10
I. Karunasagar
Table 1.3 The number of samples required, for a given level of confidence, at various frequencies
of contamination (FAO/WHO 2011b)
Confidence required that the test result is correct
Prevalence
50 %
90 %
95 %
99 %
99.9 %
Number of samples that must be tested:
Acceptable proportion
of contaminated
samples
Ä1 in 10
7
22
29
44
66
Ä1 in 100
69
229
299
459
688
Ä1 in 1,000
693
2,301
2,995
4,603
6,906
Ä1 in 10,000
6,932
23,025
29,957
46,050
69,080
Ä1 in one million 693,148 2,302,594 2,995,750 4,605,202 6,908,723
Thus, the sampling plan for any microbiological criterion should be able to
provide the desired level of confidence at these low levels of prevalence. Table 1.3
shows the relation between prevalence levels and the number of samples that need
to be tested to give various levels of confidence. To be able to detect a prevalence
level of 1 % at 95 % confidence, 299 samples need to be tested.
Salmonellosis associated with bivalve molluscan shellfish is rarely reported, as
indicated by epidemiological data. In US, since 1990, there have been three reported
outbreaks: one outbreak in 2003 involving six cases of typhoid fever associated with
a single oyster lease which was determined to be caused by overboard discharge
by immigrant workers on harvesting vessel; an outbreak caused by S. typhimurium
involving steamed clams affected five people in a private home in 1991 and another
outbreak caused by S. typhimurium involving molluscan lapas affected four people
in a retirement home in 1990. In UK, there were three bivalve molluscan shellfish
associated outbreaks between 1992 and 2007. An outbreak caused by S. enteritidis
involved cockles in 1997; another outbreak caused by S. enteritidis involved scallops
and king prawns in 2001; and the third outbreak caused by S. typhimurium involved
mussels in 2005. In France, there were 31 Salmonella outbreaks associated with
molluscan shellfish between 1996 and 2005, of which 19 were due to two common
serovars (6 due to S. typhimurium and 13 due to S. enteritidis). A study in 1997
showed that France accounts for 28 % of EU bivalve mollusc consumption (Girard
and Mariojouls 2000).
While the available epidemiological information suggests very low incidence
of salmonellosis, it must be noted that the extent to which this reflects the actual
situation is difficult to determine. Among other factors, the available data is
limited to certain geographical areas, bivalve molluscs may not be adequately
considered in investigating outbreaks of salmonellosis and information on the extent
to which bivalves are consumed raw is lacking. For those areas where information
is available, estimates of underreporting of salmonellosis have been made, and the
multipliers range from 6.9 for Australia (Hall et al. 2006) to 3.2 in England (Wheeler
et al. 1999) and 29 in the US (Scallan et al. 2011). While the limited epidemiological
record makes it difficult to determine the extent of the problem, given that we are
essentially dealing with a ready-to-eat food, the issue of under-reporting cannot be
dismissed without further investigation.
I. Karunasagar
Table 1.3 The number of samples required, for a given level of confidence, at various frequencies
of contamination (FAO/WHO 2011b)
Confidence required that the test result is correct
Prevalence
50 %
90 %
95 %
99 %
99.9 %
Number of samples that must be tested:
Acceptable proportion
of contaminated
samples
Ä1 in 10
7
22
29
44
66
Ä1 in 100
69
229
299
459
688
Ä1 in 1,000
693
2,301
2,995
4,603
6,906
Ä1 in 10,000
6,932
23,025
29,957
46,050
69,080
Ä1 in one million 693,148 2,302,594 2,995,750 4,605,202 6,908,723
Thus, the sampling plan for any microbiological criterion should be able to
provide the desired level of confidence at these low levels of prevalence. Table 1.3
shows the relation between prevalence levels and the number of samples that need
to be tested to give various levels of confidence. To be able to detect a prevalence
level of 1 % at 95 % confidence, 299 samples need to be tested.
Salmonellosis associated with bivalve molluscan shellfish is rarely reported, as
indicated by epidemiological data. In US, since 1990, there have been three reported
outbreaks: one outbreak in 2003 involving six cases of typhoid fever associated with
a single oyster lease which was determined to be caused by overboard discharge
by immigrant workers on harvesting vessel; an outbreak caused by S. typhimurium
involving steamed clams affected five people in a private home in 1991 and another
outbreak caused by S. typhimurium involving molluscan lapas affected four people
in a retirement home in 1990. In UK, there were three bivalve molluscan shellfish
associated outbreaks between 1992 and 2007. An outbreak caused by S. enteritidis
involved cockles in 1997; another outbreak caused by S. enteritidis involved scallops
and king prawns in 2001; and the third outbreak caused by S. typhimurium involved
mussels in 2005. In France, there were 31 Salmonella outbreaks associated with
molluscan shellfish between 1996 and 2005, of which 19 were due to two common
serovars (6 due to S. typhimurium and 13 due to S. enteritidis). A study in 1997
showed that France accounts for 28 % of EU bivalve mollusc consumption (Girard
and Mariojouls 2000).
While the available epidemiological information suggests very low incidence
of salmonellosis, it must be noted that the extent to which this reflects the actual
situation is difficult to determine. Among other factors, the available data is
limited to certain geographical areas, bivalve molluscs may not be adequately
considered in investigating outbreaks of salmonellosis and information on the extent
to which bivalves are consumed raw is lacking. For those areas where information
is available, estimates of underreporting of salmonellosis have been made, and the
multipliers range from 6.9 for Australia (Hall et al. 2006) to 3.2 in England (Wheeler
et al. 1999) and 29 in the US (Scallan et al. 2011). While the limited epidemiological
record makes it difficult to determine the extent of the problem, given that we are
essentially dealing with a ready-to-eat food, the issue of under-reporting cannot be
dismissed without further investigation.