Quality Criteria for Selecting an Analytical Method

The amount of information in scientific papers is generally less than is needed to establish effectively the methods described within them. There are some exceptions to this statement, but these depend initially on how the author wishes to describe the method and how the paper is reviewed before being accepted by the journal. Some methods may only have a relatively small number of samples analyzed on 1 or 2 days before publication of the manuscript. In contrast, other methods may have examined several thousand samples before the manuscript is written and submitted. The methods in the former case have virtually no robustness data and knowledge of operation compared with the latter. Some analysts develop more complex methods that are necessary to demonstrate their scientific expertise at the expense of their colleagues. 

Tire performance of an analytical method and its inherent reliability are characterized by a set of quality parameters that determine its applicability and its usefulness, b or a quantitative method, the most notable parameters are its precision, accuracy, and limit of detection. For a qualitative method, the most important characteristic is its reliability in the identification of the analyte. Since there may be found in the literature an enormous set of different methods for analyzing a particular analyte in a particular matrix, it must be decided which method is the most appropriate for the analysis. 

Quality criteria for quantitative analytical methods, in general, have been proposed or arc lo be proposed by several international organizations including the Association of Official Analytical Chemists, the Food and Drug Administration, the Codex Committee for Residues of Veterinary Drugs in Food, the International Dairy Federation, and the European Union. The European Union, in particular, has laid down minimum quality criteria for quantitative drug residue methods, 

True content (T) of reference sample ( g/kg) Accuracy as % deviation of mean from T Repeatability as coefficient of variation of mean 

Values also suitable for reference methods. Source From Ref. 18. 

---