Human errors and reliability of test results in analytical chemistry

The present paper is a review of the main theoretical and technical aspects of human error treatment (error modelling, reduction and quantification) as applied in aviation, engineering, medicine and other fields. The aim of the review is to attract the attention of analysts and specialists in metrology and quality in chemistry to the human error problem and its influence on the reliability of test results of chemical composition and associated measurement uncertainty. Therefore, the subject of human error is interpreted in the review in application to the conditions of a chemical analytical laboratory.     

Human being as a part of measuring system influencing measurement results

The role of human being as a part of a measuring system in a chemical analytical laboratory is discussed. It is argued that a measuring system in chemical analysis includes not only measuring instruments and other devices, reagents and supplies, but also a sampling inspector and/or analyst performing a number of important operations. Without this human contribution, a measurement cannot be carried out. Human errors, therefore, influence the measurement result, i.e., the measurand estimate and the associated uncertainty. Consequently, chemical analytical and metrological communities should devote more attention to the topic of human errors, in particular at the design and development of a chemical analytical/test method and measurement procedure. Also, mapping human errors ought to be included in the program of validation of the measurement procedure (method). Teaching specialists in analytical chemistry and students how to reduce human errors in a chemical analytical laboratory and how to take into account the error residual risk, is important. Human errors and their metrological implications are suggested for consideration in future editions of the relevant documents, such as the International Vocabulary of Metrology (VIM) and the Guide to the Expression of Uncertainty in Measurement (GUM).     

Performance of quantitative analyses by liquid chromatography–electrospray ionisation tandem mass spectrometry: from external calibration to isotopomer-based exact matching

Liquid chromatography–tandem mass spectrometry (LC-MS/MS) is a versatile coupling system which combines both selectivity and sensitivity and certainty. Hence, it is generally considered as the most reliable technique to quantify chemical compounds in complex matrices. In the present paper, we evaluate the performance of LC-MS/MS methods for the quantification of 3-nitrotyrosine in human urine in order to point out its dependence on the design of the quantification method, and emphasize the role of matrix effects in the performance. We compare external and internal calibrations, isotope dilution and isotopomer-based exact matching. The role of both sample preparation and multiple transitions monitoring is particularly addressed.     

Scenarios of human errors and their quantification in multi-residue analysis of pesticides in fruits and vegetables

A total of 54 scenarios of human errors in multi-residue analysis of pesticides in fruits and vegetables are mapped in a routine monitoring laboratory. The likelihood and severity of the errors and the ability of components of the laboratory quality system to prevent/block human errors, as well as the effectiveness of the quality system as a whole, are evaluated using expert judgments. It is shown that human errors in sampling are prevented less by the quality system than other steps of the analysis. Training and supervision are found as the quality system components having the highest priorities. Effectiveness of the quality system obtained a good score value of 71 % (in comparison with an ideal system preventing human errors with 100 % effectiveness). Variability of the score values resulting from changes of the expert judgments is discussed.     

Improving the quality of potentiometric pH measurements

Like all experimentally determined physical and chemical properties, pH measurements are affected by the limited precision and accuracy of the measurement procedures. Fundamental studies of pH standards, based on measurement of the potential of an electrochemical cell without transference, known as the Harned cell, containing a platinum–hydrogen electrode and a silver–silver chloride reference electrode, indicate that vapour condensation phenomena on potentiometric cell walls not immersed in the thermostatic bath are a major source of error in assessment of pH values. In this work a study was conducted on phthalate buffer, 0.05 mol kg⁻¹ KHPhth, and results are reported for the effect of this phenomenon on the assignment of pH values and on their corresponding uncertainties. Identification and quantification of this effect constitute an original contribution to improvement of the primary method of pH measurement and, therefore, more rigorous pH (PS) values.     

Isotope pattern vector based tandem mass spectral data calibration for improved peptide and protein identification

Tandem mass spectra contain noisy peaks which make peak picking for peptide identification difficult. Moreover, all spectral peaks can be shifted due to systematic measurement errors. In this paper, a novel use of an isotope pattern vector (IPV) is proposed for denoising and systematic measurement error prediction. By matching the experimental IPVs with the theoretical IPVs of candidate fragment ions, true ionic peaks can be identified. Furthermore, these identified experimental IPVs and their corresponding theoretical IPVs are used in an optimization process to predict the systematic measurement error associated with the target spectrum. In return, the subsequent spectral data calibration based on the predicted systematic measurement error enhances the data quality. We show that such an integrated denoising and calibration process leads to significantly improved peptide and protein identification. Different from the commonly employed chemical calibration methods, our IPV‐based method is a purely computational method for individual spectra analysis and globally optimizes the use of spectral data. Copyright © 2009 John Wiley & Sons, Ltd.     

Sensitivity of model-based quantitative FTIR to instrumental and spectroscopic database error sources

We report a theoretical study of the error sources associated with the quantification of gas-phase FTIR spectra using synthetic calibrations. A forward model was constructed based on a Bruker IFS66 FTIR spectrometer, modelling the instrument line shape from theory and taking line parameters from a line-by-line spectral database. Default values were set in the forward model for an ‘ideal’ system where the spectrometer is perfectly aligned and all spectroscopic parameters are exactly known. Using a re-iterative non-linear least squares routine input values were perturbed allowing assessment of the model sensitivity to each parameter. Using the sensitivity information and knowledge of the possible absolute uncertainties (e.g. the accuracy to which a user might be able to axially align the field stop aperture) the possible quantitative contribution of each parameter was ranked in the following order (greatest first); field stop aperture axial alignment, line intensity, line centre, air-broadening line width, pressure, lateral field stop aperture/collimating optic alignment, pathlength, temperature. Whilst this ranking was specific to the type of measurement modelled it was discussed which elements would remain consistent for different measurement acquisition set-ups and which would alter. Consequently, it was proposed that the error analysis presented here could be used to determine which parameters in a forward model should be set as variables (i.e. those with the highest potential error contribution) and which should remain fixed in attempting to minimise the errors in using optimisation routines for the purpose of gas quantification.     

Sensitivity of model-based quantitative FTIR to instrumental and spectroscopic database error sources

We report a theoretical study of the error sources associated with the quantification of gas-phase FTIR spectra using synthetic calibrations. A forward model was constructed based on a Bruker IFS66 FTIR spectrometer, modelling the instrument line shape from theory and taking line parameters from a line-by-line spectral database. Default values were set in the forward model for an ‘ideal’ system where the spectrometer is perfectly aligned and all spectroscopic parameters are exactly known. Using a re-iterative non-linear least squares routine input values were perturbed allowing assessment of the model sensitivity to each parameter. Using the sensitivity information and knowledge of the possible absolute uncertainties (e.g. the accuracy to which a user might be able to axially align the field stop aperture) the possible quantitative contribution of each parameter was ranked in the following order (greatest first); field stop aperture axial alignment, line intensity, line centre, air-broadening line width, pressure, lateral field stop aperture/collimating optic alignment, pathlength, temperature. Whilst this ranking was specific to the type of measurement modelled it was discussed which elements would remain consistent for different measurement acquisition set-ups and which would alter. Consequently, it was proposed that the error analysis presented here could be used to determine which parameters in a forward model should be set as variables (i.e. those with the highest potential error contribution) and which should remain fixed in attempting to minimise the errors in using optimisation routines for the purpose of gas quantification.     

Considerations for quantification of lipids in nerve tissue using matrix-assisted laser desorption/ionization mass spectrometric imaging

Matrix-assisted laser desorption/ionization (MALDI) mass spectrometric imaging is a technique that provides the ability to identify and characterize endogenous and exogenous compounds spatially within tissue with relatively little sample preparation. While it is a proven methodology for qualitative analysis, little has been reported for its utility in quantitative measurements. In the current work, inherent challenges in MALDI quantification are addressed. Signal response is monitored over successive analyses of a single tissue section to minimize error due to variability in the laser, matrix application, and sample inhomogeneity. Methods for the application of an internal standard to tissue sections are evaluated and used to quantify endogenous lipids in nerve tissue. A precision of 5% or less standard error was achieved, illustrating that MALDI imaging offers a reliable means of in situ quantification for microgram-sized samples and requires minimal sample preparation.     

Quantitative sphingosine measurement as a surrogate for total ceramide concentration—preclinical and potential translational applications

Biomarkers are an increasingly important constituent of the drug development process, offering the potential of increased efficiency through reduced compound attrition and earlier proof of mechanism and/or efficacy. Assays developed for compound screening that can be directly translated for clinical trials are especially valuable, but their successful adoption requires a careful balance between assay performance and implementation costs. One such ‘fit‐for‐purpose’ biomarker assay, the indirect measurement of pharmacological modulation of sphingolipid biosynthesis and disposition, is presented here. Among spingolipids, numerous ceramide species are readily detectable in different lipoprotein fractions of mammalian plasma, but their parallel quantification can be prohibitively expensive and time consuming. Ceramides differ in their fatty acid moiety, which is readily removed by hydrolysis, yielding a common sphingosine derivative, the measurement of which serves as an indicator of total ceramide. When followed by liquid chromatography tandem mass spectrometry (LC/MS/MS) for detection, robust analyte quantification becomes relatively straightforward. The practical utility of a method developed to be fit for the purpose of rapidly and quantitatively measuring treatment‐induced variations in total ceramide from hamster plasma and individual lipoprotein fractions is described. With a linear calibration range from 0.003 to 33.4 μm sphingosine, precision and accuracy error in plasma‐based quality controls spiked with ceramides was less than 15%. The specificity of the assay for ceramides was also assessed. The simplicity of the method would allow for its potential translation to other preclinical species, as well as for clinical applications in later‐stage drug development. Copyright © 2009 John Wiley & Sons, Ltd.     

Real-time polymerase chain reaction-based approach for quantification of the pat gene in the T25 Zea mays event

In Europe, a growing interest for reliable techniques for the quantification of genetically modified component(s) of food matrixes is arising from the need to comply with the European legislative framework on novel food products. Real-time polymerase chain reaction (PCR) is currently the most powerful technique for the quantification of specific nucleic acid sequences. Several real-time PCR methodologies based on different molecular principles have been developed for this purpose. The most frequently used approach in the field of genetically modified organism (GMO) quantification in food or feed samples is based on the 5'-3'-exonuclease activity of Taq DNA polymerase on specific degradation probes (TaqMan principle). A novel approach was developed for the establishment of a TaqMan quantification system assessing GMO contents around the 1% threshold stipulated under European Union (EU) legislation for the labeling of food products. The Zea mays T25 elite event was chosen as a model for the development of the novel GMO quantification approach. The most innovative aspect of the system is represented by the use of sequences cloned in plasmids as reference standards. In the field of GMO quantification, plasmids are an easy to use, cheap, and reliable alternative to Certified Reference Materials (CRMs), which are only available for a few of the GMOs authorized in Europe, have a relatively high production cost, and require further processing to be suitable for analysis. Strengths and weaknesses of the use of novel plasmid-based standards are addressed in detail. In addition, the quantification system was designed to avoid the use of a reference gene (e.g., a single copy, species-specific gene) as normalizer, i.e., to perform a GMO quantification based on an absolute instead of a relative measurement. In fact, experimental evidences show that the use of reference genes adds variability to the measurement system because a second independent real-time PCR-based measurement must be performed. Moreover, for some reference genes no sufficient information on copy number in and among genomes of different lines is available, making adequate quantification difficult. Once developed, the method was subsequently validated according to IUPAC and ISO 5725 guidelines. Thirteen laboratories from 8 EU countries participated in the trial. Eleven laboratories provided results complying with the predefined study requirements. Repeatability (RSDr) values ranged from 8.7 to 15.9%, with a mean value of 12%. Reproducibility (RSDR) values ranged from 16.3 to 25.5%, with a mean value of 21%. Following Codex Alimentarius Committee guidelines, both the limits of detection and quantitation were determined to be <0.1%.     

Determination of telmisartan in human blood plasma: Part II: Liquid chromatography-tandem mass spectrometry method development, comparison to immunoassay and pharmacokinetic study

A new liquid chromatography/atmospheric pressure chemical ionization-tandem mass spectrometry (LC/APCI-MS/MS) method with on-line sample clean-up for the determination of telmisartan in human blood plasma is presented. This technique is compared to a previously introduced enzyme-linked immunosorbent assay (ELISA), where fluorescence is used as detection method. For the LC/MS method applying an internal calibration via a deuterated internal standard, the limit of detection was 0.3 ng/mL, the limit of quantification was 0.9 ng/mL and the linear range extended from 0.9 to 1000 ng/mL. Forty-eight plasma samples from four healthy volunteers were analyzed in a pharmacokinetic study to obtain data for the method comparison. As a result, these two new and independent analytical methods for the determination of telmisartan in human blood plasma proved to yield comparable results for the amount of analyte.     

Development of a resolution metric for comprehensive two-dimensional chromatography

A new resolution metric for two-dimensional chromatography is proposed and tested. This resolution measurement is based on the concept of the (one-dimensional) valley-to-peak ratio, which has been adapted and modified for two-dimensional chromatography. Two questions are considered related to the computation of the resolution of a given (two-dimensional) peak. First, the concept of peak neighbourhood is revised, since it changes drastically from one- to two-dimensional chromatography. In a chromatogram resulting from a two-dimensional analysis, one peak may be surrounded by more than two neighbouring peaks. However, the neighbouring peaks can be remote from the peak or some interfering peaks may be in between. In these cases, it is not meaningful to compute the resolution between them. A method is proposed to determine whether a resolution measurement between two two-dimensional peaks is reasonable. Second, a measurement of the valley-to-peak ratio in two-dimensional chromatography is proposed. The measurement is based on the concept of the saddle point (which is defined for two-dimensional surface plots). A study of the correlation of the valley-to-peak ratio with the error obtained for quantification is presented. The new metric can be used as an estimator of the quantification errors. Also, valley-to-peak ratios can be calculated for one or more target peak(s) to estimate the separation quality of the entire chromatogram. This makes the proposed measurement suitable for optimisation purposes. Although the algorithm was developed for GC x GC, preliminary studies suggested that its application to other two-dimensional separation methods (e.g. LC x LC) should only require minor modification (if any).     

Comparison of stable-isotope labeling with amino acids in cell culture and spectral counting for relative quantification of protein expression

Protein quantification is one of the principal goals of mass spectrometry (MS)-based proteomics, and many strategies exist to achieve it. Several approaches involve the incorporation of a stable-isotope label using either chemical derivatization, enzymatically catalyzed incorporation of (18)O, or metabolic labeling in a cell or tissue culture. These techniques can be cost or time prohibitive or not amenable to the biological system of interest. Label-free techniques including those utilizing integrated ion abundance and spectral counting offer an alternative to stable-isotope-based methodologies. Herein, we present the comparison of stable-isotope labeling of amino acids in cell culture (SILAC) with spectral counting for the quantification of human embryonic stem cells as they differentiate toward the trophectoderm at three time points. Our spectral counting experimental strategy resulted in the identification of 2641 protein groups across three time points with an average sequence coverage of 30.3%, of which 1837 could be quantified with more than five spectral counts. SILAC quantification was able to identify 1369 protein groups with an average coverage of 24.7%, of which 1027 could be quantified across all time points. Within this context we further explore the capacity of each strategy for proteome coverage, variation in quantification, and the relative sensitivity of each technique to the detection of change in relative protein expression.     

Time‐resolved Fourier transform infrared/attenuated total reflection spectroscopy for the measurement of molecular diffusion in polymers

Over the past 2 decades, the use of time‐resolved Fourier transform infrared/attenuated total reflection (ATR) spectroscopy for the measurement of diffusion in polymers has grown. ATR is a powerful technique for the measurement of diffusion in polymers because it is an in situ technique that is relatively inexpensive, provides reliable short‐time data, and provides a wealth of information at the molecular level. This article highlights the technique and its application to numerous studies, ranging from the diffusion of drugs in human skin to chemical warfare agents in barrier materials. In addition to these topics, recent studies with ATR to quantify and model molecular interactions during the diffusion process are reviewed. In the future, the ATR technique may have an impact on a variety of emerging fields in which diffusion in polymers plays an important role, such as fuel cells, membrane separation, sensors, and drug delivery. © 2003 Wiley Periodicals, Inc. J Polym Sci Part B: Polym Phys 41: 2794–2807, 2003     

Convenient and rapid determination of cimetidine in human plasma using perchloric acid-mediated plasma protein precipitation and high-performance liquid chromatography

This study demonstrates the analysis of cimetidine in human plasma with HPLC using a simplified sample preparation by protein precipitation with perchloric acid. Plasma cimetidine concentration was determined by plotting peak height ratio of cimetidine to ranitidine (internal standard, IS) against cimetidine concentrations in plasma. The cimetidine and ranitidine peaks were completely separated and no interference from plasma was observed. The lower limit of quantification (LLOQ) of the method was established at 0.1 microg/mL with a precision of 4.3% and a relative error of 1.9%. The average analytical recovery was >90% over the range of cimetidine concentrations (0.1-15.0 microg/mL). The linearity of calibration curve was excellent (r(2) > 0.999). The within- and between-day precision and accuracy, expressed as the coefficients of variation and relative error, were found to be less than 5%. Compared with previously reported methods, the analytical technique for cimetidine determination in human plasma presented here demonstrates comparable accuracy and precision, an acceptable analysis time, shorter and simpler sample preparation, and a reduced need for complicated equipment. The method presented here is simple and rapid, and the precision and sensitivity are appropriate for the determination of cimetidine in plasma in pharmacokinetic studies.     

Estimation of trace vapor concentration-pathlength in plumes for remote sensing applications from hyperspectral images

A novel approach for quantification of chemical vapor effluents in stack plumes using infrared hyperspectral imaging are presented and examined. The algorithms use a novel application of the extended mixture model to provide estimates of background clutter in the on-plume pixel. These estimates are then used iteratively to improve the quantification. The final step in the algorithm employs either an extended least-squares (ELS) or generalized least-squares (GLS) procedure. It was found that the GLS weighting procedure generally performed better than ELS, but they performed similarly when the analyte spectra had relatively narrow features. The algorithms require estimates of the atmospheric radiance and transmission from the target plume to the imaging spectrometer and an estimate of the plume temperature. However, estimates of the background temperature and emissivity are not required which is a distinct advantage. The algorithm effectively provides a local estimate of the clutter, and an error analysis shows that it can provide superior quantification over approaches that model the background clutter in a more global sense. It was also found that the estimation error depended strongly on the net analyte signal for each analyte, and this quantity is scenario-specific.     

The separation of 99Tc from low and medium-level radioactive wastes and its determination by inductively coupled plasma mass spectrometry

A chemical separation method has been developed for the determination of (99)Tc in various types of radioactive wastes. Such a method includes (i) fusion with NaOH, (ii) extraction in a column containing methyltrioctylammonium chloride, (iii) extraction by solvent with N-benzoyl-N-phenylhydroxylamine and, (iv) measurement by inductively-coupled plasma mass spectrometry (ICP-MS). From the performance standpoint, the recovery of (99)Tc, using (99m)Tc as a yield tracer, is higher than 70%. This analytical method, as developed, ensures effective decontamination with respect to the radionuclides, insofar the decontamination factors are greater than 10(+5), whenever the residual activity may be measured. Taking into account a 3sigma counting error, the detection limit obtained with the ICP-MS technique is 1.9 mBq/ml; the method enabling hence to detect activities as low as 0.3 Bq/g, with analysed samples of 0.2 g and a radiochemical yield of 70%. Studies have been dedicated to the (99)Tc measurement, using the electrothermal vaporization ICP-MS technique, which lowers the detection limit by a factor 10, with the standard solution (0.3 pg/ml), compared with the previous ICP-MS technique.     

Investigation into accurate mass capability of matrix-assisted laser desorption/ionization time-of-flight mass spectrometry, with respect to radical ion species

Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOFMS) has been shown to be an effective technique for the characterization of organometallic, coordination, and highly conjugated compounds. The preferred matrix is 2-[(2E)-3-(4-tert-butylphenyl)-2-methylprop-2-enylidene]malononitrile (DCTB), with radical ions observed. However, MALDI-TOFMS is generally not favored for accurate mass measurement. A specific method had to be developed for such compounds to assure the quality of our accurate mass results. Therefore, in this preliminary study, two methods of data acquisition, and both even-electron (EE+) ion and odd-electron (OE+.) radical ion mass calibration standards, have been investigated to establish the basic measurement technique. The benefit of this technique is demonstrated for a copper compound for which ions were observed by MALDI, but not by electrospray (ESI) or liquid secondary ion mass spectrometry (LSIMS); a mean mass accuracy error of -1.2 ppm was obtained.