Apartfromthe fact that it is difficult to estimatethe numberof casesbasedon
an FSO(or the other wayaround, to derivean FSOfroman ALOP),it is also
verydifficult to set a specific ALOP.It is difficult bothto determine whatis
appropriateand also to `distribute' diseasecases overvarioustransmission
routes.For example, one can set a level for campylobacteriosis(as the public
healthgoal that is the result of foodtransmissionand othersources), and needs
to determinewhatthe FSOfor foodproducts should be. In that caseone should
selectwhatthe specific ALOPwill be for foodtransmissionof this organism,
sincethe FSOinfluencesonlythe foodtransmitted part of all the casesbut not
any other sources that can causecampylobacteriosis.Secondly,an additional
problemis that it is oftennot the productitself that gives the risk,but the fact
thatthe productcross-contaminatesotherproducts, via utensils,surfacesor
hands.The fact that it is difficult to set an ALOPdoesnot mean that it should
not be done.It is much betterto do it directlybasedon the currentstateof
knowledgeand data than to wait untilall informationis available, since this will
neverbe the case.However,if newinformation doesbecomeavailable, one
shouldevaluate whether the level shouldbe changed.
2.3.1 Distributionoverthe chain
A positiveaspect aboutthis conceptis that oncean FSOhas beenset, the
objectivescan be distributedoverthe wholechainfromprimary production to
consumption. A performance objective(PO)can be set for everylink in the
chain,so that in total the FSOis achieved.Thishas the great advantagethat the
mostefficientdistribution of the objectivesoverthe chaincan be found:one has
the flexibility to do morein the first stage,or in the last stage,or both.If the PO
has beenset for one stage,this can againbe distributed overvariousprocess
steps.Thisdefinesthe performance criterion (PC), for example for a reduction
step (pasteurisation) a 6 log reduction is necessary.Withthis criterion, one can
thendefineprocesscriteria that will attainthis reduction(e.g.72 ÎC, 15 s). This
is indicated in Fig.2.6, whichshows the relation withHACCP and critical
limits. The advantage of this concept is that one has the flexibilityto change
limits in one stage, as longas one equalises this in another. For example, a
processcriterioncan be changedso that only5 log reductionsare achievedif
this factorof 10 is balancedin another processstep,or evenin another stagein
the chain.
One of the problems in settingFSOs,and in relating FSOsto ALOPS, is the
fact that it is not the setting of a limitthat determines the healthburden, but that
in manycases extremelevelsare determining.Thiscan be illustrated by a very
largesurveypublished by Gombaset al.(2003)in which31 700 ready-to-eat
foodswere sampledforListeria. Of thesesamples, 1.8%werecontaminated
withListeria(577samples). Only2 out of the 577 positive samples (0.006% of
the 31 700 products) contained more than 105 organisms per gram. If we
determine the totalexposure of allListeriain theseproducts, these two samples
alonerepresented97.5%of the totalexposure in the 31 700 products, because of
The rangeof microbialrisksin foodprocessing 39