Food Protection Trends - July/August 2017 - 242

infectious agent; its possible transmission routes; and
features such as incubation period, duration, and magnitude
of disease (18). An epidemic curve was created using
Microsoft Excel, and a transmission web of the outbreak
was created using a basic two-column Microsoft Excel
spreadsheet (one column for people infecting others and one
column for those they infected) imported into Cytoscape 3,
an open source software program for visualizing networks
(Cytoscape Consortium, San Diego, CA).
Infectious diseases can also be characterized by their
R0, which is simply the average number of people a single
infected person is expected to infect, assuming everyone
in the population is susceptible to the disease (10, 15).
The R0 represents the transmission potential of an infectious
agent; the higher the R0 relative to a benchmark of 1.0,
the more extensive the expected spread of disease (10,
15). Weighted averages of the number of returned buttons
(corresponding to 'confirmed' cases of norovirus) were
used to calculate R0 values.
RESULTS
Response rate
All 100 bags containing five buttons each were distributed
from the exhibit booth. Overall, 248 of the 500 people
potentially infected during the experiment visited the
NoroCORE exhibit booth (49.6% response rate), becoming
"confirmed" cases of norovirus, and 198 of these people
explained where and when they were infected (79.8%).
The 252 other individuals who may have been infected but
did not visit the booth were "suspected" cases of norovirus.
Ultimately, 88 individuals were responsible for spreading the
virus. However, because of unexpected enthusiasm for the
experiment on the first evening of the conference and the
number of buttons, restrictions had to be placed thereafter
on when infected individuals could receive buttons from the
exhibit booth.
Reproduction number
Excluding Patient Zero's initial ten infections (since these
were automatically recorded at the start of the outbreak),
the number of confirmed secondary cases arising from a
single infected individual varied from 8 to 0. As 77 of the 88
participants spreading the virus could distribute only five
buttons instead of ten, an adjusted weighted average of the
number of buttons that returned to the booth for each bag of
five buttons (Fig. 1) was used to estimate an R0 of 2.48.
Location of infection
Although the vast majority (n = 166, 84%) of infections
occurred in the exhibit hall, several (n = 16, 8%) occurred
in the conference oral presentation rooms (Fig. 2). The
"Reception" area represents the ten initial infections caused
by Patient Zero at the opening session reception (5%).
The five "Other" (2.5%) responses were specifically "City

242

Food Protection Trends July/August

Downtown" (n = 2), "Poster," "Lunch," and "Lobby"; some
of these may have been referencing the exhibit hall, but this
cannot be assumed.
Time of infection
Because the participants assigned their time of infection
to one of 12 four-hour time slots, an epidemiologic curve for
the mock outbreak was easily produced (Fig. 3). Unexpected
infection times (e.g., 12:00 a.m. - 4:00 a.m.) were confirmed
with the participant when they visited the booth. Because of
the high demand and limited supply, buttons were available
from the booth from 7:30 p.m. to 9:30 p.m. on Saturday, July
25th, 2015, and from 10:00 a.m. to 12:00 p.m. and from 5:00
p.m. to 5:20 p.m. on Sunday, July 26th. Recording of new
cases ended at 1:00 p.m. on Monday, July 27th, the last full
day on which the exhibit hall was open.
Web of infection
The node-link tree displays the entirety of the outbreak and
the connections between the individual participants (Fig. 4).
Blue dots signify infected individuals who had their button
number recorded at the booth ('confirmed' cases), while gray
dots represent buttons that may have been distributed but did
not come back to the booth ('suspected' cases). The longest
chain of confirmed cases reached eight degrees of separation
from Patient Zero, and distinct branches of transmission
arose from nine of the ten individuals initially infected at the
opening reception.
DISCUSSION
Public reception
Anecdotal feedback volunteered by the participants was
overwhelmingly positive; the attendees were perceived to be
enthusiastic and curious about participating, which generated
discussion about norovirus and similarly-transmitted agents.
People of all ages and genders participated and were a
representative cross section of IAFP membership, including
individuals from academia, industry, regulatory agencies, and
public health groups. Conference staff and other exhibitors
were also seen wearing the buttons. When the option to
infect other people was available, the majority of participants
did elect to distribute buttons. However, this enthusiasm
made it necessary to set restrictions on button distribution
periods. During distribution times, around 100 buttons
were leaving the booth per hour. This unforeseen outcome
undoubtedly affected the course of the exercise, although
because of the exponential nature of the outbreak, doubling
or even tripling the number of available buttons may have led
to a similar result.
Translating the interactive to real-world epidemiology
The exercise demonstrated the expansive nature of a norovirus outbreak and physical transmission of a pathogen.
The fact that this was not a real disease outbreak and that

Table of Contents for the Digital Edition of Food Protection Trends - July/August 2017