Day 106 - 23 Mar 95 - Page 53
1 MR. MORRIS: That is the population of bacteria?
2 A. That is the statistical population, i.e. number of
3 hunks of meat in the batch or the number of birds or what
4 have you, is 400 or more; the number of samples you need to
5 take is 60. Where you are dealing, and I estimated roughly
6 that there are, perhaps, 100 or slightly over 100 pieces of
7 meet in one of those bins, and from memory reading off the
8 Ministry of Agriculture chart which is well-published,
9 widely available, and, as I say, frequently practised, you
10 would need to take 40 samples. So if you were
11 realistically looking for specific pathogens at a level of
12 5 per cent distribution within the whole or more, to give
13 you a 95 per cent probability of finding them, you would be
14 needing to take 40 samples per compo, not five.
15
16 If your expected contamination rate was less than 5 per
17 cent, you have then to take on that rate 400 samples. Now
18 one is not saying that this is realistic, but it shows the
19 statistical relationship between number of samples and
20 probability of finding. It is not, in my view, appropriate
21 to apply a sampling plan devised for determining TVCs to
22 the use of determining or detecting specific pathogens in
23 food.
24
25 Given a five sample, I cannot say what your statistical
26 probability of actually finding anything would be, but it
27 would be very unlikely, it would be remote, the chance of
28 actually finding anything unless the infection rate or
29 contamination rate was absolutely enormous. So that, in
30 effect, the rest of the sampling and, therefore, all the
31 results which stem from them are largely ritual. All
32 these massive bundles of documentation which show negative
33 results on salmonella, so on and so forth, are largely
34 meaningless, a complete waste of time and money. I must
35 say I have dealt with this many, many times in my career
36 and revised sampling programmes accordingly, and there are
37 millions of pounds spent throughout industry on these
38 ritual tests which are more of the security blanket nature
39 making everyone feel comfortable, part of the feel-good
40 factor rather than actually providing any real usable
41 information.
42
43 MR. JUSTICE BELL: Do you say that in relation to the tests for
44 all the organisms which are listed at the bottom of page
45 38?
46 A. E.coli, per se, is probably high distribution
47 throughout the meat.
48
49 Q. No. The comment which you have made, as I understand it,
50 it may be tested in due course, is that however appropriate
51 taking five samples in order to get a TCC may be, and we
52 may have to come back to what use finding out a TCC is so
53 far as finding out what numbers of pathogens may be, you
54 say it is really ritual and meaningless when one is testing
55 for specific pathogenic organisms?
56 A. That is absolutely the case, my Lord.
57
58 Q. To which of the organisms listed at the in the bottom part
59 of page 38 do you say that comment applies?
60 A. I do not take it as E.coli because E.coli is not a
