SPSFAM Heavy Metals ERP Book

Does the method perform according to the method"s applicability as written? If No, please explain why the method did not perform according to the method"s applicability as written.

No. Infant formula should be removed from the applicability statement because the infant formula samples failed for RSD(r) for all 4 elements and failed for RSD(R) for As and Pb.

2. SAFETY CONCERNS

Are there any safety concerns identified while using or regarding use of the method? If Yes, please identify the safety concerns regarding the use of this method.

No. The method does a good job of addressing safety concerns.

3. STATISTICAL REVIEW

Is the documentation for the statistical review prior to AOAC First Action Official Methods status available and completed?

Yes. However, the number of samples with RSD(r) and RSD(R) meeting the SMPR for As, Cd, Hg and Pb were 4, 3, 2, and 2, respectively. I do not think a MLT can be supported when only 2 samples were acceptable.

4. REFERENCE MATERIALS

Are there available reference materials? Were there any concerns identified while using or regarding use of the method?

Yes, reference materials are available for all 4 elements. However, it is difficult to find one material that is certified for all 4 elements at useful levels. Therefore, more than one reference material typically has to be used in an analysis. The problem with this MLT is that too few reference materials were used to cover the 4 elements.

5. SINGLE LABORATORY VALIDATION

Is there data demonstrating linearity, accuracy, repeatability, LOD/LOQ? If No, please explain.

There is a control listed for the linearity of the calibration curve of r ≥ 0.995. However, there was no data or controls listed for linearity of sample data. Accuracy was established by the analysis of 8 CRMs. LOD was calculated as 3 times the SD of 23 method blank results. The LOQ was calculated as 2 times the LOD. Repeatability and accuracy was demonstrated in the MLT by the analysis of 6 samples, 5 of which were reference materials. Repeatability and reproducibility was demonstrated in the MLT by the analysis of 6 samples. The number of samples with RSD(r) and RSD(R) meeting the SMPR for As, Cd, Hg and Pb were 4, 3, 2, and 2, respectively. Thirteen laboratories participated in the MLT. Two or three samples with acceptable data, in my opinion, is too few for a MLT. The uncertainty of the method and probability of detection were not addressed. Statistical theory can probably give an estimate of the chance for false positives and false negatives using the method's equation for detection limit calculation.

6. REPRODUCIBILITY/UNCERTAINTY AND PROBABILITY OF DETECTION

Does the available documentation demonstrate Reproducibility/Uncertainty and Probability of Detection? If yes, please specify the information provided to support the reproducibility of this method as written.