Assessing limits of detection in qualitative methods: A simple implementation of logistic regression in a web-based R...

The estimation of limits of detection (LOD) for solely qualitative methods in analytical chemistry may prove challenging because all the approaches with which chemists are familiar require some type of numeric data input. The best model to describe the binary response in these methods (detected/not detected) is a logistic model; however, these models are not easily handled by most of the laboratories and generally demand expensive statistical software packages. In this work, the advantages of applying this approach are discussed and its implementation using commercial spreadsheet software is demonstrated. A free online application based on the R environment using shinyapps was developed and its application was validated and discussed with a dataset of 57 different target compounds analyzed in urine according to the requirements of the World Anti-Doping Agency (WADA). This tool allows free, extremely quick, and easy determinations of LOD in qualitative analyses as well as the determination of the probabilities of detection in any given concentration.