical distributions, failing to update estimates, and overlooking pre-analytical factors. These mistakes lead to inaccurate uncertainty estimates and can affect clinical decision-making.

How often should measurement uncertainty estimates be updated?

Measurement uncertainty estimates should be reviewed periodically and updated when significant changes occur, such as new reagent lots, instrument maintenance, method changes, or changes in external quality assessment performance. The pragmatic measurement uncertainty letter recommends a two-stage scheme with a rolling annual estimate for analytical trends and an aggregate long-term estimate for clinical reporting.

Related Diagnostic Guides

• Absolute Quantification Using qPCR: Standard Curves and Copy Number Calculation
• qPCR Efficiency Calculation: Methods and Interpretation
• Laboratory Observation: Recording and Reporting Experimental Findings
• DNA Quantification Using a Spectrophotometer: Nanodrop and UV-Vis Methods
• Laboratory Investigation: A Step-by-Step Approach to Troubleshooting Failed Experiments

References and Further Reading

• Laboratory Quality Management System Handbook. World Health Organization.
• Laboratory Biosafety Manual. World Health Organization.
• Assay Guidance Manual. National Center for Advancing Translational Sciences.
• Bioanalytical Method Validation Guidance. U.S. Food and Drug Administration.
• NCBI Literature Resources. National Center for Biotechnology Information.
• Global incidence, prevalence, years lived with disability (YLDs), disability-adjusted life-years (DALYs), and healthy life expectancy (HALE) for 371 diseases and injuries in 204 countries and territories and 811 subnational locations, 1990-2021: a systematic analysis for the Global Burden of Disease Study 2021.. Lancet (London, England), 2024.
• Global burden and strength of evidence for 88 risk factors in 204 countries and 811 subnational locations, 1990-2021: a systematic analysis for the Global Burden of Disease Study 2021.. Lancet (London, England), 2024.
• Global, regional, and national burden of stroke and its risk factors, 1990-2021: a systematic analysis for the Global Burden of Disease Study 2021.. The Lancet. Neurology, 2024.
• Uncertainty in Measurement: Procedures for Determining Uncertainty With Application to Clinical Laboratory Calculations.. Advances in clinical chemistry, 2018.
• Uncertainty in Measurement: Procedures for Determining Uncertainty With Application to Clinical Laboratory Calculations.. Advances in clinical chemistry, 2018.
• Measurement uncertainty as a universal concept: can it be universally applicable in routine laboratory practice?. Critical reviews in clinical laboratory sciences, 2021.
• Uncertainty of Measurement in Laboratory Medicine.. Journal of medical biochemistry, 2018.
• Measurement Uncertainty in the Hemostasis Laboratory.. Seminars in thrombosis and hemostasis, 2025.
• Pragmatic measurement uncertainty: Aligning IQC data duration with clinical reality in long-term monitoring.. 2026.
• Emerging practices in the estimation of hemoglobin measurement: uncertainty in blood establishments.. 2026.
• Rapid LC-MS/MS Method for Targeted Assay of Creatine Deficiency Syndromes in Morocco.. 2026.
• Incorporating measurement uncertainty into free valproic acid monitoring: estimation and clinical impact for therapeutic drug monitoring.. 2026.
• Implementing and Transitioning a Laboratory Quality Management System from ISO 15189:2012 to ISO 15189:2022: Experience from the Malawi-Liverpool Wellcome Research Programme, Blantyre.. 2026.
• Digital Specimen Tracking- and ISO 15189-Oriented Risk Management in Anatomic Pathology: A Qualitative Study of Expert Perspectives in Western Austria.. 2026.
• Statistical distributions commonly used in measurement uncertainty in laboratory medicine. Biochemia Medica, 2020.
• Measurement uncertainty in laboratory medicine: the bridge between medical and industrial metrology. Turkish Journal of Biochemistry, 2019.
• The top-down approach to measurement uncertainty: which formula should we use in laboratory medicine?. Biochemia Medica, 2020.
• Uncertainty of measurement results in laboratory medicine. Izmeritel`naya Tekhnika, 2022.
• Permissible limits for uncertainty of measurement in laboratory medicine. Clinical Chemistry and Laboratory Medicine, 2015.
• Measurement uncertainty calculation in clinical laboratories. Biochimica Clinica, 2015.
• Assessing the measurement uncertainty of qualitative analysis in the clinical laboratory. Journal of Laboratory Medicine, 2020.
• Measurement uncertainty for clinical laboratories-A revision of the concept. Clinical Chemistry and Laboratory Medicine, 2016.
• Calculation of the uncertainty related to the measurement of γ-glutamyltransferase: Comparison of three different approaches. Biochimica Clinica, 2010.
• Measurement uncertainty in clinical biochemistry laboratories: Practical application. Turkish Journal of Biochemistry, 2011.

This article is educational and does not replace validated laboratory procedures, institutional biosafety review, manufacturer instructions, or professional interpretation.