Uncertainty in Measurement and Total Error

When bias becomes part of imprecision: how to use analytical performance specifications to determine acceptability of lot-lot variation and other sources of possibly unacceptable bias

ISO 15189 requires laboratories to estimate the uncertainty of their quantitative measurements and to maintain them within relevant performance specifications. Furthermore, it refers to ISO TS 20914 for instructions on how to estimate the uncertainty and what to take into consideration when communicating uncertainty of measurement with requesting clinicians. These instructions include the responsibility of laboratories to verify that bias is not larger than medically significant. If estimated to be larger than acceptable, such bias first needs to be eliminated or (temporarily) corrected for. In the latter case, the uncertainty of such correction becomes part of the estimation of the total measurement uncertainty. If small enough to be acceptable, bias becomes part of the long term within laboratory random variation. Sources of possible bias are (not limited to) changes in reagent or calibrator lot variation or calibration itself. In this paper we clarify how the rationale and mathematics from an EFLM WG ISO/A position paper on allowable between reagent lot variation can be applied to calculate whether bias can be accepted to become part of long-term imprecision. The central point of this rationale is to prevent the risk that requesting clinicians confuse changes in bias with changes in the steady state of their patients.

Measurement uncertainty

Over time, the metrological concept of uncertainty in measurement has been very successfully integrated into laboratory sciences. For proper implementation, an understanding of specific metrology terminology and additional concepts such as metrology traceability and commutability is necessary. Although the original thinking about measurement uncertainty in laboratory medicine suggests the complexity of the concept, it basically refers to the result as the end product of the entire laboratory process. Although the data on measurement uncertainty can be expressed quantitatively, the basis of this concept is the continuous evaluation of all phases of the laboratory process. This means that laboratory experts should keep in mind that the extra-analytical phases (on which the uncertainty of the measurement results may depend the most) must be continuously monitored. The analytical phase can be "held in check" by established internal and external quality control processes. It is the internal/external quality control data that is used to calculate the numerical value of the measurement uncertainty of the measurement results. Although over time the awareness of laboratory experts regarding the concept of measurement uncertainty has increased, there are still many challenges that need to be followed, and the last one is how to achieve a balance between understanding, evaluation process and application of measurement uncertainty data of measurement results for complete and ultimate practical use.

Clinical decision limits as criteria for setting analytical performance specifications for laboratory tests

BACKGROUND

The biological (CV_I), preanalytical (CV_PRE), and analytical variation (CV_A) are inherent to clinical laboratory testing and consequently, interpretation of clinical test results.

METHODS

The sum of the CV_I, CV_PRE, and CV_A, called diagnostic variation (CV_D), was used to derive clinically acceptable analytical performance specifications (CAAPS) for clinical chemistry measurands. The reference change concept was applied to clinically significant differences (CD) between two measurements, with the formula CD = z*√2* CV_D. CD for six measurands were sought from international guidelines. The CAAPS were calculated by subtracting variances of CV_I and CV_PRE from CV_D. Modified formulae were applied to consider statistical power (1-β) and repeated measurements.

RESULTS

The obtained CAAPS were 44.9% for urine albumin, 0.6% for plasma sodium, 22.9% for plasma pancreatic amylase, and 8.0% for plasma creatinine (z = 3, α = 2.5%, 1-β = 85%). For blood HbA_1c and plasma low-density lipoprotein cholesterol, replicate measurements were necessary to reach CAAPS for patient monitoring. The derived CAAPS were compared with analytical performance specifications, APS, based on biological variation.

CONCLUSIONS

The CAAPS models pose a new tool for assessing APS in a clinical laboratory. Their usability depends on the relevance of CD limits, required statistical power and the feasibility of repeated measurements.

Towards a Standard Method for Estimating Fragmentation Rates in Emulsification Experiments

The fragmentation rate function connects the fundamental drop breakup process with the resulting drop size distribution and is central to understanding or modeling emulsification processes. There is a large interest in being able to reliably measure it from an emulsification experiment, both for generating data for validating theoretical fragmentation rate function suggestions and as a tool for studying emulsification processes. Consequently, several methods have been suggested for measuring fragmentation rates based on emulsion experiments. Typically, each study suggests a new method that is rarely used again. The lack of an agreement on a standard method has become a substantial challenge. This contribution critically and systematically analyses four influential suggestions of how to measure fragmentation rate in terms of validity, reliability, and sensitivity to method assumptions. The back-calculation method is identified as the most promising—high reliability and low sensitivity to assumption—whereas performing a non-linear regression on a parameterized model (as commonly suggested) is unsuitable due to its high sensitivity. The simplistic zero-order method is identified as an interesting supplemental tool that could be used for qualitative comparisons but not for quantification.

Patient and clinician experiences of uncertainty in the diagnostic process: Current understanding and future directions

OBJECTIVE

Uncertainty occurs throughout the diagnostic process and must be managed to facilitate accurate and timely diagnoses and treatments. Better characterization of uncertainty can inform strategies to manage it more effectively in clinical practice. We provide a comprehensive overview of current literature on diagnosis-related uncertainty describing (1) where patients and clinicians experience uncertainty within the diagnostic process, (2) how uncertainty affects the diagnostic process, (3) roots of uncertainty related to probability/risk, ambiguity, or complexity, and (4) strategies to manage uncertainty.

DISCUSSION

Each diagnostic process step involves uncertainty, including patient engagement with the healthcare system; information gathering, interpretation, and integration; formulating working diagnoses; and communicating diagnoses to patients. General management strategies include acknowledging uncertainty, obtaining more contextual information from patients (e.g., gathering occupations and family histories), creating diagnostic safety nets (e.g., informing patients what red flags to look for), engaging in worst case/best case scenario planning, and communicating diagnostic uncertainty to patients, families, and colleagues. Potential strategies tailored to various aspects of diagnostic uncertainty are also outlined.

CONCLUSION

Scientific knowledge on diagnostic uncertainty, while previously elusive, is now becoming more clearly defined. Next steps include research to evaluate relationships between management and communication of diagnostic uncertainty and improved patient outcomes.

Great future or greedy venture: Precision medicine needs philosophy

INTRODUCTION

Over the past decade, we have witnessed the initiation and implementation of precision medicine (PM), a discipline that promises to individualize and personalize medical management and treatment, rendering them ultimately more precise and effective. Despite of the continuing advances and numerous clinical applications, the potential of PM remains highly controversial, sparking heated debates about its future.

METHOD

The present article reviews the philosophical issues and practical challenges that are critical to the feasibility and implementation of PM.

OUTCOME

The explanation and argument about the relations between PM and computability, uncertainty as well as complexity, show that key foundational assumptions of PM might not be fully validated.

CONCLUSION

The present analysis suggests that our current understanding of PM is probably oversimplified and too superficial. More efforts are needed to realize the hope that PM has elicited, rather than make the term just as a hype.

Application of the Life Cycle Management of Analytical methods concept to a HPTLC-DPPH assay method for acteoside content in industrial extracts of Plantago lanceolata L

Analytical methods used for quality control of plants and plant extracts are based on the identification and quantification of chemical markers to manage batch reproducibility and efficacy. The aim of this work was to assess the performance of a High Performance Thin Layer Chromatography (HPTLC) method developed for quality control of industrial dry extracts of ribwort plantain (P. lanceolata L.), using 2,2-diphenyl 1-picrylhydrazyle (DPPH) effect directed chemical reaction for antioxidant activity of acteoside, a phenylethanoid glycoside commonly used as a marker for P. lanceolata L., and to demonstrate the applicability of the Life Cycle Management of Analytical Methods concept to quantitative HPTLC-DPPH methods. The first step was the determination of the Analytical Target Profile (ATP) and Target Measurement Uncertainty (TMU), taking into account the quality control requirements for such extracts and the detection method applicable range. Once the desired range was established, an evaluation of the calibration function was conducted using several calibration models. Due to the lack of reference samples, spiked samples were used to evaluate the accuracy of the method by means of Total Analytical Error (TAE) determination, using prediction intervals calculation for the selected calibration functions. Measurement Uncertainty (MU) was also estimated, allowing the final choice of the calibration function to be used for quality control, giving the most fit for purpose performance level in accordance with the product specifications. As Life Cycle Management of the method also includes its routine use, the Measurement Uncertainty was checked on spiked and unspiked extract samples with different dilution levels, in order to verify the accordance of results between spiked and unspiked samples and to prepare a replication strategy to be applied during the routine use of the method.

ISO/TS 20914:2019 - a critical commentary

The long-anticipated ISO/TS 20914, Medical laboratories - Practical guidance for the estimation of measurement uncertainty, became publicly available in July 2019. This ISO document is intended as a guide for the practical application of estimating uncertainty in measurement (measurement uncertainty) in a medical laboratory. In some respects, the guide does indeed meet many of its stated objectives with numerous very detailed examples. Even though it is claimed that this ISO guide is based on the Evaluation of measurement data - Guide to the expression of uncertainty in measurement (GUM), JCGM 100:2008, it is with some concern that we believe several important statements and statistical procedures are incorrect, with others potentially misleading. The aim of this report is to highlight the major concerns which we have identified. In particular, we believe the following items require further comment: (1) The use of coefficient of variation and its potential for misuse requires clarification, (2) pooled variance and measurement uncertainty across changes in measuring conditions has been oversimplified and is potentially misleading, (3) uncertainty in the results of estimated glomerular filtration rate (eGFR) do not include all known uncertainties, (4) the international normalized ratio (INR) calculation is incorrect, (5) the treatment of bias uncertainty is considered problematic, (6) the rules for evaluating combined uncertainty in functional relationships are incomplete, and (7) specific concerns with some individual statements.

Personalized Therapy for Mycophenolate: Consensus Report by the International Association of Therapeutic Drug Monitoring and Clinical Toxicology

ABSTRACT

When mycophenolic acid (MPA) was originally marketed for immunosuppressive therapy, fixed doses were recommended by the manufacturer. Awareness of the potential for a more personalized dosing has led to development of methods to estimate MPA area under the curve based on the measurement of drug concentrations in only a few samples. This approach is feasible in the clinical routine and has proven successful in terms of correlation with outcome. However, the search for superior correlates has continued, and numerous studies in search of biomarkers that could better predict the perfect dosage for the individual patient have been published. As it was considered timely for an updated and comprehensive presentation of consensus on the status for personalized treatment with MPA, this report was prepared following an initiative from members of the International Association of Therapeutic Drug Monitoring and Clinical Toxicology (IATDMCT). Topics included are the criteria for analytics, methods to estimate exposure including pharmacometrics, the potential influence of pharmacogenetics, development of biomarkers, and the practical aspects of implementation of target concentration intervention. For selected topics with sufficient evidence, such as the application of limited sampling strategies for MPA area under the curve, graded recommendations on target ranges are presented. To provide a comprehensive review, this report also includes updates on the status of potential biomarkers including those which may be promising but with a low level of evidence. In view of the fact that there are very few new immunosuppressive drugs under development for the transplant field, it is likely that MPA will continue to be prescribed on a large scale in the upcoming years. Discontinuation of therapy due to adverse effects is relatively common, increasing the risk for late rejections, which may contribute to graft loss. Therefore, the continued search for innovative methods to better personalize MPA dosage is warranted.

A Software Tool for Calculating the Uncertainty of Diagnostic Accuracy Measures

Screening and diagnostic tests are applied for the classification of people into diseased and non-diseased populations. Although diagnostic accuracy measures are used to evaluate the correctness of classification in clinical research and practice, there has been limited research on their uncertainty. The objective for this work was to develop a tool for calculating the uncertainty of diagnostic accuracy measures, as diagnostic accuracy is fundamental to clinical decision-making. For this reason, the freely available interactive program Diagnostic Uncertainty has been developed in the Wolfram language. The program provides six modules with nine submodules for calculating and plotting the standard measurement, sampling and combined uncertainty and the resultant confidence intervals of various diagnostic accuracy measures of screening or diagnostic tests, which measure a normally distributed measurand, applied at a single point in time in samples of non-diseased and diseased populations. This is done for differing sample sizes, mean and standard deviation of the measurand, diagnostic threshold and standard measurement uncertainty of the test. The application of the program is demonstrated with an illustrative example of glucose measurements in samples of diabetic and non-diabetic populations, that shows the calculation of the uncertainty of diagnostic accuracy measures. The presented interactive program is user-friendly and can be used as a flexible educational and research tool in medical decision-making, to calculate and explore the uncertainty of diagnostic accuracy measures.

A Software Tool for Exploring the Relation between Diagnostic Accuracy and Measurement Uncertainty

Screening and diagnostic tests are used to classify people with and without a disease. Diagnostic accuracy measures are used to evaluate the correctness of a classification in clinical research and practice. Although this depends on the uncertainty of measurement, there has been limited research on their relation. The objective of this work was to develop an exploratory tool for the relation between diagnostic accuracy measures and measurement uncertainty, as diagnostic accuracy is fundamental to clinical decision-making, while measurement uncertainty is critical to quality and risk management in laboratory medicine. For this reason, a freely available interactive program was developed for calculating, optimizing, plotting and comparing various diagnostic accuracy measures and the corresponding risk of diagnostic or screening tests measuring a normally distributed measurand, applied at a single point in time in non-diseased and diseased populations. This is done for differing prevalence of the disease, mean and standard deviation of the measurand, diagnostic threshold, standard measurement uncertainty of the tests and expected loss. The application of the program is illustrated with a case study of glucose measurements in diabetic and non-diabetic populations. The program is user-friendly and can be used as an educational and research tool in medical decision-making.

Analysis of the Thermal Performance of the Embedded Composite Phase Change Energy Storage Wall

In this study, the phase change paraffin and metal powder were mixed to form the composite phase change energy-storing material. This composite material was then injected into metal coil tubings at different coil spacings to form a composite phase change energy storage tubing system, which was then embedded in a wall. The thermal performance of the embedded phase change energy storage wall was investigated at various temperatures. The results showed that among the four types of aforementioned walls, the energy storage tubes at a spacing of 20 mm exhibited the smallest heat transfer and the largest surface heat storage coefficients. Therefore, this wall can block heat flow and temperature propagation effectively, and it exhibits excellent thermal insulating and heat storage performances and increased resistance to temperature fluctuations.

Measurement uncertainty as a universal concept: can it be universally applicable in routine laboratory practice?

Measurement uncertainty (MU) of results is one of the basic recommended and accepted statistical methods in laboratory medicine, with which analytical and clinical evaluation of laboratory test quality is assessed. Literature data indicate that the calculation of MU is not a simple process, but that its assessment in daily laboratory practice should be reduced to routine and simple presentation, understandable to both laboratory professionals and physicians. In order to achieve this, it is necessary to understand the purpose of the test for which MU is to be determined. Various suggestions have been given for presentation of MU as a quantitative indicator of the quality of the final measurement result in the medical laboratory. Although MU refers to the final measurement result, this metrological concept reflects the entire laboratory measurement process. The data on estimated MU is used to interpret the measured numerical result, and represents quantitatively the quality of the measurement itself, i.e. how different are the results of multiple measurements of the analyte of interest in the same sample, as well as whether the method of determination itself is subjected to significant random and systematic deviation. Initially, in the metrological concept, the MU is viewed in relation to the true value of the analyte of interest. However, the true value of the analyte measured in the biological fluid matrix of the study population cannot be known. It is therefore considered the closest value obtained by the perfect method, for which the bias and inaccuracy, as measures of systematic and random error, are equal to zero, which is practically impossible to achieve in routine laboratory practice. Although current standards require accredited medical laboratories to estimate MU, none of these guidelines provide clear guidance on how this can be achieved in daily laboratory work. This review examines literary data and documents dealing with MU issues, but also highlights what additional terms and data should be considered when interpreting MU. This paper ultimately draws attention, and once again points out, that a simpler solution is needed for this universal concept to be formally and universally applicable in routine laboratory medicine practice.

Statistical distributions commonly used in measurement uncertainty in laboratory medicine

Uncertainty is an inseparable part of all types of measurement. Recently, the International Organization for Standardization (ISO) released a new standard (ISO 20914) on how to calculate measurement uncertainty (MU) in laboratory medicine. This standard can be regarded as the beginning of a new era in laboratory medicine. Measurement uncertainty comprises various components and is used to calculate the total uncertainty. All components must be expressed in standard deviation (SD) and then combined. However, the characteristics of these components are not the same; some are expressed as SD, while others are expressed as a ± b, such as the purity of the reagents. All non-SD variables must be transformed into SD, which requires a detailed knowledge of common statistical distributions used in the calculation of MU. Here, the main statistical distributions used in MU calculation are briefly summarized.

Assessing the measurement uncertainty of qualitative analysis in the clinical laboratory

Measurement uncertainty is a parameter that is associated with the dispersion of measurements. Assessment of the measurement uncertainty is recommended in qualitative analyses in clinical laboratories; however, the measurement uncertainty of qualitative tests has been neglected despite the introduction of many adequate methods. We herein provide an overview of three reasonable statistical methods for quantifying the measurement uncertainties of qualitative assays, namely Bayes’ theorem, the normal distribution method, and the information theoretic approach. Unlike in quantitative analysis, the measurement uncertainty of qualitative analysis is expressed using a conditional probability, likelihood ratio, and entropy. With the necessary theoretical background, the practical applications for clinical laboratories are also provided using statistical calculations. Using statistical approaches, we hope that our review will contribute to the use of measurement uncertainty in qualitative analyses in the clinical laboratory environment.

Performance characteristics of automated clinical chemistry analyzers using commercial assay reagents contributing to quality assurance and clinical decision in a hospital laboratory

Background: Clinical laboratories provide essential diagnostic services that are essential in clinical decision making, contributing to the quality of healthcare. The performance of two Siemens ADVIA 1800 analyzers was characterized in a hospital Biochemistry laboratory in order to evaluate the analytical characteristics of such automated analyzer systems using nonoriginal assay reagents attempting to support laboratory quality service and crucial clinical decision making. Methods: We independently completed performance validation studies including trueness, precision, sensitivity as well as measurement of uncertainty and sigma metrics calculation for 25 biochemical parameters. Results: Trueness expressed as bias was less than 20% for both ADVIA 1800 analyzers. Within run and total precisions expressed as CV% were ≤9.85% on both analyzers for most parameters studied with few exceptions (Mg, TB, DB, Cl, HDL and UA) observed either in low or in high level samples and between the two analyzers. LoB, LoD and LoQ values produced by the two analyzers were comparable except Cl. Uncertainty values produced by the two analyzers were comparable with no significant differences. Quality performance of reagent assays was studied using the sigma metrics system. The sigma values were plotted on normalized method decision charts for graphical representation of assay performances for each analyzer. Conclusions: The two ADVIA systems, independently evaluated, showed consistent performance characteristics with certain discrepancies by several reagents. Sigma analysis was helpful for revealing the quality performance of non-original reagents supporting the need for strict assessment of quality assurance and in some instances optimization/improvement of assay methods.

Uncertainty in measurement and total error: different roads to the same quality destination?

The debate comparing the benefits of measurement uncertainty (uncertainty in measurement, MU) with total error (TE) for the assessment of laboratory performance continues. The summary recently provided in this journal by members of the Task and Finish Group on Total Error (TFG-TE) of the EFLM put the arguments into clear perspective. Even though there is generally strong support for TE in many laboratories, some of the arguments proposed for its on-going support require further comment. In a recent opinion which focused directly on the TFG-TE summary, several potentially confusing statements regarding ISO15189 and the Evaluation of measurement data - Guide to the expression of uncertainty in measurement (GUM) were again promulgated to promote TE methods for assessing uncertainty in laboratory measurement. In this opinion, we present an alternative view of the key issues and outline our views with regard to the relationship between ISO15189, uncertainty in measurement and the GUM.

Reporting results with (Un)certainty

BACKGROUND

A degree of uncertainty occurs with every measured laboratory result due to both analytical and biological variation. The tools of Total Observed error (TE_O ) and dispersion based on biological variation have helped veterinary labs quantify the causes of variation that lead to measurement uncertainty (MU). International organizations recommend that the amount of MU in veterinary laboratory results be identified and communicated. The expanded measurement uncertainty (EMU), dispersion, and reporting interval adjustment have been recommended as tools to allow communication of MU to laboratory data users but are not commonly discussed in the veterinary literature.

OBJECTIVE

Using the vocabulary of Total Observed error and biological variation and examples from veterinary medicine, a review of the theory and application of the EMU, dispersion, and the methods for deriving an appropriate reporting interval recommended by Hawkins and Badrick, is presented.

CONCLUSIONS

By addressing the way that MU is communicated to users of laboratory results, the laboratory enables users to better understand the potential uncertainty associated with reported results, helps to prevent over and under-interpretation of data, and improves diagnostic accuracy and patient care.

Analytical Sigma metrics: A review of Six Sigma implementation tools for medical laboratories

Sigma metrics have become a useful tool for all parts of the quality control (QC) design process. Through the allowable total error model of laboratory testing, analytical assay performance can be judged on the Six Sigma scale. This not only allows benchmarking the performance of methods and instruments on a universal scale, it allows laboratories to easily visualize performance, optimize the QC rules and numbers of control measurements they implement, and now even schedule the frequency of running those controls.

Bias in analytical chemistry: A review of selected procedures for incorporating uncorrected bias into the expanded uncertainty of analytical measurements and a graphical method for evaluating the concordance of reference and test procedures

The Evaluation of measurement data - Guide to the Expression of Uncertainty in Measurement (GUM) provides the framework for evaluating measurement uncertainty. The preferred GUM approach for addressing bias assumes that all systematic errors are identified and corrected at an early stage in the measurement process. We review some procedures for treating uncorrected bias and its inclusion into an overall uncertainty statement. When bias and its uncertainty are recognised as metrological states independent of scatter in the test results, the uncertainty of the reference and uncertainty of the bias can be equated. The net standard uncertainty of a test result is the root-sum-square of the standard uncertainty of the bias and the standard uncertainty of measurements on the test. Since an incomplete and therefore potentially erroneous formula is often used for estimating bias standard uncertainty, we propose an alternative calculation. We next propose a graphical method using a simple algorithm that quantifies the discrepancy between the results of a test measurement and the corresponding reference value, in terms of the percentage overlap of two probability density functions. We propose that bias should be corrected wherever possible and we illustrate this approach using the graphical method. Even though this review is focused principally on analytical chemistry and medical laboratory applications, much of the discussion is applicable to all areas of metrology.

External quality assessment schemes for glucose measurements in Germany: factors for successful participation, analytical performance and medical impact

Background:

Determination of blood glucose concentration is one of the most important measurements in clinical chemistry worldwide. Analyzers in central laboratories (CL) and point-of-care tests (POCT) are both frequently used. In Germany, regular participation in external quality assessment (EQA) schemes is mandatory for laboratories performing glucose testing.

Methods:

Glucose testing data from the two German EQAs “Reference Institute for Bioanalytics” (RfB) and “INSTAND – Gesellschaft zur Förderung der Qualitätssicherung in medizinischen Laboratorien” (Instand) were analyzed from 2012 to 2016. Multivariable odds ratios (OR) for the probability to reach a “good” result were calculated. Imprecision and bias were determined and clinical risk of measurement errors estimated.

Results:

The device employed was the most important variable required for a “good” performance in all EQAs. Additional participation in an EQA for CL automated analyzers improved performance in POCT EQAs. The reciprocal effect was less pronounced. New participants performed worse than experienced participants especially in CL EQAs. Imprecision was generally smaller for CL, but some POCT devices reached a comparable performance. Large lot-to-lot differences occurred in over 10% of analyzed cases. We propose the “bias budget” as a new metric to express the maximum allowable bias that still carries acceptable medical risk. Bias budgets were smallest and clinical risks of errors greatest in the low range of measurement 60–115 mg/dL (3.3–6.4 mmol/L) for most devices.

Conclusions:

EQAs help to maintain high analytical performances. They generate important data that serve as the foundation for learning and improvement in the laboratory healthcare system.

Uncertainty of Measurement in Laboratory Medicine

An adequate assessment of the measurement uncertainty in a laboratory medicine is one of the most important factors for a reliable interpretation of the results. A large number of standards and guidelines indicate the need for a proper assessment of the uncertainty of measurement results in routine laboratory practice. The available documents generally recommend participation in the proficiency schemes/ external quality control, as well as the internal quality control, in order to primarily verify the quality performance of the method. Although all documents meet the requirements of the International Standard, ISO 15189, the standard itself does not clearly define the method by which the measurement results need to be assessed and there is no harmonization in practice regarding to this. Also, the uncertainty of measurement results is the data relating to the measured result itself, but all factors that influence the interpretation of the measured value, which is ultimately used for diagnosis and monitoring of the patient's treatment, should be taken into account. So in laboratory medicine, an appropriate assessment of the uncertainty of the measurement results should have the ultimate goal of reducing diagnostic uncertainty. However, good professional laboratory practice and understanding analytical aspects of the test for each individual laboratory is necessary to adequately define the uncertainty of measurement results for specific laboratory tests, which helps to implement good clinical practice. Also, setting diagnoses in medicine is a decision with a certain degree of uncertainty, rather than statistically and mathematically calculated conclusion.

Uncertainty in Measurement: Procedures for Determining Uncertainty With Application to Clinical Laboratory Calculations

In Part II of this review we consider the very common case of multiple inputs to a measurement process. We derive, using only elementary steps and the basic mathematics covered in Part I, the formula for the propagation of uncertainties from the inputs to the output. The Gaussian density distribution is briefly explained, since an understanding of this distribution is needed for the determination of so-called expanded uncertainties at the end of a measurement process. The propagation formula in general involves correlations among the inputs, although in many cases these correlations can be considered negligible. Correlations, however, need to be taken into account in related matters such as line-fitting and have particular relevance to method comparisons. These topics are addressed briefly. We next discuss the important question of bias and its incorporation into the expression of uncertainty. We present, finally, six real-world cases in clinical chemistry where uncertainty in the estimated value of the measurand is calculated using the propagation formula.

Measurement uncertainty of blood ethanol concentration in drink-driving cases in an emergency laboratory

Introduction

The quality of blood ethanol concentration (BEC) determination is important because of its legal ramifications. Measurement uncertainty provides quantitative information about the quality and reliability of test results. In this study, we aim to calculate the measurement uncertainty for the ethanol test in our laboratory measured with a Synchron Systems Ethanol assay kit by employing an enzymatic rate method on the Beckman-Coulter Olympus AU400 auto analyzer (Beckman Coulter Inc, Melville, USA).

Materials and methods

The measurement uncertainty values were calculated in accordance to the Nordtest guidelines. All vehicle drivers involved in a traffic accident were retrospectively inspected for the BEC test conducted during July to December 2016 in our emergency laboratory.

Results

A 1034 vehicle drivers had their BEC tested. The results for 181 drivers were > 0.50 g/L and reported as positive. The serum ethanol concentration in those showing a positive result was 2.04 ± 1.01 g/L, over four times the legal limit. The median BEC in those showing a negative result was 0.03 (IQR: 0.03) g/L. The expanded uncertainty obtained was 19.74%. When measurement uncertainty values were added to the results of the 1034 drivers who were retrospectively screened, eight vehicle drivers had results with 95% confidence intervals that exceeded the legal limit 0.50 g/L.

Conclusions

The BEC test results for vehicle drivers with values close to legal limits should be reported as the obtained ethanol concentration with corresponding measurement uncertainty.

Measurement Uncertainty in Spine Bone Mineral Density by Dual Energy X-ray Absorptiometry

Background

The purpose of this study was to calculate the measurement uncertainty of the process of bone mineral density (BMD) analysis using dual energy X-ray absorptiometry with traceability.

Methods

Between March 2015 and October 2016, among healthy participants in their 20s and 30s, the study included those who had not taken calcium, vitamin D supplements and steroids and were without a history of osteoporosis, osteopenia and diseases related to osteoporosis. Relational expression of the model was established based on Guide to the Expression of Uncertainty in Measurements and Eurachem and the uncertainty from each factor was evaluated.

Results

The combined standard uncertainty was 0.015, while the expanded uncertainty was 0.0298. The factor-specific standard uncertainties that occurred in the process of measuring BMD were 0.72% for the calibration curve, 0.9% for the internal quality control (IQC) using Aluminum Spine Phantom, 0.58% for European Spine Phantom (ESP), and 0.9% for the inspector precision (IP).

Conclusions

The combined standard uncertainty of the spine BMD corrected with ESP was 0.015 when measured at one time and targeting one participant. The uncertainties of the accuracy of the IQC and the IP were higher than that of the other factors. Therefore, there will be a need for establishment of protocols to lower these uncertainties.