Food Protection Trends - November/December 2017 - 427

analyses. In their study, Gallina et al. found that the excision
method was superior to swabbing methods at recovering
total viable counts from all species (19). However, sponge
swabbing on cattle, pigs and small ruminants resulted in
microbial recoveries that were very similar to those obtained
using the excision method (19).
Bolton, like Hutchinson et al., comments that a quantitative conversion factor between excision and swabbing has not
been established (4, 24), probably due to the many sources
of variation in swabbing data. Nevertheless, some studies
have attempted to calculate the relationship in different ways.
Edmonds claims that the recovery efficiencies of swabbing
with different materials compared with excision vary widely,
from 20% to 90% (14), and Pepperell et al. states that values
can even vary from 2 to 100% (35). Ghafir and Daube found
a recovery of 36% for E. coli for swabbing pig carcasses (600
cm2) compared with the destructive method (20 cm2), using
log CFU/cm2 in the calculation (20). Gallina et al. (19)
referred to the Italian ministry of Health (38), which states
that non-destructive methods recover 20% of the bacteria
compared with the destructive method. The European microbiological criteria (17) define microbial limits of "m" and
"M" for the destructive method, and m = 1.5/M = 2.5 log
CFU/cm2 Enterobacteriaceae is converted to m = 0.8/M = 1.8
log CFU/cm2 for non-destructive techniques when calculating 20% of the CFU/cm2 counts (note: not log-transformed
units). Gallina et al. found the CFU recovered by swabbing
was generally greater than 20% of the excision value, based on
median results (19). In our study, swabbing with gauze cloth
corresponded to a recovery of 28% for Enterobacteriaceae
counts (CFU/cm2) compared with the destructive method,
while the sponge stick had a recovery of 29%, sponge swab
8%, and wet-dry double swab 1%. All swab techniques had,
on average, 17% of the recovery of the destructive method on
warm meat samples. Although it is difficult to ascertain and
standardize a conversion factor, the Italian conversion factor
of 20% is similar to that observed with our results.
Growth of microorganisms during chilling is inhibited
by the low temperatures and dry meat surfaces. Our results
show that sampling warm meat provides higher recovery
of bacteria, and thus we assume that sampling from warm
carcasses provides a better overview of the slaughter hygiene
at an abattoir than samples from chilled carcasses. Samples
from chilled carcasses, however, may offer more information
about the post-slaughter and chilling processes. These will
vary between different slaughterhouses, depending on factors
such as air velocity, temperature, and relative humidity (5, 6).
The US approach and the techniques used in Denmark and
Norway involve sampling from chilled carcasses (9, 10, 33).
A major advantage of sampling chilled carcasses is that there
are fewer time constraints, so the operator has more time
to conduct the sampling appropriately. Blast chilling of pig
carcasses is reported to reduce E. coli numbers by 1 log (28).
Ware et al. found that sampling by excision achieved a higher

recovery after 24 h at 7°C than sampling with a sponge swab,
because of bacterial attachment in the intervening period and
therefore a reduction in the efficacy of swabbing (39). In our
study also, recovery using swabbing techniques, compared
with that using the destructive method, was slightly reduced,
from 17% to 14%, after chilling. This might be the effect of
bacterial attachment to the meat.
Chilled carcasses are further along on the meat value
chain towards the consumer than warm carcasses, and thus
the results from sampling may be of greater public health
significance. However, some abattoirs perform cutting and
deboning of warm carcasses immediately after slaughter (hot
boning), and sampling of chilled carcasses is therefore not
possible. Investigations of chilled carcasses also require the
use of analyses that take into account that Enterobacteriaceae
and E. coli are stressed and need resuscitation. The use of
non-selective medium (tryptone soya agar) in microbial
analyses of these bacteria (31, 32) has been recommended
for the recovery of stressed Enterobacteriaceae (12, 27). One
advantage of the NMKL125 method (Nordic Committee on
Food Analysis) is the inclusion of a resuscitation step in nonselective medium for stressed or sub-lethally injured bacteria
(32). This, together with the low incubation temperature
(37°C as opposed to 41.5 or 44°C) used for SimPlate, will
also contribute to recovery of injured bacteria (23).
The choice of hygiene indicator is another important
consideration. A quantitative bacterial indicator to be used
in a process monitoring method needs to provide results
that can be enumerated for a considerable proportion of
carcasses; at least 80% positive tests have been recommended (21). Our experiments used high inocula to ensure that
the results obtained could be interpreted, but this does not
reflect the "real life" situation should these methods be used
on routinely sampled cattle carcasses, where the contamination on many carcasses will be below the limit of detection.
Thus, E. coli that is at a lower prevalence and concentration
on carcasses will be a less suitable indicator for this purpose
than Enterobacteriaceae.
Although the destructive method gives a higher recovery, swab sampling, especially using gauze cloth, is easier
and quicker to use and has the highest recovery, after the
destructive method. This sampling technique was therefore
chosen as the method to be used for the subsequent baseline studies in European abattoirs.
ACKNOWLEDGMENTS
Funding: This work is published as part of the Hygenea
project and was supported by the Norwegian Research
Council, grant no. 244493/E50. The participating abattoirs
are acknowledged for their valuable contributions.

November/December Food Protection Trends

427

Table of Contents for the Digital Edition of Food Protection Trends - November/December 2017