Food Protection Trends - January/February 2015 - 44

Microbiological results
Simple statistics were calculated for L. casei counts
recovered at the different sampling locations to
determine the average overall spread of contamination
throughout the kitchen. The pooled results for all
treatment groups (Table 3) show the most common
kitchen locations or items contaminated by participants
during meal preparation. Regardless of participant
treatment group, 90.2% of all 123 prepared salads had
some level of contamination with the tracer culture, and
24.4% of all participants contaminated the fresh salad
heavily, with over 2.4 log CFU/g (75th percentile).
The largest source of cross-contamination during meal
preparation was the small dishcloth and large cloth
towels, which harbored mean levels of 3.74 and 4.43 log
units of L. casei contamination per towel, respectively.
ANOVA tables and mean pairwise comparisons for
L. casei counts were developed for each sampling
location for groups, meals, cues, and interactions
(P ≤ 0.10; Table 4). The oven handle counts differed
(P = 0.061) for the control and FSM groups (0.77 and
0.49 log CFU/handle, respectively). This could be
because no food safety concepts per se were provided
to the control group in their pre-meal preparation
informational session, while the FSM group had an
informational session covering the basic food safety
messages of clean, separate, cook, and chill. No
significant differences (P > 0.10) were found in
contamination levels at any sampling location between
participants who cooked entrees containing chicken
versus beef.
Differences in contamination levels were observed for
sink handles (P = 0.064) and in the finished RTE fruit
salad (P = 0.083) between the AdCM and FSM participant
groups who prepared entrees containing chicken. In
each case, the FSM participant group exhibited a lower
contamination level (approximately 0.5 log cycle) than the
AdCM group.
Both entrée recipes required participants to add salt to
the entrée during preparation. It is interesting that within
the FSM participant group only, a significantly higher
(P = 0.039) contamination level was observed on the
surface of the saltshaker for participants who utilized
ground beef as a recipe ingredient than for those who
used chicken. The ground beef-containing recipe required
participants to form meatballs, which likely resulted in
more extensive handling of the raw meat product using
either their hands or a kitchen utensil such as a dipper.
The handling/slicing of raw chicken breasts and placing
them into a bag of breading may have required less
direct hand contact, and perhaps led to less transfer of
contamination to the salt shaker. However, this difference
was observed only in the FSM group and not in the
control or AdCM groups.

44

Food Protection Trends

January/February

When participants across all groups who received cues
or no cues were compared, a significant difference was
observed only for sink handle contamination. Participants
who received external food safety cues tended to
demonstrate slightly lower contamination on sink handles
(P = 0.09), but the actual difference in counts (1.16 versus
0.94 log CFU/handle) was of little practical relevance.
Comparing participant groups by cues, small differences
(P ≤ 0.10) in contamination rates were observed only for
the small cloth towel and the oven handle. The mean
L. casei count for the small towel was 0.62 log CFU/towel
lower for AdCM participants who received cues than for
AdCM participants not receiving cues. A difference
(P = 0.089) in small towel counts between AdCM
participants receiving cues and FSM participants receiving
cues was noted, with the AdCM group resulting in slightly
lower contamination levels (0.67 log cycle difference).
Finally, a difference (P = 0.041) was observed in L. casei
contamination levels for the control group between
participants receiving cues and those not receiving
cues (mean levels of 0.56 and 0.98 log CFU/handle,
respectively). Although not a strong indication, there
may be a slight positive impact associated with the use of
external cues (i.e. refrigerator magnets) relative to lessening
cross-contamination events during meal preparation.
Comparing participant meals by cues, small differences
(P ≤ 0.10) in contamination rates were observed for the
sink handle, small cloth towel, and the large cloth towel
only. The mean L. casei count for the sink handle was
0.34 log CFU/handle lower for participants cooking beef
when cues were present versus when cues were absent.
A difference (P = 0.058) in large towel counts between
participants who handled beef while cues were present
versus participants who handled chicken with cues
present was noted, with the presence of cues resulting in
slightly lower contamination levels (0.56 log CFU/towel
difference). Finally, a difference (P = 0.093) was observed
in L. casei contamination levels on the small towel for
participants who cooked chicken with or without cues
(mean levels of 3.50 and 4.01 log CFU/towel, respectively).
Again, this may show a slight positive impact associated
with the use of external food safety cues relative to the level
of cross-contamination during meal preparation.
When the participant groups by meal by cues
interaction was compared, differences (P ≤ 0.1) in
contamination rates were observed for the small cloth
towel, oven handle, salt shaker, and refrigerator handle.
The mean L. casei count for the small dish cloth was 1.0
log CFU/towel lower for participants in the AdCM group
who were cooking chicken with external cues present
versus when cues were not present. There was also a
difference (P = 0.080) in the small dish cloth counts
between the AdCM group and the FSM group when both
were cooking chicken with cues present, with the AdCM
Food Protection Trends - January/February 2015

Table of Contents for the Digital Edition of Food Protection Trends - January/February 2015
