Concepts in the validation of neurochemical methods: the proper generation and use of statistics

Nearly every advance made in the study of neurotransmission has resulted from the development of new analytical methods or from the application of such methods to neurochemical problems in new ways. Each investigator places extremely high dependence on the laboratory method from which the data are gathered. It is therefore vital that these methods be proven valid when first selected and when used throughout the experiments being conducted. The process of method validation has flourished and been refined in the field of laboratory medicine. Methods are primarily validated by their accuracy and reproducibility in determining the analyte of interest in the tissue(s) to be studied. It is important that standards of performance be established that will allow objective decisions to be made when methods are tested for these characteristics. Performance standards for several neurotransmitters are suggested. Studies performed to collect method performance data are presented. From these data, statistics can be generated that help to estimate analytical errors. Guidelines for the proper generation and use of these statistics are discussed. Use of these validation approaches should be expanded in the whole of neurochemistry, which should enrich the data gathered within a laboratory and improve the harmonization between laboratories.