AOAC Guidance on FA Immunoassay Validation (August 2023)

6.4.1. AOAC Requirements: Collaborative studies will fulfill all other AOAC requirements for qualitative 1024 chemistry methods, including those in Appendix N. (13) 1025 6.4.2. Scope: A collaborative study is intended to evaluate the performance of the method across 1026 multiple laboratories. The collaborating laboratories should receive training, provided by the 1027 method developer, on the specific method under evaluation. 1028 6.4.3. Test Materials 1029 6.4.3.1. As in the SLV, for studies that claim applicability to finished product or ingredient analysis, 1030 incurred matrices are required. See Annex A for preparation guidelines. 1031 6.4.3.2. For studies claiming applicability to environmental surface samples, see Annex B for 1032 preparation guidelines. 1033 6.4.4. Study Design 1034 6.4.4.1. The minimum number of collaborating laboratories is based on AOAC Appendix N 1035 guidelines. (13) Currently, Appendix N requires usable datasets from a minimum of 10 1036 independent testing sites. It is recommended to recruit more than the minimum number of 1037 laboratories to ensure a sufficient number of usable datasets are acquired. 1038 6.4.4.2. Collaborating laboratories will perform the matrix/POD studies for at least one of the 1039 matrices for every five matrices claimed in the method developer study. The selection of 1040 which matrices are analyzed should be reflective of the range of difficulty associated with 1041 the claimed matrices. 1042 6.4.4.3. If the method developer study consisted of fewer than six total matrices, then the 1043 collaborator study will test at least one matrix from the single laboratory validation study. 1044 6.4.4.4. For each matrix, test materials should be prepared at each of the following concentrations: 1045 blank, 0.5x CDC (or other fractional recovery concentration, if found), CDC, and a high level 1046 at least 1.5x CDC. All samples must be blind coded when sent to collaborator sites. 1047 6.4.4.5. Each collaborator site will be provided with and will analyze a minimum of 8 test portions of 1048 each test material. Analysis of a larger number of test portions per laboratory may be 1049 considered, as more replicates will provide tighter confidence intervals for the POD 1050 estimate. (13) 1051 6.4.5. Data Analysis, Results Reporting, and Acceptance Criteria 1052 6.4.5.1. For each test material concentration, calculate the POD for each collaborator. Then 1053 calculate the LPOD (mean POD across collaborators) and LPOD 95% confidence interval. See 1054 Wehling et al. 2011 for additional details. (6) 1055 6.4.5.2. Report the LPOD estimates with 95% confidence intervals for each concentration. 1056 6.4.5.3. Calculate repeatability and reproducibility as described in LaBudde and Wehling 2023. (7) 1057 6.4.5.4. The CDC must be confirmed with a POD > 0.9, with 95% confidence. The entirety of the 1- 1058 sided 95% confidence interval for the CDC POD results must fall above 0.9. 1059 6.4.5.5. The POD(0) must be less than 0.1, with 95% confidence. The entirety of the 1-sided 95% 1060 confidence interval for the POD(0) results must fall below 0.1. 1061 6.4.6. Collaborator comments 1062