AOAC Guidance on FA Immunoassay Validation (November 2023)

= ( 0 )

�� % 100 � � − �

1686

1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707

2.2.6. Prepare an operator characteristic (OC) curve for LOQ

2.2.6.1.

Use a normal distribution calculation function to calculate the probability of obtaining a result higher than the LOQ for the given concentration using the

calculated S i and assuming a normal distribution.

2.2.6.2.

Normal distribution curves can be estimated in Excel with the following function, where LOQ is set at the estimate calculated above, S i is calculated

according to the linear regression

= 1 – NORMDIST(LOQ, mean concentration, S i , 1)

2.2.6.3.

Plot the probabilities versus concentration to prepare the OC curve.

2.2.7. Calculations/procedures for multiple matrices

Model %RSD i across a range of concentrations (as described above) for all matrices together. It is generally expected that the relationship between %RSD i and concentration will be sufficiently similar across matrices to conduct a single analysis

2.2.7.1.

for the combined data from all matrices.

2.2.7.2.

If the relationship between standard deviation and concentration appears to be substantially different between matrices, consult with the AOAC Statistics

Committee and Expert Review Panel.

2.2.8. Example Data and Calculations

Table 2: Example Data

Slope

0.0755 0.1368

Si(0)

1708

Example Data

0 20 40 60 80 100

RSD(%)

0

1

2

3

4

RSD

RSD(Q)

1709 1710