DemarcheISO17025
  • Home
  • PREVIOUS
  • NEXT
Validation

Method Validation Process

The validation process should identify the various components needed to characterize a method and validate its implementation within the laboratory.

Définitions

Definitions Related to Validation

The terminology used in method validation is highly specific. Familiarity with the relevant parameters is important before developing a validation protocol.

7.2 Selection, Verification, and Validation of Methods

7.2.1 Selection and Verification of Methods

7.2.1.1 The laboratory shall use appropriate methods and procedures for all laboratory activities and, where appropriate, for the evaluation of measurement uncertainty and for statistical techniques used to analyze data.

7.2.1.2 All methods, procedures, and supporting documentation, such as instructions, standards, manuals, and reference data relevant to laboratory activities, shall be kept up to date and readily available to personnel (see 8.3).

7.2.1.3 The laboratory shall ensure that it uses the latest valid version of a method unless it is not appropriate or possible to do so. Where necessary, the application of the method shall be supplemented with additional details to ensure consistent application.

7.2.1.4 When the customer does not specify the method to be used, the laboratory shall select an appropriate method and inform the customer of the method chosen. Methods published in international, regional, or national standards, or by reputable technical organizations, or in relevant scientific texts or journals, or as specified by the equipment manufacturer, are recommended. Laboratory-developed or modified methods may also be used.

7.2.1.5 The laboratory shall verify that it can properly perform methods before introducing them by ensuring that it can achieve the required performance. Records of the verification shall be retained. If the method is revised, verification shall be repeated to the extent necessary.

7.2.1.6 When method development is required, it shall be a planned activity assigned to competent personnel provided with adequate resources. Periodic reviews shall be carried out during development to confirm that customer needs continue to be met. Any change in requirements calling for modifications to the development plan shall be approved and authorized.

7.2.1.7 Deviations from methods for any laboratory activity shall occur only if the deviation has been documented, technically justified, authorized, and accepted by the customer.

Methodology

The laboratory should use test methods appropriate for obtaining the necessary information about the properties of the item tested (substance, material, product, or system), with a satisfactory level of confidence. This includes every stage of the method: sampling, handling, transportation, storage, and preparation of test items (valid methods).

Methods should be described sufficiently completely and unambiguously, including instructions for equipment use and, where needed, checks of proper functioning and initial setup.

When needed, clarifications or additional details may be added to a standardized method to improve its application (e.g., translation from another language, precautions not explicitly stated, lessons learned).

Try to avoid work instructions that:

- merely repeat the standard (which adds no value);

- are too superficial (a cover page and only three or four pages of text) to explain the method adequately.

Prefer instructions that:

- are written by the operator who actually performs the tests;

- accurately reflect what happens during testing;

- explain important equipment-specific points not covered in the general documentation.

This reference documentation should be accessible to personnel and updated regularly.

Deviations from these references may occur only if they are technically justified, authorized, accepted by the customer, and documented.

The methods used shall be appropriate to customer requirements and comply with the applicable technical requirements.

The laboratory should preferably use standardized methods and, where possible, reference the most recent edition of the standard.

If the customer does not specify a method, the laboratory is responsible for selecting the most appropriate one, choosing among standardized methods (where available and suitable) or non-standard methods (see below), and informing the customer.

When standardized methods are used, the laboratory must nevertheless confirm its ability to apply the method correctly.

The standard no longer makes the same distinction between standardized methods and other methods in this respect: the laboratory must document verification that it can correctly apply the method, including a standardized method, to obtain results with the required performance.

To be validated, methods should be:

- Representative: measure the relevant properties of the test item and provide an estimate sufficiently precise for the intended use of the results;

The required uncertainty (i.e., the uncertainty that must be attributed to the method) follows from this need for greater or lesser precision. It is therefore not an independent variable but depends on intended-use requirements;

- Reproducible: applicable in different environments, by different operators, using physically different but conceptually comparable equipment, while providing similar results within the uncertainty assigned to the method;

- Repeatable: provide similar results, as described above, within a given laboratory using specified equipment and personnel.

The fact that a test method is specified by the customer does not relieve the laboratory of its responsibility to ensure that the method is valid.

The laboratory shall inform the customer if it considers the proposed or required method inappropriate.

A reproducible method is also repeatable, whereas the reverse is not necessarily true.

Methods used by the laboratory can be classified into two categories:

- standardized methods;

- non-standard methods.

Non-standard methods may be developed by the laboratory (in-house methods) or defined by other organizations (e.g., research centers or technical and scientific associations).

Standardized Methods

These are test (or measurement) methods established by technical rules or national, regional, or international technical standards, or by equivalent documents issued or published by competent technical and scientific bodies.

They are considered valid by definition (meeting the requirements for representativeness, reproducibility, and repeatability), provided they are properly managed and implemented.

If the standard or reference document contains all the information necessary for correct performance, these methods do not necessarily need to be transcribed into the laboratory’s own technical procedures (although standards seldom describe the specific equipment in full).

The laboratory may nevertheless choose, for various reasons, to reproduce these methods in internal procedures, adding clarifications and details.

Special procedures are needed if proper use of the standardized method requires explanatory instructions or additional procedural detail, at the laboratory’s discretion and based on its experience.

When standardized methods are used, the latest version of the applicable standard or document should be referenced, except in specific situations where doing so is inappropriate or impossible (FLEX1 scope in the French accreditation system).

For example, methods based on the following normative references may be considered standardized:

Mandatory European regulations (directives and regulations);

Mandatory national regulations (laws and decrees);

Official pharmacopoeias;

ISO and IEC standards; EN standards (CEN and CENELEC);

UNI and CEI standards; DIN, BSI, and AFNOR standards;

MIL standards (U.S. military standards);

ANSI, ASME, and ASTM standards;

AOAC (Association of Official Analytical Chemists) and APHA (American Public Health Association) standards, etc.

Non-Standard Methods

When no standardized method is available, or when such methods are wholly or partly outdated, the laboratory may, with the customer’s prior agreement, use in-house methods developed from its own knowledge and experience or obtained from other sources.

These methods must always be documented as laboratory technical procedures and thoroughly validated.

The description of a non-standard test (or measurement or calibration) method should contain all the elements needed for its correct application, such as:

- Appropriate identification;

- Definition of the purpose;

- Identification, in more or less detail, of the items tested;

- Definition of the parameters or quantities to be determined and their applicable ranges;

- Identification of the equipment to be used and its performance characteristics, and of any measurement standards or reference materials required by the method;

- Definition of environmental conditions and all other applicable conditions (e.g., stabilization times);

- Detailed description of operations, including handling of test items, any calibration and initial adjustments, safety precautions, data-recording arrangements, etc.;

- Criteria for data analysis and presentation;

- Criteria or requirements for evaluating results and any related statements of conformity;

- Uncertainty associated with the results or procedures for estimating uncertainty.

7.2.2 Validation of Methods
7.2.2.1 The laboratory shall validate non-standard methods, laboratory-developed methods, and standardized methods used outside their intended scope or otherwise modified. The extent of validation shall be sufficient to meet the needs of the given application or field of application.

7.2.2.2 When changes are made to a validated method, the effects of the changes shall be documented and, where appropriate, a new validation shall be performed.

7.2.2.3 The range and accuracy of the values obtainable using validated methods, as assessed for their intended use, shall be relevant to customer needs and consistent with specified requirements.

7.2.2.4 The laboratory shall retain the following records as evidence of validation:
a) the validation procedure used;
b) the specification of requirements;
c) the determination of method performance characteristics;
d) the results obtained;
e) verification that the requirements can be fulfilled using the method; and
f) a statement on the validity of the method, detailing its fitness for the intended use.

Methodology

Validation of a test method means demonstrating its ability to provide, at a predefined confidence level (uncertainty), the necessary information about the properties sought in the item tested, in relation to the intended use of the results.

Validation should be planned, developed, controlled, and documented (records), and should lead to an unambiguous conclusion on the method’s suitability for its intended use.

Validation may be conducted occasionally in response to a particular innovation requirement, or systematically as part of strategic planning to improve laboratory techniques (preventive and improvement activity).

In all cases, it is preferable to avoid validating a method while it is being applied under a customer contract or during service delivery.

If validation evidence is not otherwise available, the laboratory must perform the method validation using appropriate procedures and taking the relevant requirements into account.

Validation should be more or less extensive and detailed depending on those requirements.

It should account for any critical aspects of sample handling.

Validation is required for all non-standard methods developed within or outside the laboratory, standardized methods used outside their normal scope, and substantial modifications to standardized methods.

The validation process should be proportionate to the requirements associated with the specific application.

Validation techniques are essentially divided into two types:

- The scientific approach: primarily provides evidence that the properties determined and their associated target uncertainty meet the requirements;

It then requires a demonstration that influence factors are controlled, either through calculations based on an understanding of the relevant physical phenomena or through experimental checks (evidence of reproducibility and repeatability).

- In a scientific validation approach aimed at demonstrating reproducibility and repeatability, influence factors may be varied one at a time while the other factors remain fixed, checking whether the combined effects on the final result remain within the method’s uncertainty.

- The comparative approach: validates the method by direct comparison with a similar, already-validated method, or with a measurement standard or reference material that accurately represents the quantities or properties sought, or through comparative testing programs (results should fall within the uncertainty assigned to the method).

Comparative tests do not necessarily demonstrate the method’s representativeness: a method can appear stable and accurate while remaining physically incorrect.

Copyright © DemarcheISO17025.com