Food Protection Trends - November/December 2021 - 559

were reduced to only include sites collected from the " wetend "
of the packinghouse, as this was the only area sampled
in the current study. In addition, only samplings on Thursdays
or Fridays were included in the analyses performed here
to determine the difference in samplings conducted after the
second weekly cleaning and sanitation event (performed on
Wednesdays). Logistic regression was used to determine the
relationship between the log odds of a Listeria positive sample
and if the sample was collected before or after increasing
the frequency of cleaning and sanitation. The same procedure
for logistic regression described above was used. No statistical
analyses were performed to test the quat powder, pipe
removal, and drain cleaning interventions because there were
too few samples.
RESULTS
Root cause analysis
At the RCA meeting an RCA team was assembled, which
included the packinghouse manager, the quality assurance
manager, the maintenance manager, and 2 members of the
Cornell team. The RCA team reviewed the historical results
to identify instances of repeat Listeria isolation (i.e., both
persistent and persistent transient) in the packinghouse and
opportunities for corrective actions. The historical results
showed, that while there were persistent Listeria strains
present in the packinghouse in the first years of sampling
(2017 and 2018; based on whole genome sequencing
data), the persistent strains were no longer present in the
following year (2019) (34). However, there was a pattern
of persistent transient Listeria, as indicated by the repeat
isolation of Listeria of non-matching subtypes (according to
WGS) from the same sites (e.g., drains, dead-end pipe and
catch pan area, forklift stops) within the packinghouse. In
addition, the historical results showed (i) the majority of the
positives in the packinghouse were from the wet-end (i.e.,
the area with the dump tank, flume, brush beds, and waxing
equipment), (ii) there tended to be a greater percentage of
positive samples in samplings conducted at the end of the
week compared to the beginning of the week (which is closer
to the weekly cleaning and sanitation, which was originally
performed on Saturday or Sunday), (iii) the deep square
drains in the packinghouse were commonly positive, however
after a deep cleaning event during the first year of sampling, a
decrease in drain positives was observed, (iv) forklift stops at
the dump tank loading area were commonly positive, and (v)
sites by the catch pan below the brush bed, which is drained
by a dead-end pipe, were commonly positive.
Next, brainstorming was performed by reviewing the
fishbone diagram (Fig. 2). Major bones (i.e., bones " i " to
" vi " ; see Fig. 2) were reviewed and prioritized in order of
importance; the following 3 major bones were prioritized
in this case: (iii) facilities, (iv) cleaning and sanitation, and
(vi) packinghouse equipment. Within each of these major
bones, the relevant minor bones were discussed to determine
their likelihood of contributing to the repeat isolation of
Listeria. After review and discussion of historical results with
the RCA team, the minor bones that were potential causes
requiring further explorations were determined. Once this
was completed for all relevant minor bones, the 4 minor
bones most likely to be the root cause of the repeat Listeria
positive sites were identified. The 4 minor bones selected
as being the most likely contributors to repeat isolation
of Listeria in this case were (i) cleaning and sanitation
protocols and schedules, (ii) the catch pan area (under the
" packinghouse equipment " major bone), (iii) forklift stops
(under the " facilities " major bone), and (iv) drains (under
the " facilities " major bone). For each of these 4 minor bones,
the " 5 why's procedure " was performed by asking: " what
part of this procedure likely contributed to the persistent
Listeria? " and " why is this procedure set up the way it is? "
Then, we continually asked 5 additional " why " questions
to get to the actual root cause. For instance, if the identified
problem was a persistent Listeria population at the end of the
catch pan and dead-end pipe, the " 5 Why " questions may
be: (i) Why is the persistent Listeria population found in
this area? Because the Listeria is living in the dead-end; (ii)
Why is the Listeria living in the dead-end? Because moisture,
apple juices/ organic matter, and Listeria cells get trapped in
the dead-end; (iii) Why do these things get trapped in the
dead-end? Because there is no easy way for these things to be
removed from the dead-end (iv) Why is there no easy way
for these things to be removed from the dead-end? Because
it is difficult to get cleaning and sanitation chemicals, as well
as brushes for mechanical cleaning, to reach the dead end;
and (v) Why is the pipe and dead-end designed the way it is
and why is a dead-end pipe in use in the facility? From there,
long- and short-term corrective actions to eliminate the root
cause(s) were identified. These corrective actions were then
prioritized based on the cost and ease of implementation. A
description of the root causes, long and short-term corrective
actions that were identified, and the interventions tested in
the packinghouse can be found in Table 1. See Fig. 1 for the
intervention implementation schedule.
Listeria population
Overall, 13% (22/172) of samples were positive for
Listeria spp. (including L. monocytogenes) across the 5
sampling events performed as part of the study reported here.
The 22 positive samples came from 9 sampling sites; these
sites include (i) 3 sites from a single dead-end PVC pipe that
drained the catch pan under the brush beds, (ii) 2 forklift
stops, (iii) 1 floor crack, (iv) and 3 sites within drains (2 sites
within a single square drain and 1 site from a trench drain
leading into the square drain) (Table 2).
The Listeria species isolated included L. monocytogenes,
L. innocua, and L. seeligeri; L. monocytogenes was isolated
from 11 samples, L. innocua was isolated from 5 samples, and
L. seeligeri was isolated from 15 samples (with a total of 22
November/December Food Protection Trends 559
Food Protection Trends - November/December 2021

Table of Contents for the Digital Edition of Food Protection Trends - November/December 2021
