Validation of HPLC and GC-MS systems for bisphenol-A leached from hemodialyzers on the basis of FUMI theory.

This paper proposes a method for the validation of chromatography systems in which many experiments to estimate SD or RSD are difficult or impossible to carry out because of time, cost, etc. HPLC systems with UV-Vis and fluorescence detectors and GC-MS system for bisphenol-A leached from hemodialyzers are taken as an example. Examined as validation characteristics are not only the ordinary quantities (precision, accuracy, range, limit of detection (LOD), limit of quantitation (LOQ), specificity and linearity) but also precision plots (measurement RSD vs. concentration), 95% confidence intervals of calibration lines and LOD signals over baselines. The precision plots, calibration confidence intervals and LOD signals are shown to be advantageous to validate and compare the analytical performance of the systems. The LOD, LOQ, precision plots and 95% confidence intervals of calibration lines are all derived from the SD of measurements and the reliability of these quantities and plots depends totally on the reliability of the SD estimates. This paper uses a probability theory, called the FUMI theory, to estimate as exact a measurement SD as possible without the replication. The precision of the HPLC and GC-MS systems is shown to coincide with the repeatability obtained by the repetition of measurements.