An expression of within-plate uncertainty in sandwich ELISA.

This paper puts forward a method to describe an equation of the within-plate uncertainty (relative standard deviation (R.S.D.) of measurements) as a function of analyte concentration in sandwich enzyme-linked immunosorbent assay (ELISA). A kit for thyroid stimulating hormone is taken as an example. The pipetting procedures of analyte solution and chromogen-substrate solution and absorbance inherent to the wells of a microplate are identified as major error sources and their variability is included as parameters in the uncertainty equation. These parameters can be determined by the experiments with distilled water. The theoretical R.S.D. is shown to be in good agreement with the results of the repeated experiments using the real samples. Since the theory gives a continuous plot of R.S.D. against concentration, the uncertainty structure of the ELISA kit can be recognized over a wide concentration range and the detection limit and quantitation range can easily be determined on the plot.