AOAC RI ERP Final Action Recommendations eBook

Dupont BAX Salmonella PTM Report Modification Approved 2012 / PTM Certification No. 081201 For Expert Review Panel Use Only Do Not Distribute

0

0

0-0.43

0

0

0-0.43

0

(-0.45, 0.45)

0.00

5

5

0.25

0.11-0.47

5

0.25

0.11-0.47

0

(-0.14, 0.14)

Dry pet food

BPW

0.29 (0.13, 0.53)

20

0

0

0-0.43

0

0

0-0.43

0

(-0.45, 0.45)

0.00

5

5

0.25

0.11-0.47

5

0.25

0.11-0.47

0

(-0.14, 0.14)

Dry pet food

LB

0.29 (0.13, 0.53)

20

0.00 (-0.45, 0.45) a Note: For all sample matrices, the BAX® System method was evaluated using both the reference and test method enrichments in order to validate multiple enrichment 5 0 0 0-0.43 0 0 0-0.43 0

326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358

media for each sample type.

b MPN = Most Probable Number is based on the POD of reference method test portions using the AOAC MPN calculator

c N = Number of test potions d Number of positive test portions

e POD CP = Presumptive candidate method positive outcomes divided by the total number of trials f POD cc = Confirmed candidate method positive outcomes divided by the total number of trials g dPOD CP = Difference between the candidate method presumptive and candidate method confirmed POD values h 95% CI = If the confidence interval of a dPOD does not contain zero, then the difference is statistically significant at the 5% level “If all results agree in a paired study, the measure of effect dPOD is zero, and the best estimate of s(dPod) = 0, so the t-based interval computes to zero as well. To get a non-zero confidence interval … some other method must be used. For this special case of n pairs all agreeing, one can use a binomial-based 95% confidence interval of +/- [1 - (0.05)^(1/n)] n = # pairs.” [6] For the results presented in this table, all results are paired and thus fall under this case.

AOAC INTERNATIONAL