Traceability system for elemental analysis

A complete metrological traceability system for measurement results of chemical analysis was set up. Core components are pure substances (national standards) characterised at the highest metrological level, primary solutions prepared from these pure substances and secondary solutions deduced from the primary solutions and intended for sale. The relative uncertainty of the element mass fraction of the primary substances and solutions is < 0.01 and < 0.05%, respectively. For the certification of transfer solutions and for stability testing, a precision measurement method for element contents has been developed by means of optical emission spectrometry (ICP OES) by which uncertainties between 0.1 and 0.05% can be achieved. The dissemination to field laboratories is effected with the aid of a calibration laboratory of the German Calibration Service (DKD) which certifies the element content of the secondary solutions with an uncertainty ≤ 0.3%. Calibration with these solutions enables the user to establish traceability of his measurement results to the International System of Units (SI). Currently, the system comprises Cu, Fe, Bi, Ga, Si, Na, K, Sn, W, and Pb.     

Modelling uncertainty in a concentration range

Combined uncertainty modelling in a concentration range is an important task for laboratories. Despite several models and regression methods reported, unsolved problems remain, including variability and type of distribution of combined uncertainty estimations and the influence of these on modelling. Intralaboratory data of eight trace elements in natural waters by flame atomic absorption spectrometry and interlaboratory data from some ISO standards were used as an experimental basis. Starting from these and applying the bootstrap technique, high relative variability of combined uncertainty estimations (second-order uncertainty) was found, but normal distributions or distributions with small deviations from normality were encountered. Linear and/or variance models are appropriate for modelling the analyzed data when ordinary weighted least-squares or repeat median robust regression methods are applied. The influences of combined uncertainty variability on modelling and the effect of points that do not follow the general trend of the remaining points are discussed. Some other guidelines are offered.     

Alternative calculation of measurement uncertainty with global approach in food microbiology

Calculation of measurement uncertainty is a requirement for all laboratories accredited to ISO/IEC 17025 including those carrying out microbiological analyses. Today, calculation of measurement uncertainty in microbiological analyses using precision data according to global approach principles is widely recognized by the microbiologists due to difficulties in quantification of individual uncertainty sources. In food microbiology, precision data obtained from different samples usually show over-dispersion, and the use of over-dispersed data may result in large variance. The current ISO standard on measurement uncertainty in food microbiology proposes the use of log-transformed precision data to overcome this problem. This paper proposes an alternative procedure to calculate the measurement uncertainty in food microbiology using non-logarithmic precision data. The calculations used in this procedure based on relative range of duplicate analyses can be applied to intra-laboratory reproducibility data obtained from microbiological analyses of which duplicate results show relatively low variation.     

PT-WFD: the network of PT providers to support the implementation of the European water framework directive

This paper introduces a newly formed network of proficiency testing (PT) providers, who are active and experienced in the field of PTs with relevance to the water framework directive (WFD). The objective of this new network is to support the implementation of the WFD in Europe, by providing PT schemes that meet the specific requirements of the WFD and which are run and evaluated in a harmonised way. The main value of creating such a network lies in the potential to foster the harmonised performance evaluation of European laboratories involved in WFD monitoring analyses, thus enhancing the equivalence of WFD monitoring data obtained throughout Europe. In its first year the network organised two joint PT rounds, one for volatile organic substances (VOC) and another for selected pesticides from the list of Priority Substances, with greater than 100 water analysis labs from a number of European countries participating.     

Comparison of multianalyte proficiency test results by sum of ranking differences, principal component analysis, and hierarchical cluster analysis

Sum of ranking differences (SRD) was applied for comparing multianalyte results obtained by several analytical methods used in one or in different laboratories, i.e., for ranking the overall performances of the methods (or laboratories) in simultaneous determination of the same set of analytes. The data sets for testing of the SRD applicability contained the results reported during one of the proficiency tests (PTs) organized by EU Reference Laboratory for Polycyclic Aromatic Hydrocarbons (EU-RL-PAH). In this way, the SRD was also tested as a discriminant method alternative to existing average performance scores used to compare mutlianalyte PT results. SRD should be used along with the z scores—the most commonly used PT performance statistics. SRD was further developed to handle the same rankings (ties) among laboratories. Two benchmark concentration series were selected as reference: (a) the assigned PAH concentrations (determined precisely beforehand by the EU-RL-PAH) and (b) the averages of all individual PAH concentrations determined by each laboratory. Ranking relative to the assigned values and also to the average (or median) values pointed to the laboratories with the most extreme results, as well as revealed groups of laboratories with similar overall performances. SRD reveals differences between methods or laboratories even if classical test(s) cannot. The ranking was validated using comparison of ranks by random numbers (a randomization test) and using seven folds cross-validation, which highlighted the similarities among the (methods used in) laboratories. Principal component analysis and hierarchical cluster analysis justified the findings based on SRD ranking/grouping. If the PAH-concentrations are row-scaled, (i.e., z scores are analyzed as input for ranking) SRD can still be used for checking the normality of errors. Moreover, cross-validation of SRD on z scores groups the laboratories similarly. The SRD technique is general in nature, i.e., it can be applied to any experimental problem in which multianalyte results obtained either by several analytical procedures, analysts, instruments, or laboratories need to be compared.

Inter‐laboratory comparison: Quantitative surface analysis of thin Fe‐Ni alloy films

An international interlaboratory comparison of the measurement capabilities of four National Metrology Institutes (NMIs) and one Designated Institute (DI) in the determination of the chemical composition of thin Fe‐Ni alloy films was conducted via a key comparison (K‐67) of the Surface Analysis Working Group of the Consultative Committee for Amount of Substance. This comparison was made using XPS (four laboratories) and AES (one laboratory) measurements. The uncertainty budget of the measured chemical composition of a thin alloy film was dominated by the uncertainty of the certified composition of a reference specimen which had been determined by inductively coupled plasma mass spectrometry using the isotope dilution method. Pilot study P‐98 showed that the quantification using relative sensitivity factors (RSFs) of Fe and Ni derived from an alloy reference sample results in much more accurate result in comparison to an approach using RSFs derived from pure Fe and Ni films. The individual expanded uncertainties of the participants in the K‐67 comparison were found to be between 2.88 and 3.40 atomic %. The uncertainty of the key comparison reference value (KCRV) calculated from individual standard deviations and a coverage factor (k) of 2 was 1.23 atomic %. Copyright © 2011 John Wiley & Sons, Ltd.     

Reliability of measurement uncertainty estimates for forensic analyses: evidence from recent proficiency tests

The ability of forensic laboratories to reliably estimate the uncertainty of their reported results is assessed by means of data from recent proficiency tests in which participants reported both their results and the associated uncertainties. Topics covered include the determination of blood alcohol, breath alcohol, and controlled substances—services commonly provided by forensic laboratories. It is seen that a statistically significant number of laboratories underestimate their own uncertainties, based on the incidence of the laboratories’ own reported coverage intervals failing to include the participant mean or assigned concentration of the test material.     

Quantitative analysis of trace levels of surface contamination by X‐ray photoelectron spectroscopy. Part I: Statistical uncertainty near the detection limit

We discuss the problem of quantifying common sources of statistical uncertainties for analyses of trace levels of surface contamination by using X‐ray photoelectron spectroscopy. We examine the propagation of error for peak‐area measurements by using common forms of linear and polynomial background subtraction including the correlation of points used to determine both background and peak areas. This correlation has been neglected in previous analyses, but we show that it contributes significantly to the peak‐area uncertainty near the detection limit. We introduce the concept of relative background subtraction variance (RBSV) that quantifies the uncertainty introduced by the method of background determination relative to the uncertainty of the background area itself. The uncertainties of the peak area and atomic concentration and of the detection limit are expressed using the RBSV, which separates the contributions from the acquisition parameters, the background‐determination method, and the properties of the measured spectrum. These results are then combined to find acquisition strategies that minimize the total measurement time needed to achieve a desired detection limit or atomic‐percentage uncertainty for a particular trace element. Minimization of data‐acquisition time is important for samples that are sensitive to X‐ray dose and also for laboratories that need to optimize throughput.     

Interlaboratory comparison for the determination of five residual organochlorine pesticides in ginseng root samples by gas chromatography

An interlaboratory comparison study for the determination of 5 residual organochlorine pesticides (hexachlorobenzene and 4 hexachlorocyclohexane isomers) in ginseng root was performed. This program [Asia Pacific Laboratory Accreditation Cooperation (APLAC) T049] was the first of its kind for an herbal matrix and involved the participation of 70 laboratories from 26 countries worldwide. Consensus mean values were computed statistically from the reported results, which were eventually used to assess the performance of individual laboratories in terms of the z-scores. The distribution of analytical data was found to be widespread, with standard deviation ranging from 43.9 to 55.9%, and the result patterns obtained were similar to those residue pesticide programs using other matrixes. Although the estimation of measurement uncertainty is a crucial requirement for all quantitative tests for laboratories that meet the requirements of International Organization for Standardization/International Electrotechnical Commisssion (ISO/IEC) 17025, some laboratories in this program had difficulties and showed unfamiliarity with respect to that quality criterion. It was recommended that laboratories review and rectify the situation promptly so that they would have a better understanding of measurement uncertainty or the test service provided.     

Comparison of “bottom-up” and “top-down” strategies for the estimation of the uncertainty in organic elemental analysis

The estimation of uncertainty in organic elemental analysis for C, H, N and S is reported. Both “bottom up” and “top down” strategies are used for uncertainty calculations. The bottom up approach used the results of C, H, N, and S obtained from the homogeneity study of two pure chemicals (toluene-4-sulfonamide and 4(6)-methyl-2-thiouracil). Two calibration systems, K factor and calibration curve, were applied in this study and no significant differences were obtained. For the “top down” approach, we used the data obtained from a proficiency test on both pure chemicals from among 45 Spanish laboratories. Both approaches are compared and discussed below.     

Relative sensitivity of combined standard uncertainty to changes induced in uncertainty components

This work provides a straightforward and rigorous solution to the problem of evaluating and expressing relative changes in combined uncertainty resulting from relative changes in the uncertainty components. The acquired information is essential to optimize the measurement method, indicating the most significant components to work with. This approach works for any discrete relative change in the uncertainties of the inputs, whether positive or negative, and was developed for the cases of independent and correlated quantities. Examples of both cases are treated. The results agree with that provided by direct calculation and are consistent with the ideas in EA 4/16. We propose the use of the expression “sensitivity of combined standard uncertainty to changes in uncertainty components” as a uniform quantitative reference to the individual importance of the components that form the combined uncertainty. We also discuss the reason to avoid the use of the expression “contribution of an uncertainty component to the combined uncertainty.”     

A global approach to method validation and measurement uncertainty

The enforcement of legal limits for food safety raises the question of decision-making in the context of uncertain measurements. It also puts the question of demonstrating that measurement technique that is used is fit for the purpose of controlling legal limits. A recent European Commision (EC) decision gives some indications how to deal with this question. In the meantime, the implementation of quality systems in analytical laboratories is now a reality. While these requirements deeply modified the organization of the laboratories, it has also improved the quality of the results. The goal of this communication is to describe how two fundamental requirements of ISO 17025 standard, i.e. validation of the methods and estimation of the uncertainty of measurements, can give a way to check whether an analytical method is correctly fit for the purpose of controlling legal limits. Both these requirements are not independent and it will be shown how they can be combined. A recent approach based on the “accuracy profile” of a method was applied to the determination of acrylamide and illustrates how uncertainty can be simply derived from the data collected for validating the method. Moreover, by basing on the β-expectation tolerance interval introduced by Mee [Technometrics (1984) 26(3): 251–253], it is possible to unambiguously demonstrate the fitness for purpose of a method. Remembering that the expression of uncertainty of the measurement is also a requirement for accredited laboratories, it is shown that the uncertainty can be easily related to the trueness and precision issuing from the data collected to build the method accuracy profile. The example presented here consists in validating a method for the determination of acrylamide in pig plasma by liquid chromatography–mass spectromery (LC–MS). Concentrations are expressed as mg/l and instrumental response is peak surface. The calibration experimental design included 5×5×2 measurements and namely consisted in preparing duplicate standard solutions at five concentration levels ranging from 10 to about 5000 mg/l. This was repeated for 5 days. The validation experimental design was similar.     

Precision in analytical measurements: Expected values and consequences in geochemical analyses

Over the years the authors have developed an empirical formula for the relative standard deviation among laboratories (RSD_R, %) as a function of concentration C, expressed as a decimal fraction: RSD_R=2^{(1–0.5log C)}2C^(–0.1505). This formula predicts that starting with a concentration of 100% (pure materials; C=1.00), RSD_R will be 2%, increasing by a factor of 2 for each decrease in C of 2 orders of magnitude. It has been a useful guide to acceptable precision in the fields of agricultural, pharmaceutical, and nutritional chemical analysis. It has now been found that this formula is also useful for interpreting the quality of reported analyses of various reference materials available to geochemists. If a ratio is calculated of the found RSD_R value to the RSD_R value calculated from the formula, an easily interpreted limiting parameter is generated. A ratio >2 provides a rational reference point for interpreting the uncertainty of the analytical portion of the total variability of geochemical concentration estimates. Such a high ratio indicates excessive among-laboratories precision, reflecting large systematic errors on the part of one or more laboratories. The formula quantitates the concept of the transformation of the individual systematic errors of the laboratories into the random error of the group.     

Laboratory effects models for interlaboratory comparisons

The statistical analysis of results from inter-laboratory comparisons (for example Key Comparisons, or Supplemental Comparisons) produces an estimate of the measurand (reference value) and statements of equivalence of the results from the participating laboratories. Methods to estimate the reference value have been proposed that rest on the idea of finding a so-called consistent subset of laboratories, that is, eliminating allegedly outlying participants. We propose an alternative statistical model that accommodates all participant data and incorporates the dispersion of the measurement values obtained by different laboratories into the total uncertainty of the various estimates. This model recognizes the fact that the dispersion of values between laboratories often is substantially larger than the measurement uncertainties provided by the participating laboratories. We illustrate the methods on data from key comparison CCQM–K25.     

Measurement uncertainty in microbiology

Testing laboratories wishing to comply with the requirements of ISO/IEC 17025:1999 need to estimate uncertainty of measurement for their quantitative methods. Many microbiological laboratories have had procedures available for monitoring variability in duplicate results generated by laboratory analysts for some time. These procedures, however, do not necessarily include all possible contributions to uncertainty in the calculations. Procedures for estimating microbiological method uncertainty, based on the Poisson distribution, have been published but, at times, the procedures can either underestimate uncertainty or require laboratories to undertake considerable experimental studies and more complex statistical calculations. This paper proposes procedures for estimating uncertainty of measurement in microbiology, whereby routine laboratory quality control data can be analyzed with simple statistical equations. The approaches used in these procedures are also applied to published data and examples, demonstrating that essentially equivalent results can be obtained with these procedures.     

Software support for the Nordtest method of measurement uncertainty evaluation

Software support for the Nordtest method of measurement uncertainty evaluation is described. According to the Nordtest approach, the combined measurement uncertainty is broken down into two main components??the within-laboratory reproducibility (intermediate precision) s _Rw and the uncertainty due to possible laboratory bias u(bias). Both of these can be conveniently estimated from validation and quality control data, thus significantly reducing the need for performing dedicated experiments for estimating detailed uncertainty contributions and thereby making uncertainty estimation easier for routine laboratories. An additional merit of this uncertainty estimation approach is that it reduces the danger of underestimating the uncertainty, which continues to be a problem at routine laboratories. The described software tool??MUkit (measurement uncertainty kit)??fully reflects the versatility of the Nordtest approach: it enables estimating the uncertainty components from different types of data, and the data can be imported using a variety of means such as different laboratory data systems and a dedicated web service as well as manual input. Prior to the development of the MUkit software, a laboratory survey was carried out to identify the needs of laboratories related to uncertainty estimation and other quality assurance procedures, as well as their needs for a practical tool for the calculation of measurement uncertainty.     

Use of uncertainty estimates as reported by participants in proficiency testing for the evaluation of their results: pros and cons

The aim of this paper is to study the measurement uncertainties reported in proficiency tests (PTs) using examples from PTs in the environmental sector and to compare the obtained measurement uncertainty estimates using different approaches. In addition, the paper focusses on the differences between the z-score and the zeta score. Since the year 2000, the Finnish Environment Institute has asked participants to report analytical methods as well as measurement uncertainties in connection with PT results. The measurement uncertainties of the assigned value have also been evaluated. On the basis of the results, the measurement uncertainties reported by the participants varied greatly. Participants often reported underestimated measurement uncertainties, but overestimated uncertainties were also reported. At the moment, it seems as if performance assessment should be based on the z-score because of a number of significant over- and underestimated measurement uncertainties. The zeta score should be used for information and educational purposes mainly.     

Determination of standard sample purity using the high-precision 1H-NMR process

This paper discusses the technique for high-precision quantification using ¹H-NMR to determine the purity of analytical standard samples. The procedure described is based on the use of internal reference samples in an ¹H NMR experiment in our laboratories. The sample preparation and all relevant NMR parameters were optimized for minimum uncertainty. The validation of accuracy and precision was performed by comparing different certified reference materials. It was shown that the high-precision measurement is applicable even for relatively small sample amounts down to 2.5 mg. The relative combined uncertainty of measurement was found to be 0.15%. Two different approaches for uncertainty calculation were compared; a complete uncertainty budget was calculated.     

Ion chromatographic precision measurement procedure for electrolytes in human serum: validation with the aid of primary measurement procedures

In order to make analytical measurement results traceable to the SI units in the field of clinical chemistry, an ion chromatographic (IC) measurement procedure has been developed which allows the amount of substance of the four so-called electrolytes Na, K, Mg and Ca as well as that of Li to be determined efficiently in human serum and with high accuracy. The IC measurement procedure was validated using primary measurement procedures confirmed by international comparison measurements and is proposed for use as a transfer standard when comparing measurements with clinical reference laboratories. The solutions used for calibration were gravimetrically prepared from pure substances (salts). Their chemical compositions had been iteratively fitted to those of the samples. The serum samples were mineralized by microwave-assisted digestion. The following relative expanded uncertainties for the average elemental contents were obtained: Li 0.4%, Na 0.14%, K 0.6%, Mg 0.8% and Ca 0.4%.     

An error model for evaluation and interpretation of proficiency testing

Data from proficiency testing can be used to increase our knowledge of the performance of populations of laboratories, individual laboratories and different measurement methods. To support the evaluation and interpretation of results from proficiency testing an error model containing different random and systematic components is presented. From a single round of a proficiency testing scheme the total variation in a population of laboratories can be estimated. With results from several rounds the random variation can be separated into a laboratory and time component and for individual laboratories it is then also possible to evaluate stability and bias in relation to the population mean. By comparing results from laboratories using different methods systematic differences between methods may be indicated. By using results from several rounds a systematic difference can be partitioned into two components: a common systematic difference, possibly depending on the level, and a sample-specific component. It is essential to distinguish between these two components as the former may be eliminated by a correction while the latter must be treated as a random component in the evaluation of uncertainty. Received: 20 November 2000 Accepted: 3 January 2001