SIST-TS ISO/TS 22176:2024
(Main)Cosmetics - Analytical methods - Development of a global approach for validation of quantitative analytical methods
Cosmetics - Analytical methods - Development of a global approach for validation of quantitative analytical methods
This document defines a global approach for the validation of a quantitative analytical method, based on the construction and interpretation of an accuracy profile, and specifies its characterization procedure.
This procedure is particularly applicable for internal validation in a cosmetic testing laboratory, but its scope can be extended to the interpretation of data collected for an interlaboratory study designed according to the recommendations of the ISO 5725-1. It does not apply to microbiological trials. The present approach is particularly suited to handle the wide diversity of matrices in cosmetics. This document only applies to already fully-developed and finalized methods for which selectivity/specificity have already been studied and the scope of the method to be validated has already been defined, in terms of matrix types and measurand (for example analyte) concentrations.
Cosmétiques - Méthodes analytiques - Développement d’une approche globale pour la validation des méthodes analytiques quantitatives
Le présent document définit une approche globale pour la validation d'une méthode analytique quantitative, fondée sur la construction et l'interprétation d'un profil d'exactitude, et spécifie son mode opératoire de caractérisation.
Ce mode opératoire est notamment applicable pour une validation en interne dans un laboratoire d'essais de cosmétiques, mais son domaine d'application peut être élargi à l'interprétation de données recueillies pour une étude interlaboratoires conçue conformément aux recommandations de l'ISO 5725-1. Il ne s'applique pas aux essais microbiologiques. La présente approche est notamment adaptée à la gestion de la large diversité des matrices utilisées dans les cosmétiques. Le présent document ne s'applique qu'aux méthodes déjà mises au point et totalement finalisées pour lesquelles la sélectivité/la spécificité ont déjà été étudiées et pour lesquelles le domaine d'application de la méthode à valider a déjà été défini, en termes de types de matrice et de concentrations de mesurande (par exemple, analyte).
Kozmetika - Analizne metode - Razvoj globalnega pristopa za validacijo kvantitativnih analiznih metod
Ta dokument določa globalni pristop za validacijo kvantitativnih analiznih metod na podlagi priprave in razlage profila točnosti in določa postopek njegove karakterizacije.
Ta postopek predvsem velja za notranjo validacijo kozmetičnega preskuševalnega laboratorija, njegovo področje uporabe pa je mogoče razširiti na razlago podatkov, zbranih za medlaboratorijsko študijo, pripravljeno v skladu s priporočili standarda ISO 5725-1. Ne uporablja se za mikrobiološke raziskave. Ta pristop je predvsem primeren za obravnavo različnih raznolikih matric v kozmetiki. Ta dokument se uporablja samo za povsem razvite in dokončane metode, za katere je bila že preučena selektivnost/specifičnost in je bilo že opredeljeno področje uporabe metode za validacijo v smislu vrst matric in vsebnosti merjenih količin (na primer analita).
General Information
Standards Content (Sample)
SLOVENSKI STANDARD
01-marec-2024
Kozmetika - Analizne metode - Razvoj globalnega pristopa za validacijo
kvantitativnih analiznih metod
Cosmetics - Analytical methods - Development of a global approach for validation of
quantitative analytical methods
Cosmétiques - Méthodes analytiques - Développement d’une approche globale pour la
validation des méthodes analytiques quantitatives
Ta slovenski standard je istoveten z: ISO/TS 22176:2020
ICS:
71.100.70 Kozmetika. Toaletni Cosmetics. Toiletries
pripomočki
2003-01.Slovenski inštitut za standardizacijo. Razmnoževanje celote ali delov tega standarda ni dovoljeno.
TECHNICAL ISO/TS
SPECIFICATION 22176
First edition
2020-01
Cosmetics — Analytical methods —
Development of a global approach
for validation of quantitative
analytical methods
Cosmétiques — Méthodes analytiques — Développement d’une
approche globale pour la validation des méthodes analytiques
quantitatives
Reference number
©
ISO 2020
© ISO 2020
All rights reserved. Unless otherwise specified, or required in the context of its implementation, no part of this publication may
be reproduced or utilized otherwise in any form or by any means, electronic or mechanical, including photocopying, or posting
on the internet or an intranet, without prior written permission. Permission can be requested from either ISO at the address
below or ISO’s member body in the country of the requester.
ISO copyright office
CP 401 • Ch. de Blandonnet 8
CH-1214 Vernier, Geneva
Phone: +41 22 749 01 11
Fax: +41 22 749 09 47
Email: copyright@iso.org
Website: www.iso.org
Published in Switzerland
ii © ISO 2020 – All rights reserved
Contents Page
Foreword .v
Introduction .vi
1 Scope . 1
2 Normative references . 1
3 Terms, definitions and symbols . 1
3.1 Terms and definitions . 1
3.2 Symbols . 5
4 General principles . 6
4.1 Reminder . 6
4.2 Various conditions for the estimation of precision . 6
4.3 Accuracy profile . 7
5 Procedure. 9
5.1 Definition of the measured quantity . 9
5.2 Definition of objectives . 9
5.2.1 Choice of the scope of validation . 9
5.2.2 Choice of acceptance limits . 9
5.3 Selection of validation samples .10
5.3.1 Choice of the type of matrix or types of matrices .10
5.3.2 Methods for establishing reference values.10
5.4 Characterization plan for validation .10
5.4.1 Organization .10
5.4.2 Choice of the number of series, repetitions and concentrations for the
characterization plan for validation .11
5.5 Calibration plan for the indirect methods .11
5.5.1 Organization .11
5.5.2 Choice of the number of series, repetitions and concentrations for the
calibration plan .12
5.6 Testing .13
5.7 Calculation of predicted inverse concentrations for indirect methods .14
5.7.1 General.14
5.7.2 Calculation of the calibration models .14
5.7.3 Calculation of back-calculated concentrations by inverse prediction .15
5.8 Calculation of the validation criteria by concentration level .15
5.8.1 General.15
5.8.2 Trueness criteria by series .15
5.8.3 Trueness and precision criteria by concentration .16
5.8.4 Calculation of the tolerance intervals .17
5.9 Construction of the accuracy profile .18
5.10 Interpretation of the accuracy profile for validation .19
5.10.1 General.19
5.10.2 Decision rules .20
5.10.3 Definition of the scope of validity .21
5.10.4 Choice of a calibration procedure for the routine .21
5.10.5 Influence and significance of the β proportion .21
5.10.6 Identification of outliers .22
6 Management of the outcomes during routine use .22
Annex A (normative) Calculation of repeatability, intermediate precision and
reproducibility standard deviations.23
Annex B (normative) Contents of the validation file .25
Annex C (informative) Setting-up an assay for determining the accuracy profile in the case
of NDELA in cosmetic samples .27
Annex D (informative) Influence of the value of β on the tolerance interval (R = 3 and s = 1) .37
IP
Annex E (informative) Contribution to the uncertainty calculation .38
Bibliography .39
iv © ISO 2020 – All rights reserved
Foreword
ISO (the International Organization for Standardization) is a worldwide federation of national standards
bodies (ISO member bodies). The work of preparing International Standards is normally carried out
through ISO technical committees. Each member body interested in a subject for which a technical
committee has been established has the right to be represented on that committee. International
organizations, governmental and non-governmental, in liaison with ISO, also take part in the work.
ISO collaborates closely with the International Electrotechnical Commission (IEC) on all matters of
electrotechnical standardization.
The procedures used to develop this document and those intended for its further maintenance are
described in the ISO/IEC Directives, Part 1. In particular, the different approval criteria needed for the
different types of ISO documents should be noted. This document was drafted in accordance with the
editorial rules of the ISO/IEC Directives, Part 2 (see www .iso .org/ directives).
Attention is drawn to the possibility that some of the elements of this document may be the subject of
patent rights. ISO shall not be held responsible for identifying any or all such patent rights. Details of
any patent rights identified during the development of the document will be in the Introduction and/or
on the ISO list of patent declarations received (see www .iso .org/ patents).
Any trade name used in this document is information given for the convenience of users and does not
constitute an endorsement.
For an explanation of the voluntary nature of standards, the meaning of ISO specific terms and
expressions related to conformity assessment, as well as information about ISO's adherence to
the World Trade Organization (WTO) principles in the Technical Barriers to Trade (TBT), see
www .iso .org/ iso/ foreword .html.
This document was prepared by Technical Committee ISO/TC 217, Cosmetics.
Any feedback or questions on this document should be directed to the user’s national standards body. A
complete listing of these bodies can be found at www .iso .org/ members .html.
Introduction
The purpose of this document is to propose a characterization protocol for the validation of a quantitative
analysis method in the cosmetic field and thus responds to the requirements of ISO/IEC 17025, i.e.
using the performance goals as a basis. The theoretical principles of this approach can be found in
[2]
Reference [1]. This document is based on the French Standard NF V 03-110 .
Analytical methods for analyses of cosmetics need to be validated. Validation has been long considered
as a process consisting in individually verifying several different criteria, i.e. selectivity, repeatability,
[1]
linearity, trueness, etc. The global approach, as proposed since 2003 , is based on the total error
concept and the term ‘’global” means that only a single criterion should be checked to validate a method:
the agreement between a future experimental result and the true value. This approach has already been
[1],[9] [2]
applied in the domains of pharmacy , agricultural chemistry , and is in agreement with quality
assurance guidelines such as GLP or ISO/IEC 17025. This validation process applies generally to already
developed methods and includes evaluations of the following criteria: specificity/selectivity, precision,
trueness, linearity range, LOD/LOQ, stability, ruggedness.
The large number of cosmetic products and the variety of matrices present a challenge for an analytical
laboratory requiring that standardized methods to be adapted for each type of samples. Additional
difficulties are linked to the very low concentrations to be measured, generally of the order of the mg/
kg (ppm) or µg/kg (ppb). In such context, criteria such as accuracy and uncertainty of measurement of
the analytical results are of utmost importance.
When the concentration of a substance is determined by an analytical laboratory, it is important to
evaluate the gap between the measured value and the known true value. This difference indicates
the trueness of the analysis. If cosmetic samples are analysed several times in different conditions
(laboratory, instrument, operator), the individual results will present a dispersal around the average
value which represents the precision of the measurement. As for the individual measurement, it
represents an error with the average value and an inaccuracy with regard to the reference value (i.e.
the true value).
Figure 1 — Illustration of the concepts of accuracy, precision and trueness
When a laboratory measures the concentration of a given substance in a cosmetic product sample, the
value which is obtained is thus characterized by a given accuracy which includes at the same time the
notion of trueness and precision (see Figure 1). It can also be considered as total error. The insurance
that the accuracy of a result is below acceptable limits, is thus one of the ways to make sure of the
validity of a measurement.
The accuracy profile (plot of accuracy versus concentration), such as it is developed in numerous
[3] to [9]
domains , is thus the way to know the accuracy on a result obtained with a given method applied
to a type of sample in the environment of a given laboratory.
vi © ISO 2020 – All rights reserved
To reach this accuracy profile, it is necessary to undergo a specific assay allowing to demonstrate the
validity of the analytical method, as well as the accuracy of the measurement for a given substance. In
[10]
this approach, it is necessary to determine a tolerance interval which contains a given proportion
(β) of future measured values inside (in average). If this tolerance interval is located inside a limit
of acceptability defined a priori, taking into consideration several parameters such as the type and
concentration of analyte, type of matrix, of analysis and conditions of the experiments, in this case, the
method will be considered as valid, and if it goes outside this limit of acceptability, the method will be
considered as non-valid (see Figure 2).
Key
mean value
true value
Figure 2 — Illustration of the validation principle
TECHNICAL SPECIFICATION ISO/TS 22176:2020(E)
Cosmetics — Analytical methods — Development of a global
approach for validation of quantitative analytical methods
1 Scope
This document defines a global approach for the validation of a quantitative analytical method, based on
the construction and interpretation of an accuracy profile, and specifies its characterization procedure.
This procedure is particularly applicable for internal validation in a cosmetic testing laboratory, but
its scope can be extended to the interpretation of data collected for an interlaboratory study designed
according to the recommendations of the ISO 5725-1. It does not apply to microbiological trials. The
present approach is particularly suited to handle the wide diversity of matrices in cosmetics. This
document only applies to already fully-developed and finalized methods for which selectivity/
specificity have already been studied and the scope of the method to be validated has already been
defined, in terms of matrix types and measurand (for example analyte) concentrations.
2 Normative references
The following document is referred to in the text in such a way that some or all of its content constitutes
requirements of this document. For dated references, only the edition cited applies. For undated
references, the latest edition of the referenced document (including any amendments) applies.
ISO/IEC Guide 99:2007, International vocabulary of metrology — Basic and general concepts and
associated terms (VIM)
3 Terms, definitions and symbols
3.1 Terms and definitions
For the purposes of this document, the terms and definitions given in ISO/IEC Guide 99 and the
following apply.
ISO and IEC maintain terminological databases for use in standardization at the following addresses:
— ISO Online browsing platform: available at https:// www .iso .org/ obp
— IEC Electropedia: available at http:// www .electropedia .org/
3.1.1
measurement
process of experimentally obtaining one or more quantity values that can reasonably be attributed to a
quantity
[SOURCE: ISO/IEC Guide 99:2007, 2.1, modified — Notes to entry have been excluded.]
3.1.2
measurand
quantity intended to be measured
Note 1 to entry: The term “analyte”, employed in chemistry, is a synonym of measurand, and is used more
generally.
[SOURCE: ISO/IEC Guide 99:2007, 2.3, modified — Original notes to entry have been excluded and a new
note to entry has been added.]
3.1.3
measurement trueness
trueness
closeness of agreement between the average of values obtained by replicate measurements of the same
or similar objects under specified conditions and a reference quantity value
[SOURCE: ISO/IEC Guide 99:2007, 2.14, modified — Notes to entry have been excluded.]
3.1.4
measurement precision
precision
closeness of agreement between indications or measured quantity values obtained by replicate
measurements on the same or similar objects under specified conditions
[SOURCE: ISO/IEC Guide 99:2007, 2.15, modified — Notes to entry have been excluded.]
3.1.5
repeatability condition
condition of measurement, out of a set of conditions that includes the same measurement procedure,
same operator, same measuring system, same operating conditions and same location, and replicate
measurements on the same or similar objects over a short period of time
[SOURCE: ISO/IEC Guide 99:2007, 2.20, modified — Notes to entry have been excluded.]
3.1.6
measurement repeatability
repeatability
measurement precision under a set of repeatability conditions (3.1.5) of measurement
[SOURCE: ISO/IEC Guide 99:2007, 2.21]
3.1.7
intermediate precision condition
condition of measurement, out of a set of conditions that includes the same measurement procedure,
same location, and replicate measurements on the same or similar objects over an extended period of
time, but may include other conditions involving changes
[SOURCE: ISO/IEC Guide 99:2007, 2.22, modified — Notes to entry have been excluded.]
3.1.8
intermediate measurement precision
intermediate precision
measurement precision under a set of intermediate precision conditions (3.1.7) of measurement
[SOURCE: ISO/IEC Guide 99:2007, 2.23, modified — Notes to entry have been excluded.]
3.1.9
reproducibility condition of measurement
reproducibility condition
condition of measurement, out of a set of conditions that includes different locations, operators,
measuring systems, and replicate measurements on the same or similar objects
[SOURCE: ISO/IEC Guide 99:2007, 2.24, modified — Note to entry has been excluded.]
3.1.10
measurement reproducibility
reproducibility
measurement precision under reproducibility conditions of measurement (3.1.9)
[SOURCE: ISO/IEC Guide 99:2007, 2.25, modified — Note to entry has been excluded.]
2 © ISO 2020 – All rights reserved
3.1.11
measurement accuracy
accuracy
closeness of agreement between a measured quantity value and a true quantity value of a measurand
[SOURCE: ISO/IEC Guide 99:2007, 2.13, modified — Notes to entry have been excluded.]
3.1.12
verification
provision of objective evidence that a given item fulfils specified requirements, taking into account any
measurement uncertainty
[SOURCE: ISO/IEC Guide 99:2007, 2.44, modified — Notes to entry have been excluded.]
3.1.13
validation
verification, where the specified requirements are adequate for an intended use
Note 1 to entry: The term “characterization” applies to the method, whereas the term “verification” applies to the
outcomes. Validation of the method therefore consists of checking if the results are adequate for an intended use.
[SOURCE: ISO/IEC Guide 99:2007, 2.45, modified — Example has been excluded and a Note to entry has
been added.]
3.1.14
selectivity
property of a measuring system, used with a specified measurement procedure, whereby it provides
measured quantity values for one or more measurands such that the values of each measurand are
independent of other measurands or other quantities in the measuring system
Note 1 to entry: The IUPAC considers specificity as the final stage of selectivity.
[SOURCE: ISO/IEC Guide 99:2007, 4.13, modified — Examples and original notes to entry have been
excluded. A new note to entry has been added.]
3.1.15
reference value
quantity value whose associated measurement uncertainty is generally considered small enough so
that the value may be used as a basis for comparison with quantity values of the same kind
[SOURCE: ISO/IEC Guide 99:2007, 5.18, modified — Notes to entry have been excluded.]
3.1.16
scope
all of the types of matrix (3.1.22) to which the method applies, taking into account the
range of concentrations involved in validation
3.1.17
scope of validation
all of the types of matrix (3.1.22) to which the method and range of concentrations involved in
validation applies
3.1.18
scope of validity
all of the types of matrix (3.1.22) to which the method and range of concentrations involved in validation
applies, and for which future outcomes obtained via the method will be considered valid
3.1.19
quantitative method
method of analysis which determines the quantity or weight fraction of an analyte so that it may be
expressed as a numeric value in the appropriate units
3.1.20
reference method
method of analysis recognized by experts or used as a reference by agreement between parties, which
gives, or is supposed to give the accepted reference value of the measurand
3.1.21
alternative method
method of analysis used by the laboratory instead of one or several reference methods (3.1.20)
3.1.22
matrix
set of properties of the sample and its components other than the analyte
Note 1 to entry: The matrix effect reflects the possible influence that these properties or components can have
on the instrumental response. For practical reasons, since the matrix effect can vary in the different stages
of analysis (e.g. before or after mineralisation), a type of matrix is defined as a group of materials or products
recognized by the analyst as having consistent behaviour with regard to the method of analysis used.
3.1.23
series
set of measurements carried out under a set of repeatability conditions
Note 1 to entry: For example, a series includes measurements carried out on the same day and/or by the same
operator.
3.1.24
accuracy profile
combination, in a graphic form, of one or several β-expectation tolerance intervals (3.1.25) calculated at
different concentrations, and of one or several acceptance intervals (3.1.26)
3.1.25
β-expectation tolerance interval
tolerance interval
interval which contains, on average, a defined proportion, β %, of future measurements, obtained
according to a given procedure and for a given concentration
Note 1 to entry: The limits of the interval are calculated based on trials conducted for the purpose of validation.
Note 2 to entry: A value of 80 % for β % means that, on average, one out of five results will be outside the limits of
the interval at the limit of quantitation (3.1.29). See 5.10.
3.1.26
acceptance interval
specification of the performance required for the method, expressed as an acceptable deviation around
the reference value
Note 1 to entry: The limits of the interval are set by the client or by statutory requirements, sometimes
according to the concentration. They are expressed as ±λ as absolute values and in the units of the measurand, or
(1 ± λ) × 100 as relative values.
3.1.27
linearity
establishment of a linear relationship between the deduced (or quantified) quantities
in the samples and their reference values
Note 1 to entry: Linearity of the method is different from linearity of the response function of the measuring
apparatus, which only characterizes the instrumental response during calibration and is not essential for
accurate quantitation.
4 © ISO 2020 – All rights reserved
3.1.28
validation sample
control sample
material to which the reference value may be assigned, either because it is a reference material (certified
or uncertified), or because the molecule to be assayed has been subjected to standard addition
3.1.29
limit of quantitation
the lowest and/or highest concentration of analyte that may be quantified under the experimental
conditions of the method. It corresponds to the lowest and/or highest concentration of the scope of
validity (3.1.18)
Note 1 to entry: Note 1to entry: According to the SFSTP (French society for pharmaceutical sciences and
technology), the limit of quantitation is the smallest quantity of analyte in a sample that may be assayed under
the experimental conditions described with a defined level of accuracy.
3.1.30
limit of detection
measured quantity value, obtained by a given measurement procedure, for which the probability of
falsely claiming the absence of a component in a material is b, given a probability, α, of falsely claiming
its presence
Note 1 to entry: The notation b used in this definition incurs a risk of type II error.
[SOURCE: ISO/IEC Guide 99:2007, 4.18, modified — Original notes to entry have been omitted and a
new note to entry has been added.]
3.2 Symbols
A series of i measurements (i varying from 1 to I), includes k concentrations (k varying from 1 to K),
for which j repetitions have been performed (j varying from 1 to J). The subscripts are written in the
following order: i,j,k. The random variables are written in upper case letters and their values in lower
case letters. Description of abbreviations used in formulae is given in Table 1.
Table 1 — Meaning of the different abbreviations used in formulae
Symbol Description
Reference value assigned to a calibration standard for series i (1 ≤ i ≤ I), repetition j (1 ≤ j ≤ J) and
concentration k (1 ≤ k ≤ K)
x
ijk
or
Reference value assigned to a validation sample for series i, repetition j and concentration k.
Measurement of the instrumental or experimental response observed for a calibration standard or
y
ijk
validation sample for series i, repetition j and concentration k.
Deduced value for a validation sample for series i, repetition j and concentration k, obtained either
z
ijk
by inverse prediction using a calibration model or by direct measurement.
Bias expressing the trueness error for a validation sample between the deduced value and its
b
ijk
reference value b = z − x
ijk ijk ijk
4 General principles
4.1 Reminder
The accuracy profile allows a statistical approach to validation. Formula (1) is used to describe a
measurement, z, of a measurand, Z, from a laboratory:
z = m + B + e (1)
where
m is the overall average for the homogeneous sample sent to the laboratories;
B is the bias component of the laboratory under conditions of repeatability;
e is the random error occurring in each measurement, under conditions of repeatability.
As part of an interlaboratory study, the bias component B comes from the laboratory, but it may also
come from any other source of uncertainty in an intralaboratory study, such as the day, operator,
instrument, etc.
In addition to the statistical methods for calculating the accuracy criteria, the present document also
provides details of the organization of data collection and precautions to be taken.
4.2 Various conditions for the estimation of precision
According to its definition, precision can be estimated under various conditions. In any case, precision
is quantified based on a standard deviation, be this for repeatability s , intermediate precision s or
r IP
reproducibility s . A complexity scale may be established between these different standard deviations,
R
according to the number of sources of uncertainty. Figure 3 illustrates this gradation, from conditions
of repeatability where there is no identified variation factor and/or systematic variation component for
calculating the deviation between repetitions, to the various possibilities for estimating intermediate
precision and, finally, conditions of reproducibility for which the number of sources is not known.
To simplify presentation, the notion of series refers to a set of repetitions performed under conditions
of repeatability: a series groups together all of the measurements made under the same conditions, e.g.
the same day, the same operator or a short period of time. For certain methods applying to samples that
are highly unstable over time, the chosen series effect should be the operator rather than the day; the
series will thus include repetitions performed by the same operator:
6 © ISO 2020 – All rights reserved
Figure 3 — Various estimations of the precision of a method according to the sources of
variation involved
More complicated models may be used, as in the following example, in which different laboratories,
days, operators and instruments are combined to give four series in a multi-factorial design with three
factors.
Series Laboratories Days Operators Instruments
1 1 Day 1 (−) Operator 1 (−) Instrument 1 (+)
2 1 Day 2 (+) Operator 1 (−) Instrument 2 (−)
3 1 Day 1 (−) Operator 2 (+) Instrument 2 (−)
4 1 Day 2 (+) Operator 2 (+) Instrument 1 (+)
In general, the choice of sources of variation for the measurement series should reflect as best possible
the components of variability that are likely to arise upon routine application of the method to be
validated.
NOTE For the purposes described in this document, it is essential to collect data in several series and to
control the sources of variation. Otherwise, it will not be possible to construct an accuracy profile.
4.3 Accuracy profile
From the intermediate precision or reproducibility standard deviation, calculated according to the
calculations described in Annex A, the β-expectation tolerance interval can be obtained, which includes
a proportion, β, of future outcomes.
All calculations are performed separately for each concentration k, allowing k precision standard
deviations and then k tolerance intervals to be obtained, which are brought together to construct
the accuracy profile. Figure 4 shows an example of an accuracy profile constructed using three
concentrations, 0,4 mg/L, 2,0 mg/L and 4,0 mg/L, which defines the scope or scope of validation of the
method to be validated.
The accuracy profile includes the following graphic elements:
— on the horizontal axis: the theoretical concentrations (the concentration reference values);
— on the vertical axis (simultaneously):
— the limits of the β−expectation tolerance intervals, calculated from the deduced concentrations
and expressed as percentages (recovery rate or relative accuracy);
— the acceptance intervals, defined according to the method's objective and expressed in the same
way as the tolerance intervals.
The interpretation strategy for this graph is described in detail in 5.10. Nevertheless, both the
acceptance limits and the proportion β, which is used to calculate the tolerance intervals, are strictly
dependent on the method's context of use and shall be adapted to each individual case. In Figure 4, in
the area between the broken vertical lines, the method is capable of producing a mean proportion, β,
of outcomes that lies within the acceptance limits: the method is therefore valid within this scope. The
scope of the method is the scope that was initially chosen for validation.
Key
1 mean recovery rate (%)
2 upper tolerance limit
3 upper acceptance limit
4 lower tolerance limit
5 lower acceptance limit
6 limit of quantification
7 scope of validity
8 scope of validation
X known concentration (%)
Y recovery (mg/L)
Figure 4 — Accuracy profile created using three concentrations
Each grey circle represents the ratio of the mean deduced concentration for the level, expressed as a
mean recovery rate (%) for the concentration, and quantifies the trueness. The dotted lines delimit the
acceptance interval, and the solid lines delimit the tolerance interval, calculated from the intermediate
precision standard deviations for each concentration. The vertical lines delimit the scope of validity,
within which the method is capable of producing a high and known percentage of acceptable results.
8 © ISO 2020 – All rights reserved
Generally speaking, in order to comply with the statistical models used, repetitions shall be performed
under intermediate precision or reproducibility conditions, using homogeneous validation samples.
5 Procedure
5.1 Definition of the measured quantity
Define the measured quantity according to the procedure for the method, specifying the formulae
used to calculate the final result and the method used to reach this result. In the field of cosmetics,
we are often in the case of indirect or rational methods, which require prior calibration to calculate
the concentration of the unknown samples. These methods involve a two-step approach: firstly, the
calibration curve shall be plotted, using the same physicochemical principle used for the samples;
measurements are then made using the unknown samples, and their concentrations are calculated
using the calibration model. In case of direct methods, which do not require calibration, the situation is
even simpler as there is no need of a calibration curve and the measurement on the unknown samples
provides the concentration directly.
It is essential that the quantity measured during the trial corresponds to that which will be routinely
measured. In particular, if the procedure dictates that the final outcome be expressed following a
single measurement, the result of a validation trial shall not be expressed as the mean value of several
repetitions.
5.2 Definition of objectives
5.2.1 Choice of the scope of validation
Define the scope of validation of the method in the form of a range of absolute or relative concentrations,
e.g. between 1 g and 60 g or between 1 µg/L and 500 µg/L. Practically speaking, it is through the
choices made — in terms of the range of concentrations for one or several types of matrix — when
developing the experimental design that it is possible to demonstrate the scope within which the
method is effectively valid and whether it is able to provide acceptable outcomes (5.4 and 5.5). The
scope concerned by this demonstration is called the scope of validity and may be smaller than the
previously defined scope of the method.
The choice of the method's scope may correspond to a legal requirement.
If samples are encountered during routine use of the method with concentrations that do not lie within
the scope of validity, extrapolation is not permitted. In this case, a complementary study should be
carried out to extend the scope, or a dilution may be performed if it can be shown that this has no impact.
Poor definition of the scope of the method can have significant consequences on validity, especially in
terms of specificity, interference and cross-reactivity.
5.2.2 Choice of acceptance limits
Acceptance limits are expressed as ± λ, when expressed as absolute values in the same units as the
measurand, or (1 ± λ) × 100 as relative values. Wherever possible, define the acceptance limits by
referring to a document, a professional practice or a client's requirement.
For information purposes, it could be mentioned that for pharmaceutical products the λ value is usually
[11],[12]
set at 5 % . However, in cosmetics, due to the complexity and the diversity of the matrices, it could
be proposed for ingredient metering (above 0,1 %) a reference value for λ of 10 %. For trace analysis in
[13],[14]
cosmetics, it could be between 20 % and 40 % .
Acceptance limits are generally expressed as percentages. To simplify the presentation of the examples
that illustrate this document, a fixed value has been chosen for the entire scope of the method. However,
if the scope is wide, values that vary according to the concentration may be used.
5.3 Selection of validation samples
5.3.1 Choice of the type of matrix or types of matrices
The validation samples should be materials that best represent the scope of the method.
Choose a sufficient quantity of stable and homogeneous materials to carry out all of the trials provided
in the characterization plan for validation.
5.3.2 Methods for establishing reference values
To estimate the trueness of the method, validation samples should be available whose concentration
are known as accurately as possible and with known uncertainty (ISO 11095:1996). This concentration
corresponds to the reference value assigned to the validation sample and shall be established
independently of the method to be validated. It is expressed as X. There are several possible methods
for establishing the reference value for a validation sample. These include the following:
1) use certified reference materials (CRMs), external or in-house; the traceability of these materials
decreases, respectively;
2) perform standard addition using a standard molecule of known purity. If the reference material
already contains endogenous analyte, the experimental design should include a level with no addition
so that the measured concentrations may be taken into account in the calculations (see 5.7.3);
3) prepare spiked or synthetic samples. The spiking should be carried out in the earliest stage of the
material preparation to account for the analyte-matrix interactions.
To prepare validation samples by standard addition of various concentrations, it is important to reduce
dependency between the different preparations as much as possible. Standard additions may be in
liquid form (by adding a standard solution) or in solid form (by weighing).
If successive dilutions are prepared using the same stock solution (1/2, 1/4, 1/8, etc.), this creates a
correlation between the measurements and exacerbates the consequences of errors in trueness; it
is therefore not recommended to use this approach. This note also applies to the preparation of the
standard solutions.
The rational choice of reference values is crucial in the application of this document, since it influences
the trueness of the method.
5.4 Characterization plan for validation
5.4.1 Organization
The characterization plan is used to estimate the routine performance of the method under routine
conditions of implementation.
For this, a trial consists in performing a measurement on a validation sample, with a reference value
with known uncertainty. The reference values assigned to concentrations levels k can be obtained via
the techniques described in 5.1.
The following elements are required for this plan:
— I series of measurements (1 < i < I);
— for each series, perform J repetitions (1 < j < J);
— K concentrations (1 < k < K) to cover the scope of the method.
Fill in a copy of Table 2 with the raw data. In this table, the reference value X of the validation samples
is expressed in absolute units (mg, g, etc.) or relative units (mg/kg, µg/kg, mg/L, etc.). For indirect
10 © ISO 2020 – All rights reserved
methods, the response Y is expressed in the units corresponding to the instrumental method used
(peak area, peak height, absorbance, etc.).
Table 2 — Organization of measurements for the characterization plan for validation
Measurements (instrumental
Concentrations of validation
responses) Y
Levels Series
samples (Reference values) X
1 2 … J
1 1 X y y … y
11 111 121 1J1
… … …
I …
2 1
…
I
… … …
K 1
… …
I X y
IK IJK
5.4.2 Choice of the number of series, repetitions and concentrations for the characterization
plan for validation
In the conte
...
TECHNICAL ISO/TS
SPECIFICATION 22176
First edition
2020-01
Cosmetics — Analytical methods —
Development of a global approach
for validation of quantitative
analytical methods
Cosmétiques — Méthodes analytiques — Développement d’une
approche globale pour la validation des méthodes analytiques
quantitatives
Reference number
©
ISO 2020
© ISO 2020
All rights reserved. Unless otherwise specified, or required in the context of its implementation, no part of this publication may
be reproduced or utilized otherwise in any form or by any means, electronic or mechanical, including photocopying, or posting
on the internet or an intranet, without prior written permission. Permission can be requested from either ISO at the address
below or ISO’s member body in the country of the requester.
ISO copyright office
CP 401 • Ch. de Blandonnet 8
CH-1214 Vernier, Geneva
Phone: +41 22 749 01 11
Fax: +41 22 749 09 47
Email: copyright@iso.org
Website: www.iso.org
Published in Switzerland
ii © ISO 2020 – All rights reserved
Contents Page
Foreword .v
Introduction .vi
1 Scope . 1
2 Normative references . 1
3 Terms, definitions and symbols . 1
3.1 Terms and definitions . 1
3.2 Symbols . 5
4 General principles . 6
4.1 Reminder . 6
4.2 Various conditions for the estimation of precision . 6
4.3 Accuracy profile . 7
5 Procedure. 9
5.1 Definition of the measured quantity . 9
5.2 Definition of objectives . 9
5.2.1 Choice of the scope of validation . 9
5.2.2 Choice of acceptance limits . 9
5.3 Selection of validation samples .10
5.3.1 Choice of the type of matrix or types of matrices .10
5.3.2 Methods for establishing reference values.10
5.4 Characterization plan for validation .10
5.4.1 Organization .10
5.4.2 Choice of the number of series, repetitions and concentrations for the
characterization plan for validation .11
5.5 Calibration plan for the indirect methods .11
5.5.1 Organization .11
5.5.2 Choice of the number of series, repetitions and concentrations for the
calibration plan .12
5.6 Testing .13
5.7 Calculation of predicted inverse concentrations for indirect methods .14
5.7.1 General.14
5.7.2 Calculation of the calibration models .14
5.7.3 Calculation of back-calculated concentrations by inverse prediction .15
5.8 Calculation of the validation criteria by concentration level .15
5.8.1 General.15
5.8.2 Trueness criteria by series .15
5.8.3 Trueness and precision criteria by concentration .16
5.8.4 Calculation of the tolerance intervals .17
5.9 Construction of the accuracy profile .18
5.10 Interpretation of the accuracy profile for validation .19
5.10.1 General.19
5.10.2 Decision rules .20
5.10.3 Definition of the scope of validity .21
5.10.4 Choice of a calibration procedure for the routine .21
5.10.5 Influence and significance of the β proportion .21
5.10.6 Identification of outliers .22
6 Management of the outcomes during routine use .22
Annex A (normative) Calculation of repeatability, intermediate precision and
reproducibility standard deviations.23
Annex B (normative) Contents of the validation file .25
Annex C (informative) Setting-up an assay for determining the accuracy profile in the case
of NDELA in cosmetic samples .27
Annex D (informative) Influence of the value of β on the tolerance interval (R = 3 and s = 1) .37
IP
Annex E (informative) Contribution to the uncertainty calculation .38
Bibliography .39
iv © ISO 2020 – All rights reserved
Foreword
ISO (the International Organization for Standardization) is a worldwide federation of national standards
bodies (ISO member bodies). The work of preparing International Standards is normally carried out
through ISO technical committees. Each member body interested in a subject for which a technical
committee has been established has the right to be represented on that committee. International
organizations, governmental and non-governmental, in liaison with ISO, also take part in the work.
ISO collaborates closely with the International Electrotechnical Commission (IEC) on all matters of
electrotechnical standardization.
The procedures used to develop this document and those intended for its further maintenance are
described in the ISO/IEC Directives, Part 1. In particular, the different approval criteria needed for the
different types of ISO documents should be noted. This document was drafted in accordance with the
editorial rules of the ISO/IEC Directives, Part 2 (see www .iso .org/ directives).
Attention is drawn to the possibility that some of the elements of this document may be the subject of
patent rights. ISO shall not be held responsible for identifying any or all such patent rights. Details of
any patent rights identified during the development of the document will be in the Introduction and/or
on the ISO list of patent declarations received (see www .iso .org/ patents).
Any trade name used in this document is information given for the convenience of users and does not
constitute an endorsement.
For an explanation of the voluntary nature of standards, the meaning of ISO specific terms and
expressions related to conformity assessment, as well as information about ISO's adherence to
the World Trade Organization (WTO) principles in the Technical Barriers to Trade (TBT), see
www .iso .org/ iso/ foreword .html.
This document was prepared by Technical Committee ISO/TC 217, Cosmetics.
Any feedback or questions on this document should be directed to the user’s national standards body. A
complete listing of these bodies can be found at www .iso .org/ members .html.
Introduction
The purpose of this document is to propose a characterization protocol for the validation of a quantitative
analysis method in the cosmetic field and thus responds to the requirements of ISO/IEC 17025, i.e.
using the performance goals as a basis. The theoretical principles of this approach can be found in
[2]
Reference [1]. This document is based on the French Standard NF V 03-110 .
Analytical methods for analyses of cosmetics need to be validated. Validation has been long considered
as a process consisting in individually verifying several different criteria, i.e. selectivity, repeatability,
[1]
linearity, trueness, etc. The global approach, as proposed since 2003 , is based on the total error
concept and the term ‘’global” means that only a single criterion should be checked to validate a method:
the agreement between a future experimental result and the true value. This approach has already been
[1],[9] [2]
applied in the domains of pharmacy , agricultural chemistry , and is in agreement with quality
assurance guidelines such as GLP or ISO/IEC 17025. This validation process applies generally to already
developed methods and includes evaluations of the following criteria: specificity/selectivity, precision,
trueness, linearity range, LOD/LOQ, stability, ruggedness.
The large number of cosmetic products and the variety of matrices present a challenge for an analytical
laboratory requiring that standardized methods to be adapted for each type of samples. Additional
difficulties are linked to the very low concentrations to be measured, generally of the order of the mg/
kg (ppm) or µg/kg (ppb). In such context, criteria such as accuracy and uncertainty of measurement of
the analytical results are of utmost importance.
When the concentration of a substance is determined by an analytical laboratory, it is important to
evaluate the gap between the measured value and the known true value. This difference indicates
the trueness of the analysis. If cosmetic samples are analysed several times in different conditions
(laboratory, instrument, operator), the individual results will present a dispersal around the average
value which represents the precision of the measurement. As for the individual measurement, it
represents an error with the average value and an inaccuracy with regard to the reference value (i.e.
the true value).
Figure 1 — Illustration of the concepts of accuracy, precision and trueness
When a laboratory measures the concentration of a given substance in a cosmetic product sample, the
value which is obtained is thus characterized by a given accuracy which includes at the same time the
notion of trueness and precision (see Figure 1). It can also be considered as total error. The insurance
that the accuracy of a result is below acceptable limits, is thus one of the ways to make sure of the
validity of a measurement.
The accuracy profile (plot of accuracy versus concentration), such as it is developed in numerous
[3] to [9]
domains , is thus the way to know the accuracy on a result obtained with a given method applied
to a type of sample in the environment of a given laboratory.
vi © ISO 2020 – All rights reserved
To reach this accuracy profile, it is necessary to undergo a specific assay allowing to demonstrate the
validity of the analytical method, as well as the accuracy of the measurement for a given substance. In
[10]
this approach, it is necessary to determine a tolerance interval which contains a given proportion
(β) of future measured values inside (in average). If this tolerance interval is located inside a limit
of acceptability defined a priori, taking into consideration several parameters such as the type and
concentration of analyte, type of matrix, of analysis and conditions of the experiments, in this case, the
method will be considered as valid, and if it goes outside this limit of acceptability, the method will be
considered as non-valid (see Figure 2).
Key
mean value
true value
Figure 2 — Illustration of the validation principle
TECHNICAL SPECIFICATION ISO/TS 22176:2020(E)
Cosmetics — Analytical methods — Development of a global
approach for validation of quantitative analytical methods
1 Scope
This document defines a global approach for the validation of a quantitative analytical method, based on
the construction and interpretation of an accuracy profile, and specifies its characterization procedure.
This procedure is particularly applicable for internal validation in a cosmetic testing laboratory, but
its scope can be extended to the interpretation of data collected for an interlaboratory study designed
according to the recommendations of the ISO 5725-1. It does not apply to microbiological trials. The
present approach is particularly suited to handle the wide diversity of matrices in cosmetics. This
document only applies to already fully-developed and finalized methods for which selectivity/
specificity have already been studied and the scope of the method to be validated has already been
defined, in terms of matrix types and measurand (for example analyte) concentrations.
2 Normative references
The following document is referred to in the text in such a way that some or all of its content constitutes
requirements of this document. For dated references, only the edition cited applies. For undated
references, the latest edition of the referenced document (including any amendments) applies.
ISO/IEC Guide 99:2007, International vocabulary of metrology — Basic and general concepts and
associated terms (VIM)
3 Terms, definitions and symbols
3.1 Terms and definitions
For the purposes of this document, the terms and definitions given in ISO/IEC Guide 99 and the
following apply.
ISO and IEC maintain terminological databases for use in standardization at the following addresses:
— ISO Online browsing platform: available at https:// www .iso .org/ obp
— IEC Electropedia: available at http:// www .electropedia .org/
3.1.1
measurement
process of experimentally obtaining one or more quantity values that can reasonably be attributed to a
quantity
[SOURCE: ISO/IEC Guide 99:2007, 2.1, modified — Notes to entry have been excluded.]
3.1.2
measurand
quantity intended to be measured
Note 1 to entry: The term “analyte”, employed in chemistry, is a synonym of measurand, and is used more
generally.
[SOURCE: ISO/IEC Guide 99:2007, 2.3, modified — Original notes to entry have been excluded and a new
note to entry has been added.]
3.1.3
measurement trueness
trueness
closeness of agreement between the average of values obtained by replicate measurements of the same
or similar objects under specified conditions and a reference quantity value
[SOURCE: ISO/IEC Guide 99:2007, 2.14, modified — Notes to entry have been excluded.]
3.1.4
measurement precision
precision
closeness of agreement between indications or measured quantity values obtained by replicate
measurements on the same or similar objects under specified conditions
[SOURCE: ISO/IEC Guide 99:2007, 2.15, modified — Notes to entry have been excluded.]
3.1.5
repeatability condition
condition of measurement, out of a set of conditions that includes the same measurement procedure,
same operator, same measuring system, same operating conditions and same location, and replicate
measurements on the same or similar objects over a short period of time
[SOURCE: ISO/IEC Guide 99:2007, 2.20, modified — Notes to entry have been excluded.]
3.1.6
measurement repeatability
repeatability
measurement precision under a set of repeatability conditions (3.1.5) of measurement
[SOURCE: ISO/IEC Guide 99:2007, 2.21]
3.1.7
intermediate precision condition
condition of measurement, out of a set of conditions that includes the same measurement procedure,
same location, and replicate measurements on the same or similar objects over an extended period of
time, but may include other conditions involving changes
[SOURCE: ISO/IEC Guide 99:2007, 2.22, modified — Notes to entry have been excluded.]
3.1.8
intermediate measurement precision
intermediate precision
measurement precision under a set of intermediate precision conditions (3.1.7) of measurement
[SOURCE: ISO/IEC Guide 99:2007, 2.23, modified — Notes to entry have been excluded.]
3.1.9
reproducibility condition of measurement
reproducibility condition
condition of measurement, out of a set of conditions that includes different locations, operators,
measuring systems, and replicate measurements on the same or similar objects
[SOURCE: ISO/IEC Guide 99:2007, 2.24, modified — Note to entry has been excluded.]
3.1.10
measurement reproducibility
reproducibility
measurement precision under reproducibility conditions of measurement (3.1.9)
[SOURCE: ISO/IEC Guide 99:2007, 2.25, modified — Note to entry has been excluded.]
2 © ISO 2020 – All rights reserved
3.1.11
measurement accuracy
accuracy
closeness of agreement between a measured quantity value and a true quantity value of a measurand
[SOURCE: ISO/IEC Guide 99:2007, 2.13, modified — Notes to entry have been excluded.]
3.1.12
verification
provision of objective evidence that a given item fulfils specified requirements, taking into account any
measurement uncertainty
[SOURCE: ISO/IEC Guide 99:2007, 2.44, modified — Notes to entry have been excluded.]
3.1.13
validation
verification, where the specified requirements are adequate for an intended use
Note 1 to entry: The term “characterization” applies to the method, whereas the term “verification” applies to the
outcomes. Validation of the method therefore consists of checking if the results are adequate for an intended use.
[SOURCE: ISO/IEC Guide 99:2007, 2.45, modified — Example has been excluded and a Note to entry has
been added.]
3.1.14
selectivity
property of a measuring system, used with a specified measurement procedure, whereby it provides
measured quantity values for one or more measurands such that the values of each measurand are
independent of other measurands or other quantities in the measuring system
Note 1 to entry: The IUPAC considers specificity as the final stage of selectivity.
[SOURCE: ISO/IEC Guide 99:2007, 4.13, modified — Examples and original notes to entry have been
excluded. A new note to entry has been added.]
3.1.15
reference value
quantity value whose associated measurement uncertainty is generally considered small enough so
that the value may be used as a basis for comparison with quantity values of the same kind
[SOURCE: ISO/IEC Guide 99:2007, 5.18, modified — Notes to entry have been excluded.]
3.1.16
scope
all of the types of matrix (3.1.22) to which the method applies, taking into account the
range of concentrations involved in validation
3.1.17
scope of validation
all of the types of matrix (3.1.22) to which the method and range of concentrations involved in
validation applies
3.1.18
scope of validity
all of the types of matrix (3.1.22) to which the method and range of concentrations involved in validation
applies, and for which future outcomes obtained via the method will be considered valid
3.1.19
quantitative method
method of analysis which determines the quantity or weight fraction of an analyte so that it may be
expressed as a numeric value in the appropriate units
3.1.20
reference method
method of analysis recognized by experts or used as a reference by agreement between parties, which
gives, or is supposed to give the accepted reference value of the measurand
3.1.21
alternative method
method of analysis used by the laboratory instead of one or several reference methods (3.1.20)
3.1.22
matrix
set of properties of the sample and its components other than the analyte
Note 1 to entry: The matrix effect reflects the possible influence that these properties or components can have
on the instrumental response. For practical reasons, since the matrix effect can vary in the different stages
of analysis (e.g. before or after mineralisation), a type of matrix is defined as a group of materials or products
recognized by the analyst as having consistent behaviour with regard to the method of analysis used.
3.1.23
series
set of measurements carried out under a set of repeatability conditions
Note 1 to entry: For example, a series includes measurements carried out on the same day and/or by the same
operator.
3.1.24
accuracy profile
combination, in a graphic form, of one or several β-expectation tolerance intervals (3.1.25) calculated at
different concentrations, and of one or several acceptance intervals (3.1.26)
3.1.25
β-expectation tolerance interval
tolerance interval
interval which contains, on average, a defined proportion, β %, of future measurements, obtained
according to a given procedure and for a given concentration
Note 1 to entry: The limits of the interval are calculated based on trials conducted for the purpose of validation.
Note 2 to entry: A value of 80 % for β % means that, on average, one out of five results will be outside the limits of
the interval at the limit of quantitation (3.1.29). See 5.10.
3.1.26
acceptance interval
specification of the performance required for the method, expressed as an acceptable deviation around
the reference value
Note 1 to entry: The limits of the interval are set by the client or by statutory requirements, sometimes
according to the concentration. They are expressed as ±λ as absolute values and in the units of the measurand, or
(1 ± λ) × 100 as relative values.
3.1.27
linearity
establishment of a linear relationship between the deduced (or quantified) quantities
in the samples and their reference values
Note 1 to entry: Linearity of the method is different from linearity of the response function of the measuring
apparatus, which only characterizes the instrumental response during calibration and is not essential for
accurate quantitation.
4 © ISO 2020 – All rights reserved
3.1.28
validation sample
control sample
material to which the reference value may be assigned, either because it is a reference material (certified
or uncertified), or because the molecule to be assayed has been subjected to standard addition
3.1.29
limit of quantitation
the lowest and/or highest concentration of analyte that may be quantified under the experimental
conditions of the method. It corresponds to the lowest and/or highest concentration of the scope of
validity (3.1.18)
Note 1 to entry: Note 1to entry: According to the SFSTP (French society for pharmaceutical sciences and
technology), the limit of quantitation is the smallest quantity of analyte in a sample that may be assayed under
the experimental conditions described with a defined level of accuracy.
3.1.30
limit of detection
measured quantity value, obtained by a given measurement procedure, for which the probability of
falsely claiming the absence of a component in a material is b, given a probability, α, of falsely claiming
its presence
Note 1 to entry: The notation b used in this definition incurs a risk of type II error.
[SOURCE: ISO/IEC Guide 99:2007, 4.18, modified — Original notes to entry have been omitted and a
new note to entry has been added.]
3.2 Symbols
A series of i measurements (i varying from 1 to I), includes k concentrations (k varying from 1 to K),
for which j repetitions have been performed (j varying from 1 to J). The subscripts are written in the
following order: i,j,k. The random variables are written in upper case letters and their values in lower
case letters. Description of abbreviations used in formulae is given in Table 1.
Table 1 — Meaning of the different abbreviations used in formulae
Symbol Description
Reference value assigned to a calibration standard for series i (1 ≤ i ≤ I), repetition j (1 ≤ j ≤ J) and
concentration k (1 ≤ k ≤ K)
x
ijk
or
Reference value assigned to a validation sample for series i, repetition j and concentration k.
Measurement of the instrumental or experimental response observed for a calibration standard or
y
ijk
validation sample for series i, repetition j and concentration k.
Deduced value for a validation sample for series i, repetition j and concentration k, obtained either
z
ijk
by inverse prediction using a calibration model or by direct measurement.
Bias expressing the trueness error for a validation sample between the deduced value and its
b
ijk
reference value b = z − x
ijk ijk ijk
4 General principles
4.1 Reminder
The accuracy profile allows a statistical approach to validation. Formula (1) is used to describe a
measurement, z, of a measurand, Z, from a laboratory:
z = m + B + e (1)
where
m is the overall average for the homogeneous sample sent to the laboratories;
B is the bias component of the laboratory under conditions of repeatability;
e is the random error occurring in each measurement, under conditions of repeatability.
As part of an interlaboratory study, the bias component B comes from the laboratory, but it may also
come from any other source of uncertainty in an intralaboratory study, such as the day, operator,
instrument, etc.
In addition to the statistical methods for calculating the accuracy criteria, the present document also
provides details of the organization of data collection and precautions to be taken.
4.2 Various conditions for the estimation of precision
According to its definition, precision can be estimated under various conditions. In any case, precision
is quantified based on a standard deviation, be this for repeatability s , intermediate precision s or
r IP
reproducibility s . A complexity scale may be established between these different standard deviations,
R
according to the number of sources of uncertainty. Figure 3 illustrates this gradation, from conditions
of repeatability where there is no identified variation factor and/or systematic variation component for
calculating the deviation between repetitions, to the various possibilities for estimating intermediate
precision and, finally, conditions of reproducibility for which the number of sources is not known.
To simplify presentation, the notion of series refers to a set of repetitions performed under conditions
of repeatability: a series groups together all of the measurements made under the same conditions, e.g.
the same day, the same operator or a short period of time. For certain methods applying to samples that
are highly unstable over time, the chosen series effect should be the operator rather than the day; the
series will thus include repetitions performed by the same operator:
6 © ISO 2020 – All rights reserved
Figure 3 — Various estimations of the precision of a method according to the sources of
variation involved
More complicated models may be used, as in the following example, in which different laboratories,
days, operators and instruments are combined to give four series in a multi-factorial design with three
factors.
Series Laboratories Days Operators Instruments
1 1 Day 1 (−) Operator 1 (−) Instrument 1 (+)
2 1 Day 2 (+) Operator 1 (−) Instrument 2 (−)
3 1 Day 1 (−) Operator 2 (+) Instrument 2 (−)
4 1 Day 2 (+) Operator 2 (+) Instrument 1 (+)
In general, the choice of sources of variation for the measurement series should reflect as best possible
the components of variability that are likely to arise upon routine application of the method to be
validated.
NOTE For the purposes described in this document, it is essential to collect data in several series and to
control the sources of variation. Otherwise, it will not be possible to construct an accuracy profile.
4.3 Accuracy profile
From the intermediate precision or reproducibility standard deviation, calculated according to the
calculations described in Annex A, the β-expectation tolerance interval can be obtained, which includes
a proportion, β, of future outcomes.
All calculations are performed separately for each concentration k, allowing k precision standard
deviations and then k tolerance intervals to be obtained, which are brought together to construct
the accuracy profile. Figure 4 shows an example of an accuracy profile constructed using three
concentrations, 0,4 mg/L, 2,0 mg/L and 4,0 mg/L, which defines the scope or scope of validation of the
method to be validated.
The accuracy profile includes the following graphic elements:
— on the horizontal axis: the theoretical concentrations (the concentration reference values);
— on the vertical axis (simultaneously):
— the limits of the β−expectation tolerance intervals, calculated from the deduced concentrations
and expressed as percentages (recovery rate or relative accuracy);
— the acceptance intervals, defined according to the method's objective and expressed in the same
way as the tolerance intervals.
The interpretation strategy for this graph is described in detail in 5.10. Nevertheless, both the
acceptance limits and the proportion β, which is used to calculate the tolerance intervals, are strictly
dependent on the method's context of use and shall be adapted to each individual case. In Figure 4, in
the area between the broken vertical lines, the method is capable of producing a mean proportion, β,
of outcomes that lies within the acceptance limits: the method is therefore valid within this scope. The
scope of the method is the scope that was initially chosen for validation.
Key
1 mean recovery rate (%)
2 upper tolerance limit
3 upper acceptance limit
4 lower tolerance limit
5 lower acceptance limit
6 limit of quantification
7 scope of validity
8 scope of validation
X known concentration (%)
Y recovery (mg/L)
Figure 4 — Accuracy profile created using three concentrations
Each grey circle represents the ratio of the mean deduced concentration for the level, expressed as a
mean recovery rate (%) for the concentration, and quantifies the trueness. The dotted lines delimit the
acceptance interval, and the solid lines delimit the tolerance interval, calculated from the intermediate
precision standard deviations for each concentration. The vertical lines delimit the scope of validity,
within which the method is capable of producing a high and known percentage of acceptable results.
8 © ISO 2020 – All rights reserved
Generally speaking, in order to comply with the statistical models used, repetitions shall be performed
under intermediate precision or reproducibility conditions, using homogeneous validation samples.
5 Procedure
5.1 Definition of the measured quantity
Define the measured quantity according to the procedure for the method, specifying the formulae
used to calculate the final result and the method used to reach this result. In the field of cosmetics,
we are often in the case of indirect or rational methods, which require prior calibration to calculate
the concentration of the unknown samples. These methods involve a two-step approach: firstly, the
calibration curve shall be plotted, using the same physicochemical principle used for the samples;
measurements are then made using the unknown samples, and their concentrations are calculated
using the calibration model. In case of direct methods, which do not require calibration, the situation is
even simpler as there is no need of a calibration curve and the measurement on the unknown samples
provides the concentration directly.
It is essential that the quantity measured during the trial corresponds to that which will be routinely
measured. In particular, if the procedure dictates that the final outcome be expressed following a
single measurement, the result of a validation trial shall not be expressed as the mean value of several
repetitions.
5.2 Definition of objectives
5.2.1 Choice of the scope of validation
Define the scope of validation of the method in the form of a range of absolute or relative concentrations,
e.g. between 1 g and 60 g or between 1 µg/L and 500 µg/L. Practically speaking, it is through the
choices made — in terms of the range of concentrations for one or several types of matrix — when
developing the experimental design that it is possible to demonstrate the scope within which the
method is effectively valid and whether it is able to provide acceptable outcomes (5.4 and 5.5). The
scope concerned by this demonstration is called the scope of validity and may be smaller than the
previously defined scope of the method.
The choice of the method's scope may correspond to a legal requirement.
If samples are encountered during routine use of the method with concentrations that do not lie within
the scope of validity, extrapolation is not permitted. In this case, a complementary study should be
carried out to extend the scope, or a dilution may be performed if it can be shown that this has no impact.
Poor definition of the scope of the method can have significant consequences on validity, especially in
terms of specificity, interference and cross-reactivity.
5.2.2 Choice of acceptance limits
Acceptance limits are expressed as ± λ, when expressed as absolute values in the same units as the
measurand, or (1 ± λ) × 100 as relative values. Wherever possible, define the acceptance limits by
referring to a document, a professional practice or a client's requirement.
For information purposes, it could be mentioned that for pharmaceutical products the λ value is usually
[11],[12]
set at 5 % . However, in cosmetics, due to the complexity and the diversity of the matrices, it could
be proposed for ingredient metering (above 0,1 %) a reference value for λ of 10 %. For trace analysis in
[13],[14]
cosmetics, it could be between 20 % and 40 % .
Acceptance limits are generally expressed as percentages. To simplify the presentation of the examples
that illustrate this document, a fixed value has been chosen for the entire scope of the method. However,
if the scope is wide, values that vary according to the concentration may be used.
5.3 Selection of validation samples
5.3.1 Choice of the type of matrix or types of matrices
The validation samples should be materials that best represent the scope of the method.
Choose a sufficient quantity of stable and homogeneous materials to carry out all of the trials provided
in the characterization plan for validation.
5.3.2 Methods for establishing reference values
To estimate the trueness of the method, validation samples should be available whose concentration
are known as accurately as possible and with known uncertainty (ISO 11095:1996). This concentration
corresponds to the reference value assigned to the validation sample and shall be established
independently of the method to be validated. It is expressed as X. There are several possible methods
for establishing the reference value for a validation sample. These include the following:
1) use certified reference materials (CRMs), external or in-house; the traceability of these materials
decreases, respectively;
2) perform standard addition using a standard molecule of known purity. If the reference material
already contains endogenous analyte, the experimental design should include a level with no addition
so that the measured concentrations may be taken into account in the calculations (see 5.7.3);
3) prepare spiked or synthetic samples. The spiking should be carried out in the earliest stage of the
material preparation to account for the analyte-matrix interactions.
To prepare validation samples by standard addition of various concentrations, it is important to reduce
dependency between the different preparations as much as possible. Standard additions may be in
liquid form (by adding a standard solution) or in solid form (by weighing).
If successive dilutions are prepared using the same stock solution (1/2, 1/4, 1/8, etc.), this creates a
correlation between the measurements and exacerbates the consequences of errors in trueness; it
is therefore not recommended to use this approach. This note also applies to the preparation of the
standard solutions.
The rational choice of reference values is crucial in the application of this document, since it influences
the trueness of the method.
5.4 Characterization plan for validation
5.4.1 Organization
The characterization plan is used to estimate the routine performance of the method under routine
conditions of implementation.
For this, a trial consists in performing a measurement on a validation sample, with a reference value
with known uncertainty. The reference values assigned to concentrations levels k can be obtained via
the techniques described in 5.1.
The following elements are required for this plan:
— I series of measurements (1 < i < I);
— for each series, perform J repetitions (1 < j < J);
— K concentrations (1 < k < K) to cover the scope of the method.
Fill in a copy of Table 2 with the raw data. In this table, the reference value X of the validation samples
is expressed in absolute units (mg, g, etc.) or relative units (mg/kg, µg/kg, mg/L, etc.). For indirect
10 © ISO 2020 – All rights reserved
methods, the response Y is expressed in the units corresponding to the instrumental method used
(peak area, peak height, absorbance, etc.).
Table 2 — Organization of measurements for the characterization plan for validation
Measurements (instrumental
Concentrations of validation
responses) Y
Levels Series
samples (Reference values) X
1 2 … J
1 1 X y y … y
11 111 121 1J1
… … …
I …
2 1
…
I
… … …
K 1
… …
I X y
IK IJK
5.4.2 Choice of the number of series, repetitions and concentrations for the characterization
plan for validation
In the context of this document, the minimum requirements for the number of series are:
— a number of series, I, equal to 5, possibly reduced to 4 or 3 if this can be justified. A series may be
represented by a particular day, but also by a combination of various sources of uncertainty, such as
multiple devices, multiple operators and multiple days;
— a constant number of repetitions per series and per concentration, J, equal to or greater than 2;
— a number of concentrations, K, equal to or greater than 3. It is essential that K ≥ 3 as this allows the
linearity between the concentrations of the reference values and the deduced concentrations to be
checked: three concentrations are required for this check. However, when it is necessary to validate
the method close to its limit of quantitation LOQ, it is recommended that a K equal to or greater than
4 be chosen.
If I = 5, J = 2 and K = 3, the experimental design will involve 30 tests. Apart from a minimum limit of
I = 3, a high degree of flexibility is allowed in the choice of the number of tests.
As a reminder, the higher the number of tests, the better the estimates of the validation criteria, i.e.
the interval will be narrower and validation easier. However, to improve the results, it is generally
preferable to increase the number of series rather than the number of repetitions or concentrations.
5.5
...
SPÉCIFICATION ISO/TS
TECHNIQUE 22176
Première édition
2020-01
Cosmétiques — Méthodes analytiques
— Développement d’une approche
globale pour la validation des
méthodes analytiques quantitatives
Cosmetics — Analytical methods — Development of a global
approach for validation of quantitative analytical methods
Numéro de référence
©
ISO 2020
DOCUMENT PROTÉGÉ PAR COPYRIGHT
© ISO 2020
Tous droits réservés. Sauf prescription différente ou nécessité dans le contexte de sa mise en œuvre, aucune partie de cette
publication ne peut être reproduite ni utilisée sous quelque forme que ce soit et par aucun procédé, électronique ou mécanique,
y compris la photocopie, ou la diffusion sur l’internet ou sur un intranet, sans autorisation écrite préalable. Une autorisation peut
être demandée à l’ISO à l’adresse ci-après ou au comité membre de l’ISO dans le pays du demandeur.
ISO copyright office
Case postale 401 • Ch. de Blandonnet 8
CH-1214 Vernier, Genève
Tél.: +41 22 749 01 11
Fax: +41 22 749 09 47
E-mail: copyright@iso.org
Web: www.iso.org
Publié en Suisse
ii © ISO 2020 – Tous droits réservés
Sommaire Page
Avant-propos .v
Introduction .vi
1 Domaine d’application . 1
2 Références normatives . 1
3 Termes, définitions et symboles . 1
3.1 Termes et définitions . 1
3.2 Symboles . 5
4 Principes généraux . 6
4.1 Rappel . 6
4.2 Conditions diverses pour l’estimation de la fidélité . 6
4.3 Profil d’exactitude . 7
5 Mode opératoire. 9
5.1 Définition de la grandeur mesurée . 9
5.2 Définition des objectifs . 9
5.2.1 Choix du domaine de validation . 9
5.2.2 Choix des limites d’acceptation . 9
5.3 Sélection des échantillons de validation .10
5.3.1 Choix du type de matrice ou des types de matrices .10
5.3.2 Méthodes d’établissement des valeurs de référence .10
5.4 Plan de caractérisation de la validation .10
5.4.1 Organisation .10
5.4.2 Choix du nombre de séries, de répétitions et de concentrations pour le
plan de caractérisation de la validation .11
5.5 Plan d’étalonnage pour les méthodes indirectes.12
5.5.1 Organisation .12
5.5.2 Choix du nombre de séries, de répétitions et de concentrations pour le
plan d’étalonnage .12
5.6 Essais .13
5.7 Calcul de concentrations inverses prédites pour les méthodes indirectes .14
5.7.1 Généralités .14
5.7.2 Calcul des modèles d’étalonnage .14
5.7.3 Rétro-calcul des concentrations retrouvées par prédiction inverse .16
5.8 Calcul des critères de validation par niveau de concentration .16
5.8.1 Généralités .16
5.8.2 Critères de justesse par série .16
5.8.3 Critères de justesse et de fidélité par concentration .17
5.8.4 Calcul des intervalles de tolérance .18
5.9 Construction du profil d’exactitude .19
5.10 Interprétation du profil d’exactitude pour la validation .21
5.10.1 Généralités .21
5.10.2 Règles de décision .21
5.10.3 Définition du domaine de validité . .23
5.10.4 Choix d’un mode opératoire d’étalonnage pour les analyses courantes .23
5.10.5 Influence et importance du pourcentage β .23
5.10.6 Identification des valeurs aberrantes .24
6 Maîtrise des résultats en routine .24
Annexe A (normative) Calcul des écarts-types de répétabilité, de fidélité intermédiaire et
de reproductibilité .25
Annexe B (normative) Contenu du fichier de validation .28
Annexe C (informative) Paramétrage d’un dosage visant à déterminer le profil d’exactitude
dans le cas de la NDELA dans des échantillons cosmétiques .30
Annexe D (informative) Influence de la valeur de β sur l’intervalle de tolérance (R = 3 et s = 1) .40
IP
Annexe E (informative) Contribution au calcul de l’incertitude .41
Bibliographie .42
iv © ISO 2020 – Tous droits réservés
Avant-propos
L’ISO (Organisation internationale de normalisation) est une fédération mondiale d’organismes
nationaux de normalisation (comités membres de l’ISO). L’élaboration des Normes internationales est
en général confiée aux comités techniques de l’ISO. Chaque comité membre intéressé par une étude
a le droit de faire partie du comité technique créé à cet effet. Les organisations internationales,
gouvernementales et non gouvernementales, en liaison avec l’ISO participent également aux travaux.
L’ISO collabore étroitement avec la Commission électrotechnique internationale (IEC) en ce qui
concerne la normalisation électrotechnique.
Les procédures utilisées pour élaborer le présent document et celles destinées à sa mise à jour sont
décrites dans les Directives ISO/IEC, Partie 1. Il convient, en particulier de prendre note des différents
critères d’approbation requis pour les différents types de documents ISO. Le présent document a été
rédigé conformément aux règles de rédaction données dans les Directives ISO/IEC, Partie 2 (voir www
.iso .org/ directives).
L’attention est attirée sur le fait que certains des éléments du présent document peuvent faire l’objet de
droits de propriété intellectuelle ou de droits analogues. L’ISO ne saurait être tenue pour responsable
de ne pas avoir identifié de tels droits de propriété et averti de leur existence. Les détails concernant
les références aux droits de propriété intellectuelle ou autres droits analogues identifiés lors de
l’élaboration du document sont indiqués dans l’Introduction et/ou dans la liste des déclarations de
brevets reçues par l’ISO (voir www .iso .org/ brevets).
Les appellations commerciales éventuellement mentionnées dans le présent document sont données
pour information, par souci de commodité, à l’intention des utilisateurs et ne sauraient constituer un
engagement.
Pour une explication de la nature volontaire des normes, la signification des termes et expressions
spécifiques de l’ISO liés à l’évaluation de la conformité, ou pour toute information au sujet de l’adhésion
de l’ISO aux principes de l’Organisation mondiale du commerce (OMC) concernant les obstacles
techniques au commerce (OTC), voir le lien suivant: www .iso .org/ iso/ fr/ avant -propos.
Le présent document a été élaboré par le comité technique ISO/TC 217, Cosmétiques.
Il convient que l’utilisateur adresse tout retour d’information ou toute question concernant le présent
document à l’organisme national de normalisation de son pays. Une liste exhaustive desdits organismes
se trouve à l’adresse www .iso .org/ fr/ members .html.
Introduction
Le présent document vise à proposer un protocole de caractérisation pour la validation d’une méthode
analytique quantitative dans le domaine de la cosmétique et à satisfaire ainsi aux exigences de
l’ISO/IEC 17025, c’est-à-dire en se fondant sur les objectifs de performance. Les principes théoriques
de cette approche peuvent être consultés dans la Référence [1]. Le présent document est fondé sur la
[2]
Norme Française NF V 03-110 .
Les méthodes analytiques destinées à l’analyse de cosmétiques nécessitent d’être validées. La validation
a été pendant longtemps envisagée comme un processus consistant à vérifier individuellement plusieurs
critères différents, à savoir la sélectivité, la répétabilité, la linéarité, la justesse, etc. L’approche globale,
[1]
proposée en 2003 , est fondée sur le concept d’erreur totale et le terme «global» signifie qu’il convient
de ne vérifier qu’un seul critère pour valider une méthode: l’accord entre un futur résultat expérimental
[1],[9]
et la valeur vraie. Cette approche a déjà été appliquée dans les domaines de la pharmacie , de
[2]
l’agrochimie , et est conforme aux lignes directrices d’assurance de la qualité telles que les bonnes
pratiques de laboratoire ou l’ISO/IEC 17025. Ce processus de validation s’applique généralement aux
méthodes déjà mises au point et intègre les évaluations des critères suivants: spécificité/sélectivité,
fidélité, justesse, domaine de linéarité, limite de détection/limite de quantification (LD/LQ), stabilité,
robustesse.
Le grand nombre de produits cosmétiques et la diversité des matrices représentent un défi pour un
laboratoire d’analyses car cela exige d’adapter les méthodes normalisées à chaque type d’échantillons.
Des difficultés supplémentaires sont liées aux très faibles concentrations à quantifier, en général de
l’ordre du mg/kg (ppm, partie par million) ou du µg/kg (ppb, partie par milliard). Dans ce contexte, des
critères tels que l’exactitude et l’incertitude de mesure des résultats analytiques sont de la plus haute
importance.
Lorsque la concentration d’une substance est déterminée par un laboratoire d’analyses, il est important
d’évaluer l’écart entre la valeur mesurée et la valeur vraie connue. Cette différence représente la justesse
de l’analyse. Si des échantillons cosmétiques sont analysés plusieurs fois sous des conditions différentes
(laboratoire, appareillage, opérateur), les résultats individuels présenteront une dispersion autour de
la valeur moyenne, laquelle représente la fidélité de la mesure. Tout comme pour la mesure individuelle,
elle représente une erreur par rapport à la valeur moyenne et une inexactitude par rapport à la valeur
de référence (c’est-à-dire la valeur vraie).
Figure 1 — Représentation des concepts d’exactitude, de fidélité et de justesse
Lorsqu’un laboratoire mesure la concentration d’une substance donnée dans un échantillon de produit
cosmétique, la valeur qui est obtenue est donc caractérisée par une exactitude donnée laquelle englobe
en même temps la notion de justesse et de fidélité (voir Figure 1). Elle peut également être envisagée
comme l’erreur totale. La garantie que l’exactitude d’un résultat est en dessous des limites d’acceptation
est donc l’une des manières de démontrer la validité d’une mesure.
vi © ISO 2020 – Tous droits réservés
Le profil d’exactitude (tracé de l’exactitude en fonction de la concentration), tel qu’il a été développé
[3] à [9]
dans de nombreux domaines , est donc la façon de connaître l’exactitude d’un résultat obtenu avec
une méthode donnée appliquée à un type d’échantillon dans l’environnement d’un laboratoire donné.
Pour atteindre ce profil d’exactitude, il est nécessaire de réaliser un dosage spécifique permettant de
démontrer la validité de la méthode analytique, ainsi que l’exactitude de la mesure pour une substance
[10]
donnée. Dans le cadre de cette approche, il est nécessaire de déterminer un intervalle de tolérance
dans lequel s’inscrit une proportion donnée (β) de futures valeurs mesurées (en moyenne). Si cet
intervalle de tolérance est situé à l’intérieur d’une limite d’acceptation définie a priori, en prenant en
compte plusieurs paramètres tels que le type et la concentration d’analyte, le type de matrice, le type
d’analyse et les conditions des expériences, dans ce cas, la méthode sera jugée valide, et s’il ne respecte
pas cette limite d’acceptation, la méthode sera jugée non valide (voir Figure 2).
Légende
valeur moyenne
valeur vraie
Figure 2 — Représentation du principe de validation
SPÉCIFICATION TECHNIQUE ISO/TS 22176:2020(F)
Cosmétiques — Méthodes analytiques — Développement
d’une approche globale pour la validation des méthodes
analytiques quantitatives
1 Domaine d’application
Le présent document définit une approche globale pour la validation d’une méthode analytique
quantitative, fondée sur la construction et l’interprétation d’un profil d’exactitude, et spécifie son mode
opératoire de caractérisation.
Ce mode opératoire est notamment applicable pour une validation en interne dans un laboratoire d’essais
de cosmétiques, mais son domaine d’application peut être élargi à l’interprétation de données recueillies
pour une étude interlaboratoires conçue conformément aux recommandations de l’ISO 5725-1. Il ne
s’applique pas aux essais microbiologiques. La présente approche est notamment adaptée à la gestion
de la large diversité des matrices utilisées dans les cosmétiques. Le présent document ne s’applique
qu’aux méthodes déjà mises au point et totalement finalisées pour lesquelles la sélectivité/la spécificité
ont déjà été étudiées et pour lesquelles le domaine d’application de la méthode à valider a déjà été défini,
en termes de types de matrice et de concentrations de mesurande (par exemple, analyte).
2 Références normatives
Les documents suivants sont cités dans le texte de sorte qu’ils constituent, pour tout ou partie de leur
contenu, des exigences du présent document. Pour les références datées, seule l’édition citée s’applique.
Pour les références non datées, la dernière édition du document de référence s’applique (y compris les
éventuels amendements).
ISO/IEC Guide 99:2007, Vocabulaire international de métrologie — Concepts fondamentaux et généraux et
termes associés (VIM)
3 Termes, définitions et symboles
3.1 Termes et définitions
Pour les besoins du présent document, les termes et définitions de l’ISO/IEC Guide 99 et les suivants,
s’appliquent.
L’ISO et l’IEC tiennent à jour des bases de données terminologiques destinées à être utilisées en
normalisation, consultables aux adresses suivantes:
— ISO Online browsing platform: disponible à l’adresse https:// www .iso .org/ obp;
— IEC Electropedia: disponible à l’adresse http:// www .electropedia .org/ .
3.1.1
mesurage
processus consistant à obtenir expérimentalement une ou plusieurs valeurs que l’on peut
raisonnablement attribuer à une grandeur
[SOURCE: ISO/IEC Guide 99:2007, 2.1, modifiée — Les notes à l’article ont été exclues.]
3.1.2
mesurande
grandeur que l’on veut mesurer
Note 1 à l'article: Le terme «analyte», employé en chimie, est un synonyme de mesurande, et est plus couramment
utilisé.
[SOURCE: ISO/IEC Guide 99:2007, 2.3, modifiée — Les notes à l’article d’origine ont été exclues et une
nouvelle note à l’article a été ajoutée.]
3.1.3
justesse de mesure
justesse
étroitesse de l’accord entre la moyenne de valeurs obtenues par des mesurages répétés du même objet
ou d’objets similaires dans des conditions spécifiées et une valeur de référence
[SOURCE: ISO/IEC Guide 99:2007, 2.14, modifiée — Les notes à l’article ont été exclues.]
3.1.4
fidélité de mesure
fidélité
étroitesse de l’accord entre les indications ou les valeurs mesurées obtenues par des mesurages répétés
du même objet ou d’objets similaires dans des conditions spécifiées
[SOURCE: ISO/IEC Guide 99:2007, 2.15, modifiée — Les notes à l’article ont été exclues.]
3.1.5
condition de répétabilité
condition de mesurage dans un ensemble de conditions qui comprennent la même procédure de mesure,
le même opérateur, le même système de mesure, les mêmes conditions de fonctionnement et le même
lieu, ainsi que des mesurages répétés sur le même objet ou des objets similaires pendant une courte
période de temps
[SOURCE: ISO/IEC Guide 99:2007, 2.20, modifiée — Les notes à l’article ont été exclues.]
3.1.6
répétabilité de mesure
répétabilité
fidélité de mesure selon un ensemble de conditions de répétabilité (3.1.5)
[SOURCE: ISO/IEC Guide 99:2007, 2.21]
3.1.7
condition de fidélité intermédiaire
condition de mesurage dans un ensemble de conditions qui comprennent la même procédure de mesure,
le même lieu et des mesurages répétés sur le même objet ou des objets similaires pendant une période
de temps étendue, mais peuvent comprendre d’autres conditions que l’on fait varier
[SOURCE: ISO/IEC Guide 99:2007, 2.22, modifiée — Les notes à l’article ont été exclues.]
3.1.8
fidélité intermédiaire de mesure
fidélité intermédiaire
fidélité de mesure selon un ensemble de conditions de fidélité intermédiaire (3.1.7)
[SOURCE: ISO/IEC Guide 99:2007, 2.23, modifiée — Les notes à l’article ont été exclues.]
2 © ISO 2020 – Tous droits réservés
3.1.9
condition de reproductibilité
condition de mesurage dans un ensemble de conditions qui comprennent des lieux, des opérateurs et
des systèmes de mesure différents, ainsi que des mesurages répétés sur le même objet ou des objets
similaires
[SOURCE: ISO/IEC Guide 99:2007, 2.24, modifiée — La note à l’article a été exclue.]
3.1.10
reproductibilité de mesure
reproductibilité
fidélité de mesure selon un ensemble de conditions de reproductibilité (3.1.9)
[SOURCE: ISO/IEC Guide 99:2007, 2.25, modifiée — La note à l’article a été exclue.]
3.1.11
exactitude de mesure
exactitude
étroitesse de l’accord entre une valeur mesurée et une valeur vraie d’un mesurande
[SOURCE: ISO/IEC Guide 99:2007, 2.13, modifiée — Les notes à l’article ont été exclues.]
3.1.12
vérification
fourniture de preuves tangibles qu’une entité donnée satisfait à des exigences spécifiées, en prenant en
compte l’exactitude de mesure
[SOURCE: ISO/IEC Guide 99:2007, 2.44, modifiée — Les notes à l’article ont été exclues.]
3.1.13
validation
vérification, où les exigences spécifiées sont adéquates pour un usage déterminé
Note 1 à l'article: Le terme «caractérisation» s’applique à la méthode, tandis que le terme «vérification» s’applique
aux résultats. La validation de la méthode consiste donc à vérifier si les résultats sont adéquats pour un usage
déterminé.
[SOURCE: ISO/IEC Guide 99:2007, 2.45, modifiée — L’exemple a été exclu et une note à l’article a été
ajoutée.]
3.1.14
sélectivité
propriété d’un système de mesure, utilisant une procédure de mesure spécifiée, selon laquelle le
système fournit des valeurs mesurées pour un ou plusieurs mesurandes, telles que les valeurs de
chaque mesurande sont indépendantes des autres mesurandes ou d’autres grandeurs dans le système
de mesure
Note 1 à l'article: L’IUPAC envisage la spécificité comme l’étape finale de sélectivité.
[SOURCE: ISO/IEC Guide 99:2007, 4.13, modifiée — Les exemples et notes à l’article d’origine ont été
exclus. Une nouvelle note à l’article a été ajoutée.]
3.1.15
valeur de référence
valeur d’une grandeur dont l’incertitude de mesure associée est généralement jugée comme
suffisamment faible pour que la valeur puisse servir de base de comparaison pour les valeurs de
grandeurs de même nature
[SOURCE: ISO/IEC Guide 99:2007, 5.18, modifiée — Les notes à l’article ont été exclues.]
3.1.16
domaine d’application
ensemble des types de matrice (3.1.22) auxquels la méthode s’applique, en prenant en
compte la gamme de concentrations étudiée dans le cadre d’une validation
3.1.17
domaine de validation
ensemble des types de matrice (3.1.22) auxquels la méthode et la gamme de concentrations étudiées
dans le cadre d’une validation s’appliquent
3.1.18
domaine de validité
ensemble des types de matrice (3.1.22) auxquels la méthode et la gamme de concentrations étudiée
dans le cadre d’une validation s’appliquent, et pour lesquels de futurs résultats obtenus par la méthode
seront jugés valides
3.1.19
méthode quantitative
méthode d’analyse qui détermine la quantité ou la fraction massique d’un analyte en délivrant une
valeur numérique exprimée selon les unités adéquates
3.1.20
méthode de référence
méthode d’analyse reconnue par des experts ou utilisée comme référence par accord entre les parties,
qui donne, ou est présumée donner, la valeur de référence acceptée du mesurande
3.1.21
méthode alternative
méthode d’analyse utilisée par le laboratoire à la place d’une ou de plusieurs méthodes de référence
(3.1.20)
3.1.22
matrice
ensemble de propriétés de l’échantillon et de ses constituants autres que l’analyte
Note 1 à l'article: L’effet matrice reflète l’éventuelle influence que ces propriétés ou constituants peuvent avoir
sur la réponse instrumentale. Étant donné que l’effet matrice peut fluctuer selon les différentes étapes d’analyse
(par exemple, avant ou après digestion), un type de matrice est défini pour des raisons pratiques comme un
groupe de matériaux ou produits connus de l’analyste comme présentant un comportement constant par rapport
à la méthode d’analyse utilisée.
3.1.23
série
ensemble de mesurages effectués selon un ensemble de conditions de répétabilité
Note 1 à l'article: Des mesurages effectués le même jour et/ou par le même opérateur sont un exemple de série.
3.1.24
profil d’exactitude
combinaison, présentée sous forme graphique, d’un ou de plusieurs intervalles de tolérance au niveau β
(3.1.25) calculés à différentes concentrations, et d’un ou de plusieurs intervalles d’acceptation (3.1.26)
3.1.25
intervalle de tolérance au niveau β
intervalle de tolérance
intervalle dans lequel s’inscrit, en moyenne, un pourcentage défini, β %, de futures mesures, obtenues
conformément à un mode opératoire donné et pour une concentration donnée
Note 1 à l'article: Les limites de l’intervalle sont calculées sur la base d’essais menés aux fins d’une validation.
Note 2 à l'article: Un pourcentage β % de 80 % signifie qu’en moyenne, un résultat sur cinq ne s’inscrira pas dans
les limites de l’intervalle à la limite de quantification (3.1.29). Voir 5.10.
4 © ISO 2020 – Tous droits réservés
3.1.26
intervalle d’acceptation
spécification de la performance exigée pour la méthode, exprimée sous la forme d’un écart acceptable
autour de la valeur de référence
Note 1 à l'article: Les limites de l’intervalle sont fixées par le client ou par des exigences réglementaires, parfois
en fonction de la concentration. Elles sont exprimées en valeur absolue dans l’unité du mesurande sous la forme
±λ ou sont exprimées en valeur relative sous la forme (1 ± λ) × 100.
3.1.27
linéarité
établissement d’une relation linéaire entre les quantités déduites (ou quantifiées) des
échantillons et leurs valeurs de référence
Note 1 à l'article: La linéarité de la méthode est différente de la linéarité de la fonction de réponse de l’appareillage
de mesure, cette dernière ne caractérise que la réponse instrumentale pendant l’étalonnage et n’est pas
essentielle pour une quantification exacte.
3.1.28
échantillon de validation
échantillon témoin
matériau auquel la valeur de référence peut être attribuée, soit parce qu’il s’agit d’un matériau de
référence (certifié ou non certifié), soit parce que la molécule à quantifier a fait l’objet d’un ajout dosé
3.1.29
limite de quantification
plus faible et/ou plus forte concentrations d’analyte qui peuvent être quantifiées dans les conditions
expérimentales de la méthode, correspondant à la plus faible et/ou à la plus forte concentrations du
domaine de validité (3.1.18)
Note 1 à l'article: Conformément à la SFSTP (Société Française des Sciences et Techniques Pharmaceutiques), la
limite de quantification est la plus faible quantité d’analyte qui peut être quantifiée dans un échantillon dans les
conditions expérimentales décrites avec un niveau d’exactitude défini.
3.1.30
limite de détection
valeur mesurée, obtenue par une procédure de mesure donnée, pour laquelle la probabilité de déclarer
faussement l’absence d’un constituant dans un matériau est b, étant donnée la probabilité, α, de déclarer
faussement sa présence
Note 1 à l'article: La notation b utilisée dans la présente définition implique un risque d’erreur de type II.
[SOURCE: ISO/IEC Guide 99:2007, 4.18, modifiée — Les notes à l’article d’origine ont été omises et une
nouvelle note à l’article a été ajoutée.]
3.2 Symboles
Une série de i mesures (i allant de 1 à I) couvre k concentrations (k allant de 1 à K), pour lesquelles
j répétitions ont été réalisées (j allant de 1 à J). Les indices sont notés dans l’ordre suivant: i,j,k. Les
variables aléatoires sont notées en lettres majuscules et leurs valeurs en minuscules. Une description
des abréviations utilisées dans les formules est donnée dans le Tableau 1.
Tableau 1 — Définition des différentes abréviations utilisées dans les formules
Symbole Description
Valeur de référence attribuée à un échantillon d’étalonnage pour une série i (1 ≤ i ≤ I), une répétition j
(1 ≤ j ≤ J) et une concentration k (1 ≤ k ≤ K)
x ou
ijk
Valeur de référence attribuée à un échantillon de validation pour une série i, une répétition j et une
concentration k.
Mesure de la réponse instrumentale ou expérimentale observée pour un échantillon d’étalonnage ou
y
ijk
un échantillon de validation pour une série i, une répétition j et une concentration k.
Valeur déduite pour un échantillon de validation pour une série i, une répétition j et une concentration k,
z
ijk
obtenue soit par prédiction inverse au moyen d’un modèle d’étalonnage, soit par mesurage direct.
Biais exprimant l’erreur de justesse pour un échantillon de validation entre la valeur déduite
b
ijk
et sa valeur de référence b = z − x .
ijk ijk ijk
4 Principes généraux
4.1 Rappel
Le profil d’exactitude permet de réaliser la validation selon une approche statistique. La Formule (1) est
utilisée pour décrire une mesure, z, d’un mesurande, Z, obtenue par un laboratoire:
z = m + B + e (1)
où
m est la moyenne générale pour l’échantillon homogène envoyé aux laboratoires;
B est la composante de biais du laboratoire sous des conditions de répétabilité;
e est l’erreur aléatoire survenant dans chaque mesure, sous des conditions de répétabilité.
Dans le cadre d’une étude interlaboratoires, la composante de biais B est due au laboratoire, mais elle
peut également être due à une autre source d’incertitude dans le cadre d’une étude intralaboratoire,
telle que le jour, l’opérateur, l’appareillage, etc.
Outre les méthodes statistiques de calcul des critères d’exactitude, le présent document donne
également des détails sur l’organisation du recueil de données ainsi que les précautions à prendre.
4.2 Conditions diverses pour l’estimation de la fidélité
Conformément à sa définition, la fidélité peut être estimée sous diverses conditions. Dans tous les
cas, la fidélité est quantifiée sur la base d’un écart-type, que ce soit pour la répétabilité s , la fidélité
r
intermédiaire s ou la reproductibilité s . Une échelle de complexité peut être établie entre ces
IP R
différents écarts-types, conformément au nombre de sources d’incertitude. La Figure 3 présente cette
graduation, avec en premier les conditions de répétabilité pour lesquelles aucun facteur de variation
et/ou aucune composante de variation systématique pour le calcul de l’écart entre répétitions n’est
identifié, puis les diverses possibilités pour l’estimation de la fidélité intermédiaire, et enfin, les
conditions de reproductibilité pour lesquelles le nombre de sources n’est pas connu.
Afin de simplifier la présentation, la notion de série fait référence à un ensemble de répétitions réalisées
sous des conditions de répétabilité: des groupes de séries avec la totalité des mesurages effectués sous
les mêmes conditions, par exemple le même jour, par le même opérateur ou sur un court intervalle de
temps. Pour certaines méthodes s’appliquant à des échantillons hautement instables dans le temps, il
convient que l’effet série choisi soit l’opérateur plutôt que le jour; la série inclura ainsi les répétitions
réalisées par le même opérateur:
6 © ISO 2020 – Tous droits réservés
Figure 3 — Diverses estimations de la fidélité d’une méthode conformément aux sources
de variation impliquées
Des modèles plus sophistiqués peuvent être employés, comme c’est le cas dans l’exemple ci-après, où
différents laboratoires, jours, opérateurs et appareillages ont été combinés pour donner quatre séries
selon un plan d’expériences multi-factoriel comptant trois facteurs.
Série Laboratoires Jours Opérateurs Appareillages
1 1 Jour 1 (−) Opérateur 1 (−) Appareillage 1 (+)
2 1 Jour 2 (+) Opérateur 1 (−) Appareillage 2 (−)
3 1 Jour 1 (−) Opérateur 2 (+) Appareillage 2 (−)
4 1 Jour 2 (+) Opérateur 2 (+) Appareillage 1 (+)
En général, il convient que le choix des sources de variation pour la série de mesure reflète aussi
fidèlement que possible les composantes de variabilité qui sont susceptibles d’être rencontrées lors de
l’application courante de la méthode à valider.
NOTE Pour atteindre les objectifs décrits dans le présent document, il est essentiel de recueillir des données
au moyen de plusieurs séries et de vérifier les sources de variation. Dans le cas contraire, il sera impossible de
construire un profil d’exactitude.
4.3 Profil d’exactitude
À partir de l’écart-type de fidélité intermédiaire ou de reproductibilité, calculé conformément aux
calculs présentés à l’Annexe A, l’intervalle de tolérance au niveau β peut être obtenu, lequel intègre un
pourcentage, β, de futurs résultats.
Tous les calculs sont réalisés séparément pour chaque concentration k, ce qui permet d’obtenir k écarts-
types de fidélité, et donc k intervalles de tolérance, qui sont réunis pour construire le profil d’exactitude.
La Figure 4 présente un exemple de profil d’exactitude construit à partir de trois concentrations,
0,4 mg/L, 2,0 mg/L et 4,0 mg/L, qui définit le domaine d’application ou domaine de validation de la
méthode à valider.
Le profil d’exactitude comprend les éléments graphiques suivants:
— sur l’axe horizontal: les concentrations théoriques (les valeurs de référence de concentration);
— sur l’axe vertical (simultanément):
— les limites des intervalles de tolérance au niveau β, calculées à partir des concentrations
déduites et exprimées en pourcentages (taux de recouvrement ou exactitude relative);
— les intervalles d’acceptation, définis conformément à l’objectif de la méthode et exprimés de la
même manière que les intervalles de tolérance.
La stratégie d’interprétation de ce graphique est décrite en détail en 5.10. Néanmoins, les limites
d’acceptation et le pourcentage β, qui est utilisé pour calculer les intervalles de tolérance, dépendent
strictement du contexte d’utilisation de la méthode et doivent être adaptés à chaque cas individuel.
Sur la Figure 4, dans la zone entre les droites verticales à tirets, la méthode permet de produire un
pourcentage moyen, β, de résultats qui s’inscrit dans les limites d’acceptation: la méthode est donc
valide dans ce domaine d’application. Le domaine d’application de la méthode est le domaine qui a été
initialement choisi pour la validation.
Légende
1 taux de recouvrement moyen (%)
2 limite de tolérance supérieure
3 limite d’acceptation supérieure
4 limite de tolérance inférieure
5 limite d’acceptation inférieure
6 limite de quantification
7 domaine de validité
8 domaine de validation
X concentration connue (%)
Y recouvrement (mg/L)
Figure 4 — Profil d’exactitude créé à partir de trois concentrations
Chaque cercle gris représente le pourcentage de la concentration déduite moyenne pour le niveau,
exprimé sous la forme d’un taux de recouvrement moyen (%) pour la concentration, et quantifie
la justesse. Les droites à tirets délimitent l’intervalle d’acceptation, et les droites en trait continu
délimitent l’intervalle de tolérance, calculé à partir des écarts-types de fidélité intermédiaire pour
8 © ISO 2020 – Tous droits réservés
chaque concentration. Les droites verticales délimitent le domaine de validité, au sein duquel la méthode
permet de produire un pourcentage élevé et connu de résultats acceptables.
D’une manière générale, pour satisfaire aux modèles statistiques utilisés, des répétitions doivent être
réalisées sous des conditions de fidélité intermédiaire ou de reproductibilité, sur des échantillons de
validation homogènes.
5 Mode opératoire
5.1 Définition de la grandeur mesurée
Définir la grandeur mesurée conformément au mode opératoire de la méthode, en spécifiant les
formules utilisées pour calculer le résultat final et la méthode utilisée pour parvenir à ce résultat. Dans
le secteur des cosmétiques, les méthodes utilisées sont souvent des méthodes indirectes ou rationnelles
qui exigent un étalonnage préalable pour calculer la concentration des échantillons non connus. Ces
méthodes reposent sur une approche en deux étapes: la courbe d’étalonnage doit tout d’abord être
tracée, sur la base du même principe physico-chimique que celui utilisé pour les échantillons; les
mesurages sont ensuite réalisés sur les échantillons non connus, et leurs concentrations sont calculées
à partir du modèle d’étalonnage. Dans le cas des méthodes directes, qui n’exigent pas d’étalonnage, la
situation est encore plus simple, car aucune courbe d’étalonnage n’est nécessaire et le mesurage sur les
échantillons non connus donne directement la concentration.
Il est essentiel que la grandeur mesurée pendant l’essai corresponde à celle qui sera mesurée lors des
essais de routine. Notamment, si le mode opératoire prévoit que le résultat final soit exprimé d’après un
seul mesurage, le résultat d’un essai de validation ne doit pas être exprimé sous la forme d’une moyenne
de plusieurs répétitions.
5.2 Définition des objectifs
5.2.1 Choix du domaine de validation
Définir le domaine de validation de la méthode sous la forme d’une plage de concentrations absolues
ou relatives, par exemple entre 1 g et 60 g ou entre 1 µg/L et 500 µg/L. En pratique, c’est grâce aux
choix faits (en termes de plage de concentrations pour un ou plusieurs types de matrice) au moment
de l’élaboration du plan d’expériences qu’il est possible de démontrer le domaine d’application au sein
duquel la méthode est effectivement valide et fournit des résultats acceptables (5.4 et 5.5). Le domaine
d’application couvert par cette démonstration est appelé domaine de validité et peut être plus réduit
par rapport au domaine d’application de la méthode précédemment défini.
Le choix du domaine d’application de la méthode peut correspondre à une exigence réglementaire.
Si, lors de l’application courante de la méthode, des échantillons dont la concentration ne se situe pas
dans le domaine de validité sont analysés, l’extrapolation n’est pas admise. Dans ce cas, il convient de
réaliser une étude complémentaire pour étendre ce domaine, ou une dilution peut être effectuée s’il
peut être démontré qu’elle n’a pas d’impact.
Une mauvaise définition du domaine d’application de la méthode a des conséquences notables sur la
validité, en particulier en termes de spécificité, d’interférence et de réactivité croisée.
5.2.2 Choix des limites d’acceptation
Les limites d’acceptation sont exprimées en valeur absolue dans l’unité du mesurande sous la forme ±λ
ou sont exprimées en valeur relative sous la forme (1 ± λ) × 100. Définir les limites d’acceptation en se
référant autant que possible à un document, à une pratique professionnelle ou à une exigence du client.
À titre informatif, il pourrait être mentionné que pour les produits pharmaceutiques, la valeur λ est
[11],[12]
habituellement fixée à 5 % . Toutefois, dans le domaine des cosmétiques, en raison de la complexité
et de la diversité des matrices, une valeur de référence de 10 % pourrait être proposée pour λ dans le
cas d’un dosage de constituant (au-dessus de 0,1 %). Pour l’analyse de traces dans les cosmétiques, cette
[13],[14]
valeur pourrait être comprise entre 20 % et 40 % .
Les limites d’acceptation sont généralement exprimées sous la forme de pourcentages. Afin de simplifier
la présentation des exemples qui illustrent le présent document, une valeur fixe a été choisie pour
l’ensemble du domaine d’application de la méthode. Toutefois, si le domaine d’application est large, des
valeurs variables en fonction de la concentration peuvent être utilisées.
5.3 Sélection des échantillons de validation
5.3.1 Choix du type de matrice ou des types de matrices
Il convient que les échantillons de validation soient des matériaux qui représentent le mieux le domaine
d’application de la méthode.
Choisir des matériaux stables et homogènes en quantité suffisante pour l’ensemble des essais prévus
dans le plan de caractérisation en vue de la validation.
5.3.2 Méthodes d’établissement des valeurs de référence
Pour estimer la justesse de la méthode, il convient de disposer d’échantillons de validation dont la
concentration est connue aussi exactement que possible avec une incertitude connue (ISO 11095:1996).
Cette concentration correspond à la valeur de référence attribuée à l’échantillon de validation et doit
être établie indépendamment de la méthode à valider. Elle est notée X. Il existe plusieurs méthodes
possibles pour établir la valeur de référence d’un échantillon de validation, notamment les suivantes:
1) utilisation de matériaux de référence certifiés (MRC), externes ou internes; la traçabilité de ces
matériaux décroît, dans cet ordre, en fonction de leur nature;
2) réalisation d’ajouts dosés d’une molécule étalon de pureté connue. Si le matériau de référence
contient déjà un analyte endogène, il convient que le plan d’expériences inclue un niveau
correspondant à l’absence d’ajout de sorte que les concentrations mesurées puissent être prises en
compte dans les calculs (voir 5.7.3);
3) préparations d’échantillons dopés ou de synthèse. Il convient d’effectuer le dopage dès la première
étape de la préparation du matériau afin de prendre en compte les interactions analyte-matrice.
Lors de la préparation d’échantillons de validation par ajouts dosés de diverses concentrations, il est
important de réduire autant que possible la dépendance entre les différentes préparations. Les ajouts
dosés peuvent se faire sous forme liquide (par ajout d’une solution étalon) ou sous forme solide (par pesée
...












Questions, Comments and Discussion
Ask us and Technical Secretary will try to provide an answer. You can facilitate discussion about the standard in here.
Loading comments...