Quality Control and Uncertainties

The backbone of the BALTIC atlas is the data holding in the lOW database, collected 

In BALTIC, following a recommendation of the Intergovernmental Oceanographic Commission (IOC) endorsed by its 11th Session in Mexico City, March 1979, the measured parameters are labeled by GF3 codes (IOC, 1987). The subsetprocessedinBALTIC is given in Table 11.3. The related data holdings are described by the Tables 11.4 and 11.5 and in Fig. 11.2. 

The data sets provided to BALTIC by the national partner institutes as well as by the data banks of ICES (and BED) were quality controlled (QC) data. The measuring devices used 

Before including a particular data sample into the BALTIC data set, it was checked for plausibility one more time. This check includes 

Each particular value found outside the plausibility bounds was manually inspected in its measurement context in order to decide between alternative decisions to be taken discard the value as wrong, correct the value for obvious reasons, or expand the confidence interval. The percentage of discarded data was surprisingly high with regard to the fact that the data were already quality-checked before. Frequent typical cases were data on land, obviously invalid, or erratic data, temperatures significantly below the freezing point and extremely low salinities likely measured inside fjords or near the shore. This detailed and careful procedure took most of the efforts in the BALTIC project development from 2000 to 2007. 

---

To provide a practical, understandable and common way of measurement uncertainty calculations, mainly based on already existing quality control and validation data covering all uncertainty sources in a integral way... 

At the conclusion of the study, surplus test materials are offered for sale as reference materials for quality control and method development purposes. These materials have been studied for homogeneity and stability and have been assigned values for specific analytes. The more recently produced materials (since July 1999) have traceable assigned values with rigorously evaluated uncertainties. Older materials have consensus-based assigned values. 

The values were then quite usable for quality control and identification especially when the system was calibrated with standards at frequent intervals. The main residual source of variation was then batch-to-batch differences between columns, although these differences have been reduced in recent years, and uncertainties in the preparation of the mobile phase, which can be reduced by close control of the protocols used. 

A broad range of technical requirements is important in a laboratory quality system. These include aspects such as the selection of appropriately qualified and experienced personnel sampling, sample handling and preparation laboratory accommodation and environmental conditions equipment and reagents calibration reference standards and reference materials traceability (of standards and of samples) the selection or development, validation, and control of methods estimation of the uncertainty of measurements reporting of results and quality control and proficiency testing. 

Quality assurance and quality control are the unifying themes of this chapter as well as the previous one. Uni- and multivariate statistics provide the quantitative framework to measure reliability and uncertainty, two indispensable descriptors of any forensic data. Uncertainty and errors are inevitable in even the best-designed and validated method quality control and statistics provide the methods to characterize and incorporate those errors into data analysis and reporting. If problems arise during an anal5reis, quality assurance and quality control procedures and practices serve to raise a red flag and allow the forensic chemist to take corrective action. Thus, QA/QC are as much a part of forensic chemistry as are stoichiometry, organic chemistry, and sample preparation. 

In the case of these plastic plaques the major component of uncertainty of measurement is the levelness contribution. It is incumbent upon the spectroscopist, quality control and quality assurance personnel, colorists, and laboratory managers to assess the total measurement system uncertainty under their conditions. 

The approaches to analytical procedures uncertainty prognosis are developed. The correctness of these approaches is confirmed in 3rd and 4th rounds of pharmaceutical laboratories inter-laboratory testing ( Phamia-Test program of State Inspection for Quality Control of Medicines, Ministry of Health of Ukraine). 

A quality control laboratory had a certain model of HPLC in operation. One of the products that was routinely run on the instrument contained two compounds, A and B, that were quantitated in one run at the same detector wavelength setting. At an injection volume of 20 /tL, both compounds showed linear response. The relatively low absorption for compound B resulted in an uncertainty that was just tolerable, but an improvement was sought. 

By now, it should be clear what role RMs play in measurement science. This puts great responsibility on the producers of RMs, as they must see how to satisfy the requirements set impHcitly or explicitly by the users regarding matrix, parameters, uncertainty, and traceability. Laboratories use RMs often as a quality control measure, but it this obviously only vahd if the RM is produced under proper conditions. 

In addition to the requirements regarding traceability of measurement results, the measurement methods employed should represent "state-of-the-art in the particular field. Failing to do so would lead to a reference material with an uncertainty that has become too large to serve as a quality control. The better the methods perform in terms of uncertainty and traceability, the better the reference material will serve the interests of the (potential) users. 

Of the four strategies given above, the best condition for obtaining independent data for quality control (QC) are satisfied when INAA and RNAA results are compared, because the use of RNAA dramatically improves the selectivity of signal measurement and eliminates or greafiy reduces the measurement uncertainty sotuces, such as spectral interferences. A variety of radiochemical separations and... 

Some of the intended categories of use of radioisotopic reference material have been reviewed recently by Fajgelj et al. (1999). They include assignment of property values, establishing the traceability of a measurement result, determining the uncertainty of a measurement result, calibration of an apparatus, assessment of a measurement method, use for recovery studies and use for quality control purposes. It should be noted however that, in general, natural matrix reference materials are not recommended for calibration purposes. This should preferably be done with pure chemical forms of the element labelled with the isotope of interest. Calibrated isotopic sources of this kind are available from a number of commercial suppliers and are not the subject of this review. 

Ingamells CO, PiTARD FF (1986) Applied Geochemical Analysis, pp L-84.Wiley, New York. International Federation of Clinical Chemistry (IFCC) (1978) Expert Panel on Nomenclature and Principles of Quality Control in Clinical Chemistry. Clin Chim Acta 83 L89F-202F. International Organization for Standardization (ISO) (1993) Guide to the expression of uncertainty. Geneva. 

On most occasions CRMs are used as Quality Control materials, rather than as calibrations . As outlined above, this common application adds significantly to the user s uncertainty budget, since at a minimum it is necessary to consider at least two independent measurement events (Um). so increasing the combined uncertainty of the results. Again this process rapidly increases the combined uncertainty with increasing complexity of the analytical system and so the usefulness of a control analysis may be downgraded when a correct uncertainty budget is formulated. 

The principles of quality assurance are commonly related to product and process control in manufacturing. Today the field of application greatly expanded to include environmental protection and quality control within analytical chemistry itself, i.e., the quality assurance of analytical measurements. In any field, features of quality cannot be reproduced with any absolute degree of precision but only within certain limits of tolerance. These depend on the uncertainties of both the process under control and the test procedure and additionally from the expense of testing and controlling that may be economically justifiable. 

Internal quality control (IQC) is one of a number of concerted measures that analytical chemists can take to ensure that the data produced in the laboratory are of known quality and uncertainty. In practice this is determined by comparing the results achieved in the laboratory at a given time with a standard. IQC therefore comprises the routine practical procedures that enable the analyst to accept a result or group of results or reject the results and repeat the analysis. IQC is undertaken by the inclusion of particular reference materials, control materials , into the analytical sequence and by duplicate analysis. 

---