Food Protection Trends - May/June 2020 - 149

Figure 1. Schematic of the division of a TMD into sides A and B.

Recovery and detection of microorganisms
A dry cotton-tipped swab (Puritan Medical Products
Co. LLC, Guilford, ME) was used to recover bacteria
from the TMD. Side A was swabbed first, and the swab
was immediately spread onto two tryptic soy agar (TSA;
Hardy Diagnostics, Santa Maria, CA) plates. This method
was chosen to reduce participant concerns about applying
a solution to the TMD surface that is not specifically
for cleaning or disinfection purposes. Prior to swabbing
side B, either treatment 1 or treatment 2 was applied to
the TMD. For treatment 2, the TMD was allowed to dry
completely prior to swabbing. After swabbing side B, the
swab was spread onto two TSA plates. Two negative control
TSA plates were also incubated to ensure that no outside
contamination had occurred. The plates were incubated at
37°C for 48 h, after which bacterial CFU were counted. For
each TMD, an average CFU count was calculated separately
for side A and B. Using the surface area of the TMD, the
bacterial concentrations in CFU/cm2 were determined.
For each TMD, up to 10 colonies were randomly
selected from the two TSA plates for both side A and B, to
differentiate between Gram-positive and Gram-negative
bacteria. To do this, a simple, non-staining potassium
hydroxide (KOH) test was used, as outlined by Suslow et
al. (17). A sterile loop was used to place one colony on a
sterile glass slide. One drop of 3% KOH was added to the
top of each colony, and the mixture was incubated at room
temperature for one minute. The mixture was then slowly
drawn upward with a loop. If the mixture contained viscous
strings attached to the loop, the colony was marked as
Gram-negative. If there was no change in the viscosity of

the mixture after one minute, the colony was categorized as
Gram-positive.
Statistical analysis
Prior to data analysis, all microbial data were log-transformed (e.g., log10 CFU/cm2+1) for visual convenience without loss of generality in statistical results, as described previously (6). More specifically, in some instances, the log value
would be negative because the CFU/cm2 is < 1 but > 0; by
adding one to the CFU/cm2 prior to log transformation for
all samples, all values are positive. Descriptive statistics were
performed to determine the mean and standard deviation for
bacterial concentrations on each side within each treatment
group. Thereafter, one-way analysis of variance (ANOVA)
was performed to compare treatments and to identify any
statistically significant differences, as defined by P ≤ 0.05.
Multiple comparisons of the means were performed using
the Steel-Dwass method, because of the lack of normality in
the data. For the survey data, Likert item data were treated as
nominal data on a continuous scale, and descriptive statistics
were used to calculate the mean and standard deviation for
responses to each question. In addition, ANOVA was performed to compare bacterial concentrations on Side A (i.e.,
prior to treatment) with self-reported cleaning habits. JMP®
Pro 13 (SAS, Cary, NC) statistical analysis software was used
for all data analysis.
RESULTS
Survey
The demographic information collected from the survey
revealed that of 100 participants, nine were faculty, 36 were

May/June Food Protection Trends

149

Food Protection Trends - May/June 2020

Table of Contents for the Digital Edition of Food Protection Trends - May/June 2020