326
8 Analytical Chemistry and Quality
Like traceability, validation has lately become a buzzword even though it has
always been an intrinsic, obvious concept with regard to the typical duties of the
analytical laboratory. Validation is the formal demonstration that a system has
performed as expected and continues to do so - or that an object has possessed
given characteristics and retains them. In the analytical domain, validation is
applicable to samples, data and methods. Validating a method entails formally
demonstrating that it features given levels of the basic analytical parameters
(accuracy, precision, sensitivity, selectivity, range, linearity, throughput, robustness) and is fit for the intended purpose within the framework of the analytical
problem to be solved (see Sect. 4.6).
The robustness of an analytical method is a property of a high practical
significance. It is related to constancy in the results in response to slight changes
in the operating conditions, in such a way that the relative significance of each
variable and the ranges over which it can fluctuate without altering the results in
a statistically significant manner can be established. Robustness is related to
precision (both change in the same direction but under different conditions),
reliability (the ability to withstand slight alterations in the operating conditions
with no change in the results) and transferability (the ability to lead to the
same results when used by different laboratories). Robustness is essential for
validation.
Quality of the results in terms of accuracy rests on two essential cornerstones
of traceability, viz. correct, rigorous calibration and proper use of reference
materials. Both are systematically discussed in Chap. 3. Chemometrics is of
great help with the former. Reference materials, as primary standards, are a key
to quality of results. While the availability of highly pure materials has grown
dramatically over the last few years, certified real samples continue to be the
greatest weakness of Analytical Chemistry as barely 5 -10 % of those needed are
currently available. Much effort need still be devoted to obtaining certified
reference materials to meet such a heavy demand.
8.S Analytical Quality Control
Quality control activities are an essential element of Quality Assurance. They
are planned, well-documented actions performed by laboratory staff and involve direct inspection of the laboratory organization, work or tools, as well as
of the results produced. These activities are examined by quality assessment
systems, which derive pertinent corrective actions. All should be properly recorded.
Quality Control possesses essentially quantitative connotations. As a rule, it is
based on comparisons of data (preferably accompanied by their uncertainties)
under the umbrella of Statistics (Chemometrics). They are thus required
numerical references for quality control that are provided by data from reference
materials (RMs) or certified reference materials (CRMs), the results of various
methods applied to the same sample, etc.
8 Analytical Chemistry and Quality
Like traceability, validation has lately become a buzzword even though it has
always been an intrinsic, obvious concept with regard to the typical duties of the
analytical laboratory. Validation is the formal demonstration that a system has
performed as expected and continues to do so - or that an object has possessed
given characteristics and retains them. In the analytical domain, validation is
applicable to samples, data and methods. Validating a method entails formally
demonstrating that it features given levels of the basic analytical parameters
(accuracy, precision, sensitivity, selectivity, range, linearity, throughput, robustness) and is fit for the intended purpose within the framework of the analytical
problem to be solved (see Sect. 4.6).
The robustness of an analytical method is a property of a high practical
significance. It is related to constancy in the results in response to slight changes
in the operating conditions, in such a way that the relative significance of each
variable and the ranges over which it can fluctuate without altering the results in
a statistically significant manner can be established. Robustness is related to
precision (both change in the same direction but under different conditions),
reliability (the ability to withstand slight alterations in the operating conditions
with no change in the results) and transferability (the ability to lead to the
same results when used by different laboratories). Robustness is essential for
validation.
Quality of the results in terms of accuracy rests on two essential cornerstones
of traceability, viz. correct, rigorous calibration and proper use of reference
materials. Both are systematically discussed in Chap. 3. Chemometrics is of
great help with the former. Reference materials, as primary standards, are a key
to quality of results. While the availability of highly pure materials has grown
dramatically over the last few years, certified real samples continue to be the
greatest weakness of Analytical Chemistry as barely 5 -10 % of those needed are
currently available. Much effort need still be devoted to obtaining certified
reference materials to meet such a heavy demand.
8.S Analytical Quality Control
Quality control activities are an essential element of Quality Assurance. They
are planned, well-documented actions performed by laboratory staff and involve direct inspection of the laboratory organization, work or tools, as well as
of the results produced. These activities are examined by quality assessment
systems, which derive pertinent corrective actions. All should be properly recorded.
Quality Control possesses essentially quantitative connotations. As a rule, it is
based on comparisons of data (preferably accompanied by their uncertainties)
under the umbrella of Statistics (Chemometrics). They are thus required
numerical references for quality control that are provided by data from reference
materials (RMs) or certified reference materials (CRMs), the results of various
methods applied to the same sample, etc.