Planning a validation study involves careful consideration of several key points. This process ensures that the study is designed and executed effectively, yielding reliable and accurate results. By addressing these points, researchers can enhance the validity and credibility of their study findings.
1. The method to be validated
Before starting a validation study a detailed written procedure (such as a standard operating procedure) describing the method to be evaluated should be available. The formal validation should be considered separately from any method development activities. It is the ‘final’ version of the method – after completion of method development – that is validated.
2. Critical steps in the method and instrument requirements
Before starting the validation study the analyst should be familiar with the method and aware of any critical steps that require particular attention. Any specific requirements relating to equipment/instrumentation should also be considered.
3. Supporting information
There may be existing information available that can help with planning the validation study and/or demonstrating the fitness for the purpose of the method. This includes data from participation in interlaboratory comparisons (ILC), such as proficiency testing (PT) schemes, results from internal quality control (IQC), and results from previous routine use of similar methods.
4. Extent of the validation
One of the main issues facing laboratories when planning a validation is deciding which performance characteristics should be studied and the level of replication required.
5. Order of evaluation of performance characteristics
With careful planning, it is possible to obtain information on several performance characteristics from a single set of experiments. However, some characteristics should ideally be evaluated before carrying out a full precision or bias study.
Selectivity is generally studied very early in the validation process as without knowledge that the selectivity is acceptable, other performance characteristics will be of little value. In some situations, it may be advantageous to carry out a ruggedness study before the full precision and bias studies as it will provide information on the critical steps in the measurement process that need to be controlled.
However, regulatory requirements in some sectors stipulate that a ruggedness study should be carried out as the final stage of the validation.
ALSO READ: SOP for Analytical Method Validation
6. Materials to be analyzed
Guidance on the types of materials (e.g. reference materials (RMs), test samples) that can be analyzed is given in the sections for the individual performance characteristics. When planning the study, the scope of the method should be taken into account.
The aim is for the validation to cover a representative range of sample types in terms of matrix and analyte levels. This may require the analysis of several different materials including certified reference materials (CRMs), spiked samples, and test samples. It is also important to establish how much of each material will be required during the validation to ensure that sufficient material will be available.
7. Experimental design
Choosing suitable experimental designs is a key part of validation planning. With appropriate planning, it is possible to maximize the amount of information obtained from a particular experiment. For example, it may be possible to obtain information on more than one performance characteristic.
Several experimental designs can be used in a validation study. These include:
- Simple replication: This involves making a series of measurements on a single material. It is useful for estimating precision (particularly repeatability). If a reference value is available (e.g. if the material being analyzed is a CRM) the results from a simple replication study can also be used to evaluate bias.
- Linear calibration: This type of design is commonly used for instrument calibration, and studies of linearity and working range. This type of design involves observations at a range of levels (usually different analyte concentrations).
- Nested design (also known as a hierarchical design): This is an experimental design in which each level of a given factor appears in only a single level of any other factor. For example, in a study of repeatability and intermediate precision, replicate measurements obtained in a short period of time are ‘nested’ within days or analytical runs. Figure 1 shows an example of a single-factor nested design.
Figure 1: Example of a nested design for an experiment in which different precision measures can be evaluated if the groups represent different analytical runs (ideally carried out on different days).
The results from this type of experiment can be analyzed using a one-way analysis of variance (ANOVA).
Fractional factorial design: This is a factorial design from which some carefully chosen combinations of levels have been removed. This reduces the total number of measurements required in a study while still providing useful information. A fractional factorial design commonly used in method validation is a simple seven-factor design, known as a Plackett-Burman design.
ALSO READ: Checklist for Analytical Method Validation
Tags
Validation
