Food Protection Trends - November/December 2017 - 394

investigated; nine out of ninety (10%) bottles had a built-in
carbon filter.
Sample collection and surveying process
The sample collection and surveying process was set up
in a linear fashion in a high traffic corridor of a Midwestern
college campus building. Passersby were asked if they had
a reusable water bottle with them and whether they would
be willing to participate. Upon agreement, they handed
over their water bottles, which were then emptied of any
liquid. Gloved researchers took each bottle, labeled it with
a sequential number, and, using an iPad set to the Qualtrics
survey, took an embedded photo of the bottle and label. The
iPad was then handed to the respondent, who completed
the survey. Surveys were coded with an anonymous number, which was paired to the contamination testing by the
use of adhesive labels with matching numbers.
Exterior surface testing by ATP bioluminescence
While the respondent was completing the survey, the
researchers placed a 4" × 4" template over the middle of the
bottle. This template was used to ensure that ATP swabbing
was consistent across all bottles. Ultrasnap ATP test strips
(Hygiena, Camarillo, CA) were removed from a refrigerator
and were allowed to warm up to room temperature for ten
minutes prior to use. The delineated space was swabbed
vertically and horizontally while the swab was rotated and
slight pressure was applied, after which the swabs were run
through the SystemSURE plus luminometer (Hygiena,
Camarillo, CA).
Interior surface sampling
Two 100 ml bottles of prepared 3M Phosphate Buffer
solution plus Tween20 (PBS-T) were added to each water
bottle. The solutions were prepared by taking purchased
99ml bottles of sterile Phosphate Buffer Solution (PBS) and
aseptically pipetting 1 ml of 10% Tween20 to each bottle,
thus filling the bottle to a full 100 ml. Prepared solutions
were held under refrigerated conditions overnight until use.
Once PBS-T was added to each reusable water bottle,
the lid was replaced and the bottle was shaken by hand 60
times, using a one-foot stroke. Subsequently, the sample
was aspirated through the opening of the bottle (either
poured or squeezed); a minimum of 80% of the PBS-T
volume was collected in a new, sterile container. Two new
water bottles were purchased from a local retailer and
processed through the same ATP and microbial testing
procedures as a means of control and comparison. Once
all bottles were photographed, swabbed, and sampled,
the PBS-T samples were shipped overnight in refrigerated
conditions to the NSF International Applied Research
Center (789 North Dixboro Road, Ann Arbor, MI 48105)
for HPC and coliform counts analyses.

394

Food Protection Trends November/December

HPC and coliform counts
Membrane filtration and pour plating were used to
enumerate heterotrophic bacteria and coliform bacteria
from PBS-T samples, respectively. Samples were diluted
in PBS as needed to ensure reliable colony counts. For
HPC processing, aliquots were pour plated with Standard
Plate Count (SPC) agar (one agar plate per dilution). For
coliform processing, 100 mL of each sample was filtered
through a 0.4 micron pore size membrane and plated onto
mEndo agar. All plates were incubated at 35 ± 1°C for 24
± 2 h. After incubation, colonies were enumerated and
reported as CFU/mL of eluent for HPC and CFU/100 mL
of eluent for coliforms.
Post-test cleaning
Once the respondent moved through the sample
collection and surveying process, their water bottles were
subsequently cleaned through the use of a standardized
three compartment sink cleaning process, which included
(1) washing in soapy water (Apex™ Presoak detergent,
Ecolab Inc.), using a bottle brush to clean the interior
of each bottle, (2) rinsing in clean tap water, and (3)
immersing in a sanitizing solution of approximately 300
ppm Oasis 146 Multi-Quat sanitizer (Ecolab Inc.) for
a minimum of five seconds, after which the bottle was
inverted to allow the interior to air/drip dry, while the
exterior was wiped dry. In addition to cleaning the bottle,
the owner was offered a choice of cookie as a thank you for
completing the testing and surveying process.
Survey design
The survey assessed respondents' behaviors in terms of
bottle usage and cleaning (Table 1). The first section of
the survey was descriptive in nature, asking respondents
to describe their bottle both in text and by selecting a
graphic representation. Respondents were then asked to
estimate the age of their bottle, and also to indicate what
types of liquids (water, soda, juice, energy drinks, etc.)
they had put into their bottle over the prior seven days.
The following questions addressed the frequency of bottle
use (both filling and refilling) as well as how often they
emptied their bottle and whether or not they had shared
their bottle with others. The second section of the survey
focused on cleaning behaviors, with questions pertaining
to frequency of cleaning and method of cleaning. The last
section solicited demographic information, such as gender
and age. The survey was submitted to the University's
Human Research Protection Program for approval by the
Institutional Review Board, which was granted prior to
commencing the study
Statistical analysis
Upon completion of data collection, a database of ATP,
HPC and coliform counts was compiled and sequenced by

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