Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Guide To Method Validation For Quantitative Analysis in Chemical Testing Laboratories 17025 PDF 36 Pages 349KB

Download as pdf or txt
Download as pdf or txt
You are on page 1of 27
At a glance
Powered by AI
The document provides guidance on method validation for quantitative analysis in chemical testing laboratories in accordance with ISO/IEC 17025.

The document provides guidance on interpreting the requirements for method validation according to ISO/IEC 17025. It aims to ensure consistent assessment of laboratories and place Irish laboratories on par with others internationally. However, it does not cover sampling or measurement uncertainty.

The main steps involved in method validation include appointing an individual responsible for the method, specifying requirements, determining performance characteristics, documenting the validation work in a report, and assessing if requirements can be met.

Guide to Method Validation for Quantitative PS15

Analysis in Chemical Testing Laboratories


(ISO 17025)

TABLE OF CONTENTS

1. Foreword 2

2. Scope 3

3. References 4

4. Definitions 6

5. Scale of method validation 12

6. Review of performance characteristics for a method of analysis 17

7. Follow-up of method validation 26

8 Laboratory competence 26

9 Laboratory documentation 27

10 INAB Contact 27

PS15 Guide to Method Validation for Quantitative Analysis in Chemical Testing Laboratories
Issue 4 February 2016 Page 1 of 27
1. FOREWORD

With the introduction of EN ISO/IEC 17025, the requirements governing the documentation of methods,
including method selection and validation of methods, have been amplified. The level of documentation
required by INAB in the field of chemical testing still depends on whether standard methods, modified
methods or a laboratorys own methods are involved.

As there is scope for interpretation in any standard, it has long been desirable from many quarters to have
guidance drawn up to ensure more uniform interpretation of the requirements for validation.

In drawing up this document, it has proved necessary not only to interpret the requirements of the
standard with respect to laboratories in the sector but also to provide guidance on the assessment of
laboratories. This document is generally to be regarded as a guideline for chemical analysis laboratories,
but also contains a few sections that are purely aimed at INAB assessors in this area. Where sections are
purely aimed at INAB assessors, this is apparent from the text. In addition, the document contains
instructions and examples of how method validation can be carried out.

PS15 Guide to Method Validation for Quantitative Analysis in Chemical Testing Laboratories
Issue 4 February 2016 Page 2 of 27
2. SCOPE

INAB guidelines concern interpretation of relevant sections of the accreditation criteria, drawn up with a
view to ensuring consistent case administration and assessment of applicant and accredited laboratories
and to place the treatment of Irish laboratories on a par with laboratories in other countries.

This guideline lays down INABs interpretation of point 5.4 and in particular point 5.4.5 of EN ISO/IEC
17025:1999 (as revised by ISO/IEC 17025:2005) for chemical analysis laboratories, concerning the
laboratories validation of methods of analysis adopted and to the assessment of the results of such
validation work. The guideline has been drawn up on the basis of current international standards and
guidelines in the sector.

Measurement uncertainty (uncertainty budgets), which should be part of method validation, is not dealt
with at length in this document. Reference is in this respect made to the GUM document [15] and the
Eurachem/CITAC document [16].

In this issue, the guideline covers exclusively quantitative, analytical methods; however, it does not cover
sampling in connection with the performance of these methods.

Where specific validation requirements are laid down by Regulatory Authorities (in legislation, etc.)
compliance with the specified validation protocols will be accepted by INAB as demonstration of fitness
for purpose.

A few sections purely describe guidance for INABs assessors and the document is therefore also called
Internal Guidance. In addition, the document contains guidance and examples of how method validation
can be performed (section 6). The guideline cannot, however, as such be regarded as a manual for
method validation in connection with the laboratories compliance with the requirements of ISO 17025.

The guideline applies to all applicant and accredited chemical analysis testing laboratories.

PS15 Guide to Method Validation for Quantitative Analysis in Chemical Testing Laboratories
Issue 4 February 2016 Page 3 of 27
3. REFERENCES

[1] McNaught & Wilkinson: Compendium of Chemical Terminology, 2nd ed., IUPAC, 1997 (ISBN 0-86542-
684-8).
Also known as IUPACs Gold Book
Available on: www.iupac.org/publications/compendium/index.html

[2] International vocabulary of basic and general terms in metrology (VIM) (BIPM, IEC, IFCC, ISO, IUPAC,
IUPAP, OIML, 2nd ed., 1993).

[3] ISO/IEC 17025: General Requirements for the competence of testing and calibration laboratories,
2nd edition, May 2005.

[4] CITAC/Eurachem Guide: Guide to Quality in Analytical Chemistry, An Aid to Accreditation. Prepared
jointly by CITAC (The Cooperation on International Traceability in Analytical Chemistry) and
Eurachem (A Focus for Analytical Chemistry in Europe), 2002.
a) The guide is based on the previous documents CITAC Guide 1 and Eurachem Guidance
Document No. 1/WELAC Guidance Document No. WGD 2, with account being taken of the
change from EN 45001 to ISO/IEC 17025 as accreditation standard.
b) The guide is based in particular on the CITAC document, with relatively few amendments
and additions being made.
c) The previous document has been abolished as an official EA document (previously EA-
4/05); this document has (for the time being) not been granted equivalent status.

[5] IUPAC: Harmonized guidelines for single-laboratory validation of methods of analysis (IUPAC
Technical report). Resulting from the Symposium on Harmonization of Quality Assurance Systems
for Analytical Laboratories, Budapest, Hungary, 4-5 November 1999, held under the sponsorship of
IUPAC, ISO & AOAC INTERNATIONAL. Prepared for publication by Michael Thompson, Stephen L.R.
Ellison and Roger Wood. Published in Pure Appl. Chem, Vol. 74, No. 5, 2002 (pp. 835 855).

[6] IUPAC: Harmonized guidelines for the use of recovery information in analytical measurement
(Technical report). Resulting from the symposium on harmonization of internal quality assurance
systems for analytical laboratories, Orlando DC, USA, 4-5 September 1996, held under the
sponsorship of IUPAC, ISO & AOAC International.

[7] Eurachem Guide: The Fitness for Purpose of Analytical Methods. A Laboratory Guide to Method
Validation and Related Topics, 1st edition, Eurachem, Dec. 1998.

[8] EN ISO 9000: Quality control systems Basic principles and glossary. .

[9] Youden, W.J. & Steiner, E.H.: Statistical Manual of the AOAC, Association of Official Analytical
Chemists, 1975.
[10] ISO 5725-1: Accuracy (trueness and precision) of measurement methods and results. Part 1: General
principles and definitions, 1995.

[11] ISO 3534-1:1993: Statistics Vocabulary and symbols. Part 1: Probability and general statistical
terms, ISO, 1993.

[12] ISO Guide 30:1992, Terms and definitions used in connection with reference materials, ISO 1992.

[13]

PS15 Guide to Method Validation for Quantitative Analysis in Chemical Testing Laboratories
Issue 4 February 2016 Page 4 of 27
ISO/IEC 17043:2010 Conformity assessment -- General requirements for proficiency testing

[14]

[15] BIPM, IEC, IFCC, ISO, IUPAC, IUPAP, OIML: Guide to the expression of uncertainty in measurement,
ISO, Geneva, 1993.
This document is often referred to as GUM. It is interpreted in the chemical, analytical sector in
the form of the document Quantifying Uncertainty in Analytical Measurement [16].

[16] Eurachem, CITAC: Quantifying Uncertainty in Analytical Measurement, 2nd edition, 2000
This document is designated by Eurachem and CITAC as QUAM:2000 P1.
EA refers to the document as an official interpretation of the accreditation requirements.

[17] Poulsen, Holst & Molin Christensen: Method evaluation and quality control of chemical
measurement processes, The National Institute of Occupational Health, Copenhagen 1993. (ISBN
87-7534-440-8).

[18] Funk, Dammann & Donnevert: Quality Assurance in Analytical Chemistry, VCH Verlagsgesellschaft
mbH, Weinheim, Germany, 1995 (ISBN 3-527-28668).

[19] IUPAC: Nomenclature in evaluation of analytical methods including detection and quantitation
capabilities. Recommendations, 1995. Published in Pure Appl. Chem. Vol. 67 (1995) pp. 1699
1723.

[20] Directive 96/23/EC Measure to Monitor certain Substances and Residues thereof in Live Animals and
Animal Products

[21] INAB Policy on Flexible Scopes, PS11

PS15 Guide to Method Validation for Quantitative Analysis in Chemical Testing Laboratories
Issue 4 February 2016 Page 5 of 27
4. DEFINITIONS

The terms used in this guideline are defined here in order to ensure a clear and unambiguous
understanding of the terms. Where possible, internationally accepted definitions have been taken as the
starting point; however, there are cases where these are not clear and so a definition of our own has been
chosen for this guideline.
Many of the terms may also have other more or less sector-specific designations, which are acceptable
if the laboratory has made the meaning clear within the context of validation.
It should be pointed out that notes connected with the definition in the quoted reference (NOTE(S)) have
been included, where possible. In some cases, certain comments important for understanding the
definition in question in relation to this guideline have also been added (COMMENT(S)).

4.1 Method of Analysis


Method describing in detail a procedure of chemical analysis with a view to determining/identifying one
or more analytical parameters.
COMMENT:
a) Method of analysis is the term generally accepted and adopted in laboratories, but there is no
official definition of the term.
b) Measurement method is the logical sequence of operations, generically described, used for
the performance of measurements.
c) Measurement procedure is the set of operations, described in detail, used in the performance
of certain measurements in accordance with a given measurement method.

4.2 Analytical Parameter / Analyte


The element, chemical compound or property to be detected and quantified by the performance of a
method of analysis.
COMMENT:
In this guideline, the term analytical parameter is used not only for purely chemical substances and
compounds but also for magnitudes such as dry matter, pH, conductivity, various collective parameters,
etc.

4.3 Bias [10]


The difference between the calculated mean of the measurement results and an accepted reference
value.
NOTE:
Bias is the total systematic error, as opposed to random error. There may be one or more systematic
error components contributing to bias. A substantial departure from the accepted reference value is
manifested in a high level of bias.

4.4 Blank Value


A blank value is obtained as a result of analysis of a specimen which does not, as far as possible, contain
the analytical parameter(s) in question - and with the implementation of the method of analysis.
COMMENTS:
a) The quoted type of blank value, also called test blank, is the magnitude referred to in the rest of
this guideline.
b) Unlike a test blank, a reagent blank is also adopted in certain contexts [see 4.30].

PS15 Guide to Method Validation for Quantitative Analysis in Chemical Testing Laboratories
Issue 4 February 2016 Page 6 of 27
4.5 Limit of Detection
There are at least two different interpretations in relation to the definition of the term limit of
detection:
1. The limit of detection defined on the basis of the desire to avoid a specimen not containing analytical
parameters erroneously being ascribed a concentration (false positive), and at the same time avoid a
specimen containing the analytical parameter being erroneously reported as undetected (false
negative). The definition of the limit of detection then becomes the lower content of an analytical
parameter that can be detected with a specifically defined level of confidence (e.g. 95, 99 or 99.7%)
using a given method of analysis.
2. The limit of detection defined on the basis of a desire for a certain maximum uncertainty in the result.
The definition of the limit of detection then becomes the minimum content that can be measured
with a reasonable statistical certainty [9].
COMMENTS:
a) Expressed in statistical terms, definition 1 above can also be formulated as The result of a
measurement performed with a given measurement procedure, for which the probability of an
analytical false negative is , given that the probability of an analytical false positive is . IUPAC
recommends that and should in principle be assigned the value 0.05.
b) If a measurement gives a value below the level of detection, the result can be quoted as less than
the limit of detection NOT as zero.

4.6 Degrees of Freedom


The number of independent determinations (estimates) of a given statistical magnitude (e.g. mean or
standard deviation) that can be performed on the basis of a given data set.
COMMENT:
The greater the number of degrees of freedom, the better the statistical basis for the
determination of the magnitude in question.

4.7 Sensitivity [2,9]


The change in response on a measuring instrument divided by the corresponding change in stimulus.
NOTE:
Stimulus may for example be the amount of analyte present. Sensitivity may depend on the value for the
stimulus in question. Although this definition clearly concerns measuring instruments, it can also be used
in connection with methods of analysis as a whole, with account also being taken of other factors such as
specimen preparation, etc.
COMMENT:
Sensitivity can often be expressed by the gradient of the linear calibration curve.

4.8 Recovery
4.8 Performance Parameter/Characteristic
Characteristics for a method of analysis determining its possibilities and limitations (measuring range,
specificity, precision, etc.)

4.9 Trueness [10]


Level of conformity between the mean for a population of measurement results and the accepted
reference value.

PS15 Guide to Method Validation for Quantitative Analysis in Chemical Testing Laboratories
Issue 4 February 2016 Page 7 of 27
NOTE:
Trueness is normally expressed as the bias of the method.

4.10 Limit of Quantitation


The lowest concentration that can be determined with acceptable laboratory reproducibility and
trueness.
COMMENT:
The limit of quantitation forms the lower level for the measuring range.

4.11 Laboratory Reproducibility


Precision under reproducible conditions, within the laboratory.

4.12 Linearity
The ability to induce a signal (response) that is directly proportional to the concentration of the given
analytical parameter.
COMMENTS:
a) Determination of linearity is used among other things in connection with the formulation of a
calibration curve.
b) For some methods, calibrations may give non-linear fits e.g. quadratic fit

4.13 Linear Range [1]


Concentration range within which the intensity of the signal obtained is directly proportional to the
concentration of the analytical parameter inducing the signal.
COMMENT:
The original text [1] uses the slightly specific expression species for what has here been translated as
analytical parameters/analyte.

4.14 Matrix [1]


All components in the specimen, with the exception of the analytical parameter(s) in question.

4.15 Measuring Range


The concentration range within which the analytical parameter in question can be determined with
specifically determined trueness and precision.
NOTES:
a) The measuring range may certainly contain a broader concentration range than a linear range, with
the link between response and concentration not in all cases being regarded as linear.
b) The measuring range must be defined in relation to the analyte, matrix and relevant range of use.
c) Trueness and precision must be acceptable throughout the validated measuring range.

4.16 Measurement Uncertainty [2]


A parameter linked to the measurement result and characterising the spread of the values that can
reasonably be added to the measurement magnitude.
PS15 Guide to Method Validation for Quantitative Analysis in Chemical Testing Laboratories
Issue 4 February 2016 Page 8 of 27
NOTES:
a) This parameter may, for example, be a standard deviation (or a given multiple thereof) or half the
breadth of an interval that has a quoted confidence level.
b) Measurement uncertainty generally comprises many components. Some of these can be assessed
on the basis of the statistical distribution of the results of measurement series and can be
characterised by experimental standard deviations. Other components, which are also
characterised by standard deviations, are assessed on the basis of assumed probability
distributions based on experience or on other information.
c) It should be understood that the measurement result is the best estimate of the value of the
measurement magnitude and that all uncertainty components, including those originating from
systematic effects (e.g. components associated with corrections and reference standards),
contribute to the spread.
[This definition is the same as in the Guide to the expression of uncertainty in measurement
[15] in which the logical reasons are detailed (particularly in section 2.2.4 and Annex D)].

4.17 Accuracy [10]


Degree of conformity between a measurement result and the accepted reference value.
NOTE:
Used in connection with a range of measurement results, the expression accuracy covers a combination
of random components and a common systematic error or bias component
COMMENT:
In GUM [15], the definition is expressed in relation to the true value of the measurement magnitude
instead of the accepted reference value.

4.18 Precision [10]


Degree of conformity between independent measurement results obtained under prescribed conditions.
NOTES:
a) Precision depends only on the distribution of the random error, and is unrelated to a true value, a
conventional true value or an accepted reference value.
b) Precision is often expressed as a lack of precision and is calculated as the spread of the
measurement results. Less precision corresponds to greater spread.
c) The phrase independent measurement results means that the results are obtained in a manner
such that they are unaffected by any previous results for the same or similar measurement
objects. Quantitative measures for precision depend critically on the prescribed measurement
conditions. Repeatability and reproducibility conditions are extreme points for prescribed
measurement conditions.

4.19 Repeatability [10]


Precision under repeatable conditions.

4.20 Repeatable Conditions [10]


Conditions in which independent measurement results are obtained by the same method on identical
specimens in the same laboratory, by the same operator, with the same equipment and within short
intervals of time.

PS15 Guide to Method Validation for Quantitative Analysis in Chemical Testing Laboratories
Issue 4 February 2016 Page 9 of 27
4.21 Repeatability Standard Deviation [10]
The spread in the distribution of measurement results obtained under repeatable conditions.
NOTES:
a) Repeatability standard deviation is an expression of the variation in the distribution of
measurement results under repeatable conditions.
b) Accordingly, repeatability variance and repeatability variation coefficient could be defined and
used as an expression of the variation of measurement results under repeatable conditions.

4.22 Reproducibility [10]


Precision under reproducible conditions.

4.23 Reproducible Conditions [10]


Conditions under which independent measurement results are obtained with the same method on
identical specimens in different laboratories by different operators with different equipment.

4.24 Reproducibility Standard Deviation [10]


The spread in the distribution of measurement results obtained under reproducible conditions.
NOTES:
a) Reproducibility standard deviation is an expression of the variation in the distribution of
measurement results under reproducible conditions.
b) Accordingly, reproducibility variance and reproducibility variation coefficient could be defined
as an expression of the variation in the distribution of measurement results under reproducible
conditions.

4.25 Robustness/Ruggedness [8]


The degree of independence of the method of analysis from minor deviations in the experimental
conditions of the method of analysis.

4.26 Selectivity [4]


The selectivity of a method of analysis refers to the degree to which the method of analysis is usable for
determining the presence of specific analytical parameters in a complex mixture (matrix) without
interference from other analytical parameters in the mix.
COMMENTS:
a) The ability of a method of analysis to measure only what it is intended to measure (the specificity
of the method).
b) A lack of selectivity in a method of analysis is manifested, among other things, in the interference
which can hamper determination of the analytical parameter sought.
c) Selectivity should be viewed in relation to the analytical parameter in question whether this is
an individual substance or a group of substances.

PS15 Guide to Method Validation for Quantitative Analysis in Chemical Testing Laboratories
Issue 4 February 2016 Page 10 of 27
4.27 Validation [8]
Confirmation via the provision of objective evidence* - that the requirements for specifically intended
use or application have been met.
NOTES:
a) The expression validated is used to denote the corresponding status.
b) In the case of validation, the use/application conditions may be actual or simulated.
c) *Objective evidence is defined in [8] as data substantiating the presence or veracity of
something (Objective evidence can be furnished by observation, measurement, testing or in
other ways).

COMMENT:
The difference between the terms validation and verification in terms of their practical use
is discussed at greater length in section 5 of this guideline.

4.28 Recovery
Detection of a known amount of an analytical parameter added to a specimen and included throughout
the method of analysis. After deducting any detected content of the analytical parameter in question in
the original specimen without addition, the recovery percentage can be calculated as a percentage of the
amount added.
COMMENTS:
a) The aim must be for the addition to be at a level in proportion to the content of the specimen, so
different dilutions need not be performed, but also so large that the difference between the
content of the specimen and the specimen plus added analytical parameter can be determined
with an uncertainty no greater than for determination of the content of the specimen.
b) In validation, recovery tests are used to monitor the overall effect of different procedure steps, for
example evaporation losses, extraction, clean-up, but should not be used purely to assess any one
of the effects in isolation.
c) Recovery tests are used to assess bias.

4.29 Reagent Blank


Reagent blank determination means the complete analytical procedure applied without the test portion
or using an equivalent amount of suitable solvent in place of the test portion [20].

PS15 Guide to Method Validation for Quantitative Analysis in Chemical Testing Laboratories
Issue 4 February 2016 Page 11 of 27
5 SCALE OF METHOD VALIDATION method validation
The laboratory must, in connection with any testing within the accreditation area, have undertaken a
conscious assessment of the fitness of the method of analysis used for the purpose of the given test.
Fitness must be adapted to the applicants needs and be documented by appropriate validation, with a
suitable validation programme (a protocol) being set up and implemented for each matrix type for which
the method of analysis is intended to be used.

There are a number of general, overriding requirements concerning how extensive such a validation
programme should be (included in ISO/IEC 17025). It is also apparent from this, however, that the scale of
a validation programme as a whole and the scale of the assessment of the individual performance
characteristics must be laid down in relation to the specific situation.
This section of the guideline examines a number of important aspects concerning definition of the
relevant scale of method validation. The individual performance characteristics and their importance in
connection with method validation are examined in section 6 of this guideline.

5.1 Validation/Verification
The terms validation and verification are often linked when referring to the necessary scale of method
validation, with verification often being regarded as a process that is less extensive (and thus less
demanding) than validation.
INAB has developed a separate policy on verification of test methods which is documented in PS24.

It is important that the laboratory has undertaken an assessment of needs for the specific task and, based
on this, laid down a validation programme (a protocol), considering
1) the nature of the task including the applicants requirements in this respect
2) the method of analysis in question and the documentation of this in relation to a task of this
nature

If new requirements concerning the method of analysis have not been laid down in this situation and all
relevant performance characteristics have been fully documented in connection with previous validation,
the laboratory may choose to verify one or more (where appropriate, all) performance characteristics.
The laboratory must in this respect have documented the fact that, during performance of the method of
analysis in the given situation, the specifications laid down in connection with previous validation work
are observed. This previous validation work may be performed by the laboratory itself or originate from
publication and documentation of an accepted method.
Depending on the situation, validation work may thus consist of anything from full validation of all
performance characteristics (i.e. no verifications) to only verifications of all relevant performance
characteristics (see the summary in section 5.3).

In cases where the laboratory can justify verification only of a previously validated performance
characteristic, the requirements in PS24 shall be followed,

Any use of verification as part of validation work must be well-founded particularly if the laboratory has
accreditation for a flexible scope (see INAB Policy Statement PS11), with importance being attached to
the laboratorys competence and routines in the area in question.

5.2 General Requirements Concerning the Scope of Method Validation


Validation must be as extensive as necessary to meet the requirements in connection with the given use
or the given application.

Validation should (as a minimum) encompass:


1) a specification of the applicants need in connection with the testing in question;
PS15 Guide to Method Validation for Quantitative Analysis in Chemical Testing Laboratories
Issue 4 February 2016 Page 12 of 27
2) a definition of the performance characteristics of the method of analysis in relation to the respective
requirement;
3) an investigation of whether the needs can be met by using the method of analysis with the
performance characteristics laid down;
4) a declaration concerning the validity of the method of analysis.

In the resulting declaration, there must be justification as to why some of the performance characteristics
of the method of analysis may not have been found relevant in connection with the validation in question.
This might for instance be omission of validation of the limit of detection if the method is exclusively to be
used at high concentrations.

If there are sector-specific requirements in relation to the form and scope of a method validation, the
laboratory shall follow these requirements.

5.3 Scope of Method Validation Dependent on Type of Method of Analysis


The scope of method validation required shall be assessed for each specific analyte and shall establish the
methodology to be used.
In the accreditation standard, ISO/IEC 17025, it is stated in point 5.4.2 that:

The laboratory shall use test and/or calibration methods, including methods for sampling, which
meet the needs of the customer and which are appropriate for the tests and/or calibrations it
undertakes preferably those published as international, regional or national standards and further:
When the customer does not specify the method to be used, the laboratory shall select appropriate
methods that have been published either in international, regional or national standards, or by
reputable technical organizations, or in relevant scientific texts or journals, or as specified by the
manufacturer. Laboratory-developed methods or methods adopted by the laboratory may also be
used if they are appropriate for the intended use and if they are validated. The customer shall be
informed as to the method chosen.
The laboratory shall confirm that it can properly operate standardized methods before introducing
the tests or calibrations. If the standard method changes, the confirmation shall be repeated.

Standardised methods of analysis are published by standardisation bodies (e.g. CEN, ISO) and recognised
technical organisations (e.g. NMKL, AOAC, etc.), etc., but the documentation for the standardised method
of analysis varies. In some cases there is no validation report as a basis for the method of analysis, or the
performance characteristics are not or only partially validated. If this is the case, verification of the
laboratorys ability to use the method of analysis is not directly possible, and validation is necessary.
In this guideline, the expression accepted method of analysis is also used at several points for methods
of analysis of the above nature, with there therefore being no certainty as to how well such a method of
analysis is documented.

PS15 Guide to Method Validation for Quantitative Analysis in Chemical Testing Laboratories
Issue 4 February 2016 Page 13 of 27
5.4 Situation-Dependent Requirements Concerning the Scope of Method Validation
Some examples that can be used as the starting point for assessment in various situations are set out
below. In general, it should be mentioned that limits of quantitation only need to be included if this is
required for the specific testing field.

Examples of the scale of method validation/verification in a number of different situations:

1. The method of analysis is Completely internal validation: the laboratory must give
developed internally. This consideration to all performance characteristics (see section 6), and
includes use of a standardised perform appropriate validation of them. In the event of omission of
method of analysis on a matrix certain performance characteristics in the validation work, the
which is not described in the laboratory shall account for the omission.
method
2a The method of analysis is Full internal validation as indicated under point 1.
published in the scientific
literature, but lacks important
performance characteristics
2b The method of analysis is The laboratory shall verify its ability to use the method of analysis
published in the scientific with respect to the specified performance characteristics. The
literature and specifies important laboratory must assess the need for validation of any performance
performance characteristics characteristics.
3a The method of analysis is The laboratory shall assess the need for validation of those
standardised or otherwise well- performance characteristics, which are critical for the method of
established, but data from analysis. The laboratory must as a minimum document its ability to
external validation is unavailable use the method of analysis with quality that meets the customers
needs.
3b The method of analysis is The laboratory shall verify its ability to meet the specified
published by one of the performance characteristics for the method of analysis. Selectivity
standardisation organisations, and robustness can reasonably be expected to be investigated in
e.g., S.I., CEN, AOAC, NMKL, ISO, connection with the standardisation process.
etc. The method of analysis is
externally validated

PS15 Guide to Method Validation for Quantitative Analysis in Chemical Testing Laboratories
Issue 4 February 2016 Page 14 of 27
4. The method of analysis is well- The laboratory shall undertake and document an assessment of
established within the testing what possible effects modification of the method of analysis may
field (cf. point 2b or 3b) but is have on the individual performance characteristics.
used with modification In cases where a relevant effect is involved, the performance
characteristics in question shall be validated.
The laboratory shall verify the modified method against the
specifications for performance characteristics in the original
method.
5. Change of method of analysis The scale is in principle the same as indicated under point 2 and 3. If
the laboratorys existing method of analysis for
validation/verification proves to be free of known systematic error,
the procedure for investigating trueness may be a comparison of the
results obtained with the new and the old method of analysis. The
possibility of investigating trueness is thus improved as any natural
specimen can be included in the work.
6. Extension of the measuring range Trueness and precision shall be investigated for the new measuring
of a method of analysis already range, e.g. in the case of investigation of homogeneous inspection
commissioned and validated materials (point 5.5) or participation in inter-laboratory
investigations.
7. Change from equipment to new Measuring range, linearity, limit of detection, trueness and precision
equipment but with the same shall be compared with data from the old equipment with the aid of
measurement principle homogeneous inspection materials (verification) (see section 5.5).
8. Extension of range of matrix types Trueness and precision (performance characteristics) shall be
investigated for the new matrix type, as is deemed fit for purpose.

PS15 Guide to Method Validation for Quantitative Analysis in Chemical Testing Laboratories
Issue 4 February 2016 Page 15 of 27
5.5 Use of Homogeneous Inspection Materials
In this guideline, reference is repeatedly made to the use of homogeneous inspection materials as the
basis for laying down the trueness, precision, etc. of the method of analysis. Depending on whether
traceability is to be documented in the given phase of the validation work, the laboratory may in principle
choose between several types (levels) of inspection materials. When determining trueness, the highest
level of traceability of the inspection material shall always be sought.

The homogeneous inspection materials have traditionally been described as follows (ascending level of
traceability):
1. Inspection material at user level i.e. according to experience, stable and homogeneous test
material (not necessarily monitored systematically for stability and homogeneity) with one or more
properties that have a well-defined composition that corresponds to the specimens. Inspection
materials are suitable for the ongoing monitoring of equipment or the performance characteristics
of the method of analysis to guarantee and document compliance with prescribed quality
requirements. They are used only to determine repeatability and reproducibility and not for
calibration.
2. Commercially available inspection material, which is monitored for homogeneity and stability,
where appropriate with a specified assigned value which is not determined by a procedure
warranting the description of reference material. Such materials may be used in internal quality
control instead of the materials described above. In addition, various laboratories may compare
analysis results for these materials, i.e. comparative tests between laboratories without traceability.
Such materials are often described as reference materials, which, according to international
standards, is not correct.
3. Reference materials are stable and homogeneous materials with one or more properties that are
well-defined to be used for calibration of equipment, assessment of a method of analysis or to
guarantee the traceability of inspection materials.
4. Certified reference materials are reference materials with one or more properties whose values are
determined with the aid of technically valid measurement procedures, and which are accompanied
by or traceable to a certificate or other documentation.
These certified reference materials may be used for calibrating equipment, for validating and
monitoring methods of analysis, for monitoring the competence of laboratory personnel or when
two or more methods of analysis are to be compared.

The definitions of 3 and 4 are in accordance with the official definitions in ISO Guide 30 [12]. The
laboratory should have considered the use of homogeneous inspection materials at a relevant level in
connection with their validation work, with the need to use inspection material with a relevant matrix and
traceability requirements being catered for as far as possible.

5.6 Assessment of Scale


INAB assessors shall:
1. assess whether the laboratory has adequately covered requirements for the method of analysis
(e.g. official requirements or other generally applicable requirements in connection with routine
analyses; if new or revised methods of analysis are involved in connection with the extension of
accreditation under flexible scopes). The laboratory must be able to document close dialogue
with the customer with a view to determining the requirements in question;
2. assess whether the validation programme established (protocol) and the experimental work
performed can be considered sufficient to document compliance with the established
requirements and needs;
3. assess whether the result of the validation has shown that the established requirements and
needs have been met.

PS15 Guide to Method Validation for Quantitative Analysis in Chemical Testing Laboratories
Issue 4 February 2016 Page 16 of 27
6 REVIEW OF PERFORMANCE CHARACTERISTICS FOR A METHOD OF ANALYSIS

The parameters that characterise a method of analysis and which therefore define the suitability of the
method of analysis in a given test situation are examined in the following section.

The following performance characteristics are considered:


selectivity, measuring range, linearity, sensitivity, limit of detection, limit of quantitation,
robustness, trueness (expressed in the form of bias), precision, measurement uncertainty and
accuracy.

The performance validation exercises must encompass all stages of preparation, extraction and analysis
encountered by samples during routine analyses.

6.1 Selectivity

6.1.1 Definition
See point 4.26 (including the reference to the term specificity)

6.1.2 Content
The validation programme for selectivity/specificity should comprise investigation of possible
interference. This should as a minimum encompass:
investigation of the pure (respective) matrix or a natural specimen with the lowest possible
known content;
investigation of a solution of the relevant analytical parameters in a simple medium;
investigation of a matrix spiked with pure analytical parameters.
COMMENT: Potential interference may be assessed by spiking the sample matrix with the interfering
substances under consideration.

6.1.3 Scale
In the case of a recognised method of analysis (cf. section 5.3), in which specific interference
problems are mentioned, and where there is no justified suspicion of additional interference,
the laboratory is not required to carry out further interference testing.
In cases involving the use of a method of analysis developed by the laboratory itself, foreseeable
interference problems must be investigated and documented, (6.1.2).
If a specific usage situation might give rise to suspicions of additional interference problems,
these must be clarified by supplementary validation before the method of analysis is adopted.
The laboratorys declaration concerning the scale of validation chosen for the method of analysis
in question must incorporate considerations of possible interference e.g. chemical, spectral,
chromatographic, detection-related, extraction-related, distillation-related, etc. Interference is
assessed with reference to (where possible) specifically described test matrices, and (where
relevant) with low concentrations of the analytical parameter(s) concerned.

6.2 Measuring Range


6.2.1 Definition
See point 4.15

PS15 Guide to Method Validation for Quantitative Analysis in Chemical Testing Laboratories
Issue 4 February 2016 Page 17 of 27
6.2.2 Content
To establish the measuring range, trueness and precision are determined with specimens whose
concentrations represent the desired measuring range. The measuring range of a method of analysis
includes, where appropriate, dilution steps.

6.2.3 Scale
The laboratory must have established a validation programme (protocol) with details of how validation of
the measuring range of the method of analysis can be established. This includes documentation of
linearity, trueness and precision [see 6.3, 6.8, 6.9].

Trueness and precision must as a minimum be documented at the lowest and highest concentration level
of the measuring range. It will often be relevant for the laboratory also to have undertaken
supplementary determinations at a number of levels within the measuring range.

If the laboratory uses the method of analysis on matrices whose content of analytical parameters are
beyond the measuring range, and which the laboratory consequently wishes to dilute, additional
validation/verification must be undertaken.

6.3 Linearity/non-linearity
6.3.1 Definition
See points 4.13

6.3.2 Content
Evaluation of linearity can be performed in various steps of a method of analysis: linearity can be
investigated for calibration function and thus only cover instrumental measurement, or linearity can be
investigated for the method as a whole and thus be an investigation of trueness as a function of the
concentration of the specimen. The investigations described here are aimed at establishing the shape of
the calibration function. The principles can, however, to a large extent also be applied in assessing
linearity for the method as a whole.
The graphical shape of calibration curves (linearity/non-linearity) is established by analysing a number of
calibration standards that cover the entire desired instrumental measuring range. Calibration standards
may be reference materials containing the analytical parameters in question in the correct concentration
range in relevant matrices. In-house calibration standards at the desired concentration level are
generally used, preferably consisting of the correct matrix.

In the event of pure solutions being used as calibration standards, there is a risk of matrix effects that are
not taken into account. In this event, particular care must be taken in determining the trueness of the
method.

6.3.3 Scale
The calibration function forms the basis for all subsequent quantitative determinations, and the
laboratory must therefore have displayed particular care in validating the process. Experience shows that
ignorance of the precise shape of the calibration curve is a source of erroneous results in subsequent
routine use of the method of analysis.

It is important that the linearity investigation is performed with calibration standards of the same type as
those used in routine analysis.

There are many methods for calibrating and assessing linearity, partly dictated by the various analytical
techniques and their specific use. This document does not therefore set out generally applicable
PS15 Guide to Method Validation for Quantitative Analysis in Chemical Testing Laboratories
Issue 4 February 2016 Page 18 of 27
guidelines appropriate to all cases. All laboratories must, however, have undertaken detailed
consideration of how the validity of the calibration function adopted and its graphical course are
investigated and documented. As a minimum, such considerations should be based on the following
approach:

1. Measurement of a minimum of 6 calibration standards evenly distributed over the measuring range.
The measurements should as far as possible be performed in one series with at least double
determinations (where, appropriate triple determinations) at each concentration level and in random
order. If the method of analysis is to be used at low concentrations, the measurements should
further comprise supplementary concentration levels in this range.
2. Graphical representation of response as a function of concentration both for individual values and
for the mean.
3. Assessment of whether there is (reasonable) variance homogeneity throughout the measuring range
(based on the spread at individual points).
4. Performance of linear regression. The regression calculation should be weighted if the assessment
of variance homogeneity under point 3 indicates its relevance.
5. Plotting of residual values relative to the regression line.
6. Assessment of any outliers. Following possible rejection or replacement of one or more data points,
parts 4 6 above shall be repeated. (Note that rejection should normally only have taken place if the
laboratory has been able to identify a cause for the extreme value, and if it otherwise leaves
sufficient points for performing this procedure).
7. Assessment of linearity (based on a plot of the mean, see part 2 above) and residual plot).
8. If a linear link cannot be substantiated, the laboratory may choose one of several approaches:
Reduction of the measuring range used and subsequent documentation of linearity;
Description of a piecemeal linear link, with the difference in gradient coefficients being assessed;
Determination of a second- or third-level polynomial function and subsequent assessment of
residuals (NB: normally requires a minimum of 9 or 12 data points respectively; fourth-level
polynomials and above will normally be unrealistic).

Since, according to experience, many laboratories assess the level of linear correlation exclusively by
examining the correlation (r) or determination coefficient (r2) for the data set in question, it should be
emphasised that this does not provide adequate documentation of linearity. The laboratory shall, as a
minimum, also have undertaken a visual assessment of plots, but in addition r or r2 can certainly be used
as a supplementary parameter in the assessment. A goodness-of-fit or lack-of-fit test may be utilised to
underpin the visual assessment.

The calibration standards used by the laboratory must be of the same type, which are subsequently used
in routine analysis.

For certain types of method, the laboratory shall have assessed the calibration function of the method by
studying a graphical representation of response factors at various levels. The term response factor
should in this context be understood to mean the relationship between response for a given analytical
parameter and the actual content of the analytical parameter (e.g. the content in certified reference
material or spiked content in natural specimens).

If the laboratory has used, for example, instrument-integrated software for automatically calculating
regression, it should at the very least have been ascertained that various plots are suitable for a visual
assessment, and that algorithms used are otherwise in accordance with the above principles.

The investigations performed document the waveform of the calibration function, but not necessarily that
the calibration leads to a correct result. This will only become apparent when the trueness of the method
is investigated (see section 6.8).
PS15 Guide to Method Validation for Quantitative Analysis in Chemical Testing Laboratories
Issue 4 February 2016 Page 19 of 27
If the laboratory has, on the basis of the above, documented and described a linear relationship within the
measuring range, subsequent routine calibration of equipment may often be simpler (e.g. 2-3 point
calibration incorporating the lowest and highest calibration points which would be used in the full
calibration), with the laboratory nevertheless having taken account of any lack of variance homogeneity in
the measuring range. Investigations to validate the method shall accordingly be performed with the
calibration procedure chosen for routine use.

6.4 Sensitivity
6.4.1 Definition
See point 4.7

6.4.2 Content
Sensitivity is calculated with the aid of the gradient of the calibration curve with determination of the
linear range.
Sensitivity will be dependent on the equipment and its optimisation.

6.4.3 Scale
In so far as sensitivity affects one or more of the other performance characteristics, it shall be laid down
with the associated acceptance limit. It shall be documented to what extent sensitivity affects other
performance characteristics.

6.5 Detection Limit


6.5.1 Definition
See point 4.5
The laboratory shall follow the definition that generally applies to the testing field in question.

6.5.2 Content
The detection limit is based primarily on the standard deviation for natural specimens without or with a
very low content of the analytical parameter. Alternatively, blank specimens or natural specimens or
homogeneous inspection materials with a very low content of the analytical parameter may be used, with
the term low content being understood to mean a content that is as close as possible to the expected
detection limit. These may, however, give lower values for the detection limit.
In determining the detection limit, every specimen shall have undergone the complete method of
analysis.
If a definition of detection limit based on the desire to avoid false positives or false negatives is used, the
standard deviation used shall be based on the repeatability of the method.
If a definition of detection limit based on the desire to limit the uncertainty of the measurement is used,
the standard deviation should be based on the laboratory reproducibility of the method, i.e. a standard
deviation that caters both for variation within a day and day-to-day variation. Within certain testing
fields, there is, however, a tradition or even a requirement to use repeatability standard deviation instead
of total standard deviation. The choice of measure for variation shall therefore be determined in
accordance with what is the current standard or requirement for the testing field.
The laboratory shall substantiate the choice of definition of detection limit and the choice of measure for
variation in its calculation.

Determination of standard deviation is carried out in the same way as described in section 6.9, Precision.

If the detection limit (DL) is based on the desire to avoid false positives or false negatives,
calculation should be performed as follows:
PS15 Guide to Method Validation for Quantitative Analysis in Chemical Testing Laboratories
Issue 4 February 2016 Page 20 of 27
DL 2 t1 ( f ) sr

Where:
2 is due to the desire to have a risk of false negatives of . This () is also the risk of false positives at
a concentration equal to 1 t1 ( f ) s r . At a concentration equal to the detection limit, the risk of
false positives is at least 10 times less than .
t1- the tabulated value in the Students t-distribution with accepted risk for false negatives and f
degrees of freedom
f the number of degrees of freedom for determining sr,
sr, the repeatability standard deviation for a natural specimen without a content or a specimen with a
content of the analytical parameter as close as possible to the expected detection limit

If a blank value is deducted in the calculation of the analytical result, the following expression
should be used for calculation purposes:

Where:
sr,blank the repeatability standard deviation for a blank specimen, and
n the number of blank specimens in a series in routine analysis.

If the detection limit (DL) is based on the desire to limit the uncertainty of the measurement, the
calculation should be performed as follows:

DL = blank value + k.s

Where:
k a factor that is multiplied by the standard deviation to calculate the uncertainty. In this context, 3 is
traditionally used.
s standard deviation for a natural specimen without content, for a specimen with a very low content
of the analytical parameter or for blank specimens. The standard deviation should be the
laboratory reproducibility unless the laboratory with technical justification has argued for use of the
repeatability standard deviation.

In most cases, the blank value is deducted from the measurement result before final calculation of the
analytical result. If this is the case, it can also be deducted from the detection limit, which thereby only
becomes k.s

The laboratory should be alert to sector-specific requirements for defining the detection limit.

6.5.3 Scale
The laboratory shall have validated/verified the detection limit in all cases where a method of analysis is
used at low concentrations, with the term low concentrations in this context being understood to mean
concentrations less than 5 times the expected detection limit.
The standard deviation, which is used as the basis for calculating a detection limit, shall be determined
with at least 6 degrees of freedom. This may, for example, have occurred through the performance of 7
determinations in a single analytical series or by double determinations in each of 6 analytical series.

PS15 Guide to Method Validation for Quantitative Analysis in Chemical Testing Laboratories
Issue 4 February 2016 Page 21 of 27
6.6 Limit of Quantitation
6.6.1 Definition
See point 4.10

6.6.2 Content
The limit of quantitation should as far as possible be based on laboratory reproducibility for natural
specimens without content of the analytical parameter, blank specimens or natural specimens or
homogeneous inspection materials with a very low content of the analytical parameter, i.e. the same
specimens as used in determining the detection limit. The standard deviation is determined as described
in section 6.9 Precision.

The quantitation limit is determined traditionally as the blank value plus 10 times the repeatability
standard deviation, or in some contexts 3 times the detection limit is used, which gives largely the same
figure. Other factors can be used within certain testing fields and the laboratory shall in that case refer to
the factor that is current within the testing field in question. In many cases, the blank value is deducted
from the measurement result before final calculation of the analytical result. If this is the case, it can also
be deducted from the quantitation limit, which thereby only becomes a factor times s.

6.6.3 Scale
Many testing fields do not operate with the limit of quantitation. The limit of quantitation shall therefore
only be included in validation if this is the requirement for the testing field in question.

6.7 Robustness
6.7.1 Definition
See point 4.25

6.7.2 Content
Changes in the experimental conditions for a method of analysis, such as source and age of reagents, pH,
specimen composition and temperature, etc. may possibly affect the results of the method, as may occur
in the performance of the method of analysis by other laboratories.

It is possible to systematically investigate whether small changes in the procedure result in significant
changes in the analytical result in a robustness test.
The effects on the analytical method are investigated in this way when changes are made in either
surroundings and/or in the method procedure relative to precision and/or accuracy.

For a more detailed description of a test set-up for robustness testing, see the Statistical Manual of the
AOAC [9]

6.7.3 Scale
The conditions that may have a significant effect on the results of the method of analysis shall be
identified.
Trials shall be conducted with the aid of blank specimens, (certified) reference material, specimens of
known composition, etc., in which one or more changes in the procedure have been made and the effect
on the analytical result has been assessed. In the case of a recognised method of analysis for which
specific robustness problems are mentioned, and for which there is no justified suspicion of other
problems, the robustness test is considered adequate.

PS15 Guide to Method Validation for Quantitative Analysis in Chemical Testing Laboratories
Issue 4 February 2016 Page 22 of 27
6.8 Trueness/Bias
6.8.1 Definition
See points 4.3 and 4.9

6.8.2 Content
The description of the trueness of the method of analysis can be visualised by determining the bias of the
method by at least 2 of the following procedures:
use of a certified reference material;
comparison of the method with a primary method or a reference method with little or no bias;
use of traceable, commercially available reference materials or internal inspection materials with
an assigned value;
use of spiked specimens, based on blank or positive specimens;
participation in a performance testing programme in which setting of the reference values meets
the quality criteria specified in ISO Guide 43 [13,14].
If possible, several of the above-mentioned procedures shall be incorporated in the determination of the
bias of the method of analysis.

6.8.3 Scale
To take account of any variation between runs, bias shall be determined over several days and preferably
throughout the measuring range through the use of a suitable combination of different specimens. As the
method of analysis cannot be expected to have the same bias throughout the measuring range (e.g. with
non-linear calibration curves), several concentration levels shall be incorporated in the determination of
bias (at least one determination at a high and a low level). Otherwise, the laboratory should be able to
prove that the method of analysis employed has the same trueness throughout the measuring range.

Bias can be calculated as the relative deviation from a reference value, measured as a mean determined
with at least 6 degrees of freedom. Bias shall be tested with a t-test or by assessment of the observed
values confidence interval (calculated from the laboratory reproducibility standard deviation) relative to
the reference value and its indication of uncertainty.

If the laboratory has used spiking trials in connection with documentation of the trueness of the method
of analysis, the laboratory shall be able to prove that certain relevant considerations have been catered
for concerning the problems associated with spiking and recovery of pure analytical parameters in
relation to the presence of naturally occurring analytical parameters.

6.9 Precision
6.9.1. Definition
See points 4.18

6.9.2 Content
Repeatability:
The repeatability standard deviation, sr, is determined as the standard deviation for a suitable number of
determinations measured under repeatability conditions.

Laboratory Reproducibility:
The reproducibility standard deviation, sR, is determined as the standard deviation of a suitable number of
determinations on identical specimens analysed over several days with at least 2 different calibration
standards. In addition, the analyses should as far as possible be performed by several competent

PS15 Guide to Method Validation for Quantitative Analysis in Chemical Testing Laboratories
Issue 4 February 2016 Page 23 of 27
individuals. If the method of analysis is to be used on more than one item of equipment (e.g. several
GCs), the relevant measuring equipment must be used for determining sR.

6.9.3 Scale
The repeatability standard deviation shall be determined with at least 6 degrees of freedom. This can be
achieved, for example, by analysing 7 times in series with one specimen (f=6), 4 times in series with 2
specimens (f=6), 3 times in series with 3 specimens (f=6), etc.
In the case of sr , it is therefore the case that this can be determined as the standard deviation of several
specimens run in the series or as the pooled standard deviation of a number of multiple double
determinations run over several series.
The number of degrees of freedom achieved with various combinations of numbers of series and numbers
of specimens in each series is set out in the diagram below:

Number of Number of Degrees of freedom for Degrees of freedom for laboratory


specimens per series repeatability standard deviation reproducibility standard deviation
series
7 1 6 Not determined
4 2 6 7
3 3 6 8
2 6 6 11
N M (n-1)*m n*m 1

As repeatability and reproducibility will typically vary within the measuring range of a method of analysis,
these should be determined at several concentration levels (min. 3), with one of these levels being close
to the lower value for the measuring range.
With variations within the matrix in question, determinations at several levels will be relevant.
All the matrices for the method of analysis (or groups thereof) should form part of the laboratorys
considerations with respect to validating the precision of the method.

6.10 Measurement Uncertainty


This subject is not dealt with in detail in this guideline, but documentation of measurement uncertainty
should form an integral part of method validation. It will be natural for the laboratory to combine a
method validation study with the formulation of an uncertainty budget, not least owing to the fact that
data from the experimental work will form the basis for formulating an uncertainty budget.

6.10.1 Definition
See point 4.16

6.10.2 Content
As is apparent from the definition (4.16), measurement uncertainty is a property of a measurement result.
A method of analysis cannot therefore be assigned measurement uncertainty, but forms a source of
uncertainty that is usually characterised by its repeatability/reproducibility (ISO 5725-1 [10]). The
contribution of the method of analysis to the uncertainty of the analytical result depends on many other
things, as is apparent from the guidance in the GUM document [15] and the EURACHEM/CITAC document
[16].

Determination of measurement uncertainty comprises the following steps:


1. Identification of all sources of uncertainty;
2. Quantitation of uncertainty components;
PS15 Guide to Method Validation for Quantitative Analysis in Chemical Testing Laboratories
Issue 4 February 2016 Page 24 of 27
3. Calculation of the combined measurement uncertainty for the method of analysis.
The procedure for determining measurement uncertainty in connection with chemical methods of
analysis should as far as possible follow the guidelines in the EURACHEM/CITAC guide [16].

6.10.3 Scale
Following identification of all sources of uncertainty, it shall be assessed whether individual components
occur several times, whether the contribution from several components could, where appropriate, be
grouped and whether uncertainty components and groups of them can be quantified on the basis of data
already available including data from method validation. For example, determination of laboratory
reproducibility may cover uncertainty contributions from sub-specimen collection, volumetric equipment
and weights used in specimen preparation (but not in preparation of the calibration standards),
contributions from random noise in the detection process, contamination/loss of analytical parameters,
recovery, estimation of calibration curve and blank correction.

After the determination of measurement uncertainty has been simplified, the components left shall be
quantified. Lastly, the model according to which the contributions are combined shall be taken into
account.

When reporting measurement uncertainty, so-called expanding measurement uncertainty shall be used.
This is calculated by multiplying the combined standard uncertainty with a coverage factor, k, normally 2.

6.11 Accuracy
Accuracy is (like measurement uncertainty) a property of a measurement result, and is therefore not
included as such in the method validation itself.

6.11.1 Definition
See point 4.17

The link between the terms precision, trueness and accuracy can, among other things, be illustrated by the
following figure:

PS15 Guide to Method Validation for Quantitative Analysis in Chemical Testing Laboratories
Issue 4 February 2016 Page 25 of 27
6.11.2 Content
Accuracy can be described on the basis of the bias of the method of analysis (trueness) and precision.

6.11.3 Scale
The determination of accuracy does not require extra validation work.

7. FOLLOW-UP METHOD VALIDATION

The laboratory shall continually check that a method of analysis meets the values for the performance
characteristics documented in connection with validation.
The laboratory shall have established and implemented a procedure for internal monitoring in relation to
relevant performance characteristics for the method. In this context, it should be stressed that calibration
standards and control specimens must have a different origin.

8. LABORATORY COMPETENCE

The laboratory shall have appointed individuals (individuals responsible for the method) who can assume
responsibility for performing validation work on the basis of documented knowledge and experience,
including:
technical (theoretical and practical) knowledge of the area of analysis in question and analytical
parameters;
an understanding of the intended use of the method of analysis e.g. via contact with customers;
knowledge of the validity of methods of analysis.

A declaration concerning the validity of a method of analysis is purely the responsibility of the individual
responsible for the method, and this individual also has overall responsibility for the entire course of

PS15 Guide to Method Validation for Quantitative Analysis in Chemical Testing Laboratories
Issue 4 February 2016 Page 26 of 27
validation. The practical work on performing the validation work itself can be delegated to other
competent staff within the laboratory.

The individual responsible for the method may, where appropriate, be the same party who, for
accreditation purposes, authorises the method of analysis.

9. LABORATORY DOCUMENTATION

9.1 Procedure(s)
The laboratory shall have one or more procedure(s) defining the laboratorys approach in connection with
the performance of validation work. Such procedures shall as a minimum contain the laboratorys
guidelines for:
appointment of an individual responsible for the method validation work (individual responsible
for the method);
specification of the customers/laboratorys requirements concerning the testing in question;
guidelines for laying down the scale of validation with a view to determining performance
characteristics of the method of analysis in relation to the requirements in question;
drafting of a protocol describing the validation work;
investigation of whether the customers/laboratorys requirements can be met with the method
of analysis after the performance requirements have been laid down;
drafting a declaration concerning the validity and approval of the method of analysis.

Different validation procedures may be necessary / appropriate for different testing fields.

9.2 Reporting
The laboratory shall, for every validation job, have drawn up a validation report, which shall as a minimum
contain:
a declaration concerning the scale of validation decided upon;
a protocol for the validation work in question (or a reference thereto);
results of validation work (or reference thereto);
a final assessment of these results;
a declaration concerning the validity of the method of analysis with, where appropriate, release of
the method (approval).

The report shall be signed by the individual responsible for the method (cf. point 8) and must be kept in
the laboratory as part of the documentation for the method of analysis in question.

10. Contact

For further information about this statement please contact an INAB officer at The Irish National
Accreditation Board.

Phone: 1890 289 389


E-mail: inab@inab.ie
Website: www.inab.ie

PS15 Guide to Method Validation for Quantitative Analysis in Chemical Testing Laboratories
Issue 4 February 2016 Page 27 of 27

You might also like