Methods Reviewed by OMB in 2015

98 / 471

The mean log10 counts of the 3M Petrifilm RYM Count Plate and FDA BAM/ISO 21527 1
results were compared statistically. One (1) laboratory was identified in the FDA BAM/ISO 2
21527 high contamination level as an outlier by the Single Grubbs’ test and two laboratories in 3
the 3M Petrifilm RYM Count Plate high contamination level were identified as an outlier by the 4
double Grubbs’ test for outliers. However, no evidence of physical cause or suspicion of cause 5
was noted and it was determined that they would be included in the statistical analysis. 6
There were no statistically significant differences at the 5% level in log transformed means 7
between the 3M Petrifilm RYM Count Plate and FDA BAM/ISO 21257 methods, at any of the 8
three contamination levels. 9
10
3M RYM Count Plate incubated at 28oC and enumerated at 60 hours 11
12
The mean log10 counts of the 3M Petrifilm RYM Count Plate and FDA BAM/ISO 21527 13
results were compared statistically. One (1) laboratory was identified in the FDA BAM/ISO 14
21527 high contamination level as an outlier by the Single Grubbs’ test and two laboratories in 15
the 3M Petrifilm RYM Count Plate high contamination level were identified as an outlier by the 16
double Grubbs’ test for outliers. However, no evidence of physical cause or suspicion of cause 17
was noted and it was determined that they would be included in the statistical analysis. 18
There were no statistically significant differences at the 5% level in log transformed means 19
between the 3M Petrifilm RYM Count Plate and FDA BAM/ISO 21257 methods, at any of the 20
three contamination levels.. 21
22
23
Discussion 24
25
No negative feedback was reported to the study directors from the collaborating laboratories 26
in regards to the performance of the 3M PetrifilmRYM Count Plate. A few laboratories 27
indicated difficulty in spreading the test portion liquid aliquot over the entire surface of the plate 28
before the sample began to solidify in the Petrifilm medium. Several laboratories also indicated 29
that the colonies were very easy to interpret due to their size and color. One laboratory indicated 30
that the “Space saving benefits of Petrifilm over the traditional FDA-BAM method are readily 31
apparent.” 32
During the analysis of samples, several laboratories indicated deviating from the approved 33
protocol and therefore, their data was subsequently removed from statistical analysis. No 34
laboratories were removed from statistical analysis as a result of their data. For the raw ground 35
beef four laboratories were removed from statistical analysis. Laboratory 12 was unable to 36
enumerate samples by both the Petrifilm RYM Plate and reference methods on the same day. 37
All samples were plated onto DRBC, then stored at 2-8oC overnight and plated onto the Petrifilm 38
RYM plates the following day. Laboratories 6 and 8 prepared the reference method samples 39
using duplicate plating and not the required triplicate plating. Laboratory 4 (for both raw ground 40
beef and raw almonds) failed to enumerate the 3M RYM Petrifilm plates at the 60 hour time 41
point. Due to these deviations, these laboratories were removed from statistical analysis. 42
There were no statistically significant differences at the 5% level in log transformed means 43
between the 3M Petrifilm RYM Count Plate and FDA BAM/ISO 21257 methods for either 44
matrix evaluated in this study, at any of the three contamination levels. . No statistically 45
significant differences were observed between the 3M Petrifilm RYM Count Plate results when 46
enumerated at 48 hours vs. 60 hours and compared to the FDA BAM/ISO 21527 methods. No 47
statistically significant differences were also observed between 3M Petrifilm RYM Count Plates 48
that were incubated at 25oC and 28oC when compared to the FDA BAM/ISO 21527 methods. 49
Candidates for 2016 Method of the Year
84