Validation and verification of an analytical method: what is the difference?
Validation and verification of an analytical method are two essential steps to demonstrate that a method produces reliable results suitable for its intended use. However, these terms are still frequently confused in testing laboratories. The distinction is particularly important in a management system compliant with ISO/IEC 17025, especially when a laboratory introduces a standardized method, modifies an existing method, or develops its own.
Summary
What is the validation of an analytical method?
Validation consists of demonstrating, through objective evidence and experimental data, that a method possesses the necessary performance for its intended use .
According to the Eurachem approach, it is therefore not enough to demonstrate that a method "works". It is necessary to establish the analytical requirements corresponding to the problem to be solved, evaluate the performance of the method and determine whether this performance actually allows these requirements to be met.
A full validation becomes particularly relevant when a laboratory:
develops an internal method;
uses a non-standard method;
substantially modifies a published method;
applies a validated method to a domain that was not covered by its initial validation;
must demonstrate that the performance achieved meets a new analytical use.
Validation should therefore be considered as a documented demonstration of fitness for purpose .
What is the verification of an analytical method?
Verification usually occurs when a method has already undergone appropriate validation, for example a standardized method or a recognized method from another organization.
The laboratory does not necessarily have to repeat the entire original validation. However, it must demonstrate that it is capable of achieving, under its own conditions , performance levels compatible with those required.
The FDA defines method verification as the process of demonstrating that a laboratory is capable of reproducing a validated method with an acceptable level of performance.
In other words:
Validation: "Is this method suitable for its intended use?"
Verification: "Is our laboratory capable of correctly using this method and achieving the required performance?"
Validation and verification of an analytical method: the essential difference
The difference lies mainly in the prior status of the method and the extent of evidence to be produced .
Situation | Generally appropriate approach |
New method developed in the laboratory | Validation |
Non-standard internal method | Validation |
Standardized method used in accordance with its scope of application | Verification |
A proven business method | Laboratory suitability verification |
Significant modification of a validated method | Validation or additional validation |
Extension to a new matrix or a new analyte | Evaluation of the change and further validation based on its impact |
The fundamental point is that the existence of a published or standardized method does not exempt the laboratory from demonstrating that it can apply it correctly . Eurachem specifically emphasizes that a laboratory adopting an already validated procedure must still confirm its ability to apply it.
What performance parameters should be evaluated?
The scope of the work depends on the method, its use, the matrix, and the risk associated with the results.
For a quantitative method, the characteristics that can be studied include, in particular:
selectivity or specificity;
the limit of detection (LOD);
the limit of quantification (LOQ);
the working interval;
analytical sensitivity;
accuracy and bias;
the recovery rate;
repeatability;
sensitivity;
intermediate precision;
reproducibility;
robustness.
For qualitative methods, other characteristics such as threshold concentration, diagnostic sensitivity and diagnostic specificity may become relevant.
Therefore, it is not scientifically justified to automatically apply the same validation protocol to all methods. The selected characteristics and their acceptance criteria should be established before the study , based on the intended use. Eurachem specifically recommends defining the desired performance, the materials used, the number of replicates, the statistical analysis, and the criteria for concluding suitability for use.
When should a method be validated or verified?
Let's take a concrete example from an agri-food laboratory.
A laboratory wishes to introduce a recognized standardized method for determining a contaminant in a food matrix within the method's scope of application. A targeted verification of the locally obtained performance will generally be appropriate.
On the other hand, if the laboratory significantly adapts this method to a very different matrix, modifies the extraction, substantially changes the instrumental conditions or extends the use beyond the validated range, a simple verification may become insufficient.
The scope of the work must then be determined based on the nature of the modifications and their potential impact on analytical performance.
Why is this distinction important under ISO/IEC 17025?
For an accredited laboratory, the issue is not merely terminological. Inadequate validation or verification can call into question the laboratory's ability to demonstrate the validity of its results.
A robust case file should allow us to trace, in particular:
analytical need → performance requirements → experimental protocol → results → statistical analysis → acceptance criteria → conclusion on suitability for employment.
This approach also facilitates the scientific defense of the case during an accreditation audit, a regulatory inspection, an investigation concerning a result or a dispute by a client.
Conclusion
The validation and verification of an analytical method share a common goal: to produce results reliable enough to enable informed decision-making. However, they do not address exactly the same question.
Validation primarily demonstrates that the method is fit for its intended use , while verification confirms that the laboratory can achieve the necessary performance with an already validated method .
A properly established strategy thus avoids two common errors: carrying out an unnecessarily costly full validation when a check would suffice, or, conversely, carrying out too limited a check when the changes made would require a more thorough validation.
For agri-food laboratories and other testing laboratories, MMNA Consultants can support the implementation of validation and verification protocols, statistical performance analysis, gap analysis and document preparation for ISO/IEC 17025 accreditation or audit.
Learn more about MMNA Consultants' services regarding the validation and verification of an analytical method .
Comments