External validation of analytical methods

Method validation is aftermath of quality assurance systems development in analytical laboratories. The initial aim was to create a confidence relationship between a service provider and a customer. Initially perceived as a constraint, it quickly became a critical issue, as analysts used to work fairly independently. This attitude soon disappeared, as the laboratory had everything to gain from providing results the client-decision-maker could trust in. In fact, this approach proved highly successful, as the demand increased for analyses of all kinds, and still does. In just a few years, analysts have become validation experts. However, it is necessary to distinguish between in-house validation and external validation.

Historically, different in-house validation procedures have been set up on a sector-by-sector approach, as water analysts thought that they did not face the same problems as steel industry. For this reason, there are still grey areas in the validation parameter definitions. As a result, many method performance parameters are open to interpretation. The best example is the limit of quantification: it is widely used, even though there are dozens of calculation methods that produce different values. Better standardization of parameter calculation modes is certainly the approach that will resolve these problems.

This is also the case for external, or interlaboratory, validation. As we shall see, the principles for conducting corresponding studies and the related calculations are all covered by various international standards. They also call on dedicated statistical methods

Finally, validating a method means making standardized statistical calculations. This remark is not accidental since standardization effort is in line with obvious economic and technical concerns. Indeed, the progress made in analytical methods, and the confidence that people are placing in their results, means that they are now essential decision-making tools in trade, environmental or medical diagnostics. The main aim of external validation is therefore to provide a means of ensuring good comparability of measurements, preferably at international level.