Classification of methods
Abstract In this section three classes (A, B and C) of methods are identified and described. Class-A methods are validated; In class B one will find methods that are verified. Class C methods have been demonstrated.
This chapter has been written by Peter Schoenmakers, Stella Brudin and Peter Pruim.
KeywordsThis chapter was written by Peter Schoenmakers and Stella Brudin.
In this section three classes (A, B and C) of methods are identified and described. Class-A methods are validated; they are carefully tested and developed. The critical aspects are standardized and these are described in detail. The methods in class A are regarded as very reliable.
In class B one will find methods that are verified. The biggest difference between Class A and Class B methods is the degree of validation that has been preformed. This does not mean that this is a less 'good' method; the opposite is quite often true. The technical level of these methods can exceed those of the Class-A methods, a Class B method is much easier to customize for new purposes.
Class C methods have been demonstrated. In its simplest form it can be a chromatogram, where peaks belonging to the relevant components have been marked. Typical sources for these kinds of methods are suppliers of columns and equipment, but they can also be found on the Internet.
- Class A. Validated methods are "developed and comprehensively validated”
- Class B. Verified methods have been well developed in a specific lab.
- Class C. Demonstrated methods, where the effect has been shown.
A: Validated methods
Validated methods are thoroughly and carefully developed and tested.
In some areas, collections of methods which are very thorough and carefully developed are available. These methods have been tested and discussed in detail and where necessary improved and retested. The critical aspects are standardized and these are described in detail.
- Validated methods for the analysis of pharmaceuticals/drugs can be found in the Pharmacopoeia.
- In the American Society for Testing and Materials (ASTM) a wide range of methods in other areas can be found. In addition, there are a large number of national and international organizations that are active in the field of the distribution of analytical methods.
'Class A' methods are regarded as very reliable. As long as the correct method is used within the specified scope, the risk of encountering significant method related problems is unlikely. Since all the obvious problems already occurred during the extensive testing of the method they have been resolved. This detailed testing is the most important aspect of these kinds of methods.
In these types of validated methods almost all data is known. In a few cases the origin of the method is unknown; the method can have been added anonymously to the collection. The scope of a Class-A method including the results obtained (detection limit, repeatability, reproducibility) are well established, accurate data is also available. This does not always mean that the correct numbers for this type of method are shown. Sometimes there are systematic deviations and assumptions which are not entirely correct.
The methods provide consistent data with which the samples can be compared, even if they have been measured at different locations and/or laboratories. In order to establish relationships between chemical composition and properties of substances and materials, consistent data is needed; it is also needed for investigations (of pharmacologists, chemists, etc.).
The methods are progressing slowly and a number of laboratories are involved. As a result, the methods can serve as an illustration of an average level rather than a top level. Changing and modifying a method for new developments - is a slow and laborious process, renewed reproducibility test have to be made.
|Pharmacopeia, NEN, EPA, ASTM, etc.;
A Delineation scope
A Sample pre-treatment
A All metering
A Alternative parameters
A Limit of detection and workspace
A Possible interference / issues
A Other validation features
Not always correct
|Background via publisher / other users
|Use control standards or reference
Only few methods are developed to the level as specified in this table.
B: Verified methods
The common practice is that a method is developed, described and applied in one place. Subsequently, the method can be implemented at another location or to another analyst. The biggest difference between the Class A and Class B methods is the degree of validation that has been done. A Class B method does not have to be evaluated in detail at various locations and is not subject to a reproducibility test, but the detection limit, the upper limit of workspace (maximum concentration of the injection quantity) and the repeatability are almost always evaluated.
This does not mean that this is a less 'good' method; quite often the opposite is true. Good practice (in terms of accuracy, precision, sensitivity, resolution, analysis time, etc.) methods developed for their own use can be found in public literature, but also within companies and institutions. The technical level of these methods can exceed those of the Class-A methods, a Class B method is much easier to customize for new purposes. One consequence of the great flexibility and lower degree of standardization for a Class B methods compared to a Class-A method is that each implementation will vary.
Often, the scope of Class B methods, in regards of both the analysis and the components in terms of the matrix, is less firm than those of Class-A methods. Class B methods are usually investigated on one system and one type of column.
|Own job; Lab.journals; Public literature
P Delineation scope
P Sample (for) processing
P All measuring circumstances
P Alternative parameters
Limit of detection and workspace
P Potential interferences / issues
P Other validation features
|Variable (sometimes without problems, sometimes unexpected problems)
|usually necessary (gives rise to different versions)
Results should be consistent
|Reliable, provided that no unexpected disturbance occur
Changeable provided accuracy verified
|More information from the developer
Any other users
|Search and maintaining contact with developer / user
Class B methods are not subject to ring tests.
C: Demonstrated methods
A demonstrated method is in its most simple form a chromatogram, where peaks belonging to the relevant components have been marked. The components to be analyzed are generally known, but often not with the matrix taken into account. Typical sources for these kinds of methods are suppliers of columns, equipment or can be found on the Internet.
- The purpose of a publishing an application is to show that something works (a column, equipment, detectors, etc.), and not to develop an entire method.
- A Class-C method will rarely be a good and fitting method and the chromatographer searching for an application must always be aware of that and that an alternative method (column, detector, etc.) may give much better results.
- It is always advisable to search in different sources.
Specific date and details for Class-C method are usually unknown. This is particularly true for dynamic parameters which may briefly be summarized, so are the details of the sample, quantification is not always possible. These methods rarely discuss validation.
|Applications suppliers; Internet
N Delineation scope
N Sample (for) processing
N All measuring circumstances
N Alternative parameters
N Limit of detection and workspace
N Potential interferences / issues
N Other validation features
|• Many degrees of freedom
• Sometimes a lot of work
• Final result can greatly differ from starting conditions
|• Not without reliable validation
• Exchangeable provided accuracy has been verified
|Background through supplier
|• Use whenever possible better defined methods
• Consult different sources
• Ask for information from the source
• Find out about other potential users of the method
• Be critical and selective!