Food Protection Trends - March/April 2021 - 233

consumers to understand their role to properly (i) cook
raw poultry products to 165°F, (ii) clean to avoid crosscontamination, and (iii) chill leftovers promptly. Efforts to
encourage consumers to use thermometers to properly cook
poultry have been ongoing for almost two decades (15).
In addition, safe handling instructions have been required
on poultry packaging since 1994 (30), and although these
education methods have shown effectiveness, it has been
found that they are not sufficient to ensure proper handling
and cooking by all consumers (3, 17). Over the same
time that these two education efforts have been in place,
Salmonella and Campylobacter have remained the two leading
bacterial causes of foodborne illness (23). A major challenge
is that it is not clear what percentage of illnesses caused by
poultry and these pathogens is the result of undercooking
and what percentage is the result of cross-contamination.
Although it has been suggested that cross-contamination by
poultry is more problematic than undercooking (16, 18),
consumer observation research found that when consumers
cooked chicken breasts to what they believed was done, they
were not yet at the proper temperature of 165°F (3).
Recent and ongoing research and education efforts have
sought to better understand and address knowledge and
behavior gaps regarding handling of raw chicken (8, 10, 14).
This research has identified unique practices and conditions
that might result in increased risk for illness from poultry.
These include the common practice of washing raw chicken
(6, 10, 11, 12, 13), the presence of pathogens on the outside
of raw poultry packaging (2), and potential exposure
of young children to raw meat and poultry products in
shopping carts (21). Education efforts have been developed
and piloted to show efficacy around some of these practices
(7, 8, 20). However, most of these efforts have focused on
chicken and chicken parts.
Dewaal et al. found that from 1990 to 2003, chicken was
responsible for 179 outbreaks and 3,363 cases of illness,
whereas turkey was responsible for 88 outbreaks and
5,146 cases of illness (5). Similarly, in poultry-associated
outbreaks from 1998 to 2012, whereas the overall number
of outbreaks because of chicken was greater, the mean
number of illnesses associated with turkey-associated outbreaks was 45, compared with 31 illnesses associated with
chicken-associated outbreaks (4). Of the 52 turkey-associated outbreaks, 42% occurred during November and
December and 55% of those were caused by C. perfringens,
compared with only 13% of turkey-associated outbreaks
being associated with C. perfringens during other months
(4). Outbreaks of C. perfringens associated with turkey
around institutional Thanksgiving meals have also been
identified (19, 22). This would indicate that there is evidence of increased foodborne illness around the months
when most whole raw turkeys are prepared, both from
mishandling, because C. perfringens is associated with
improper holding temperatures, and potentially from

Salmonella or Campylobacter, because of cross-contamination or undercooking.
Formative research in our laboratory identified cooking
a whole turkey overnight to be a practice among a subset of
consumers (9, 10). Although cooking a turkey overnight may
result in a turkey that has reached an appropriate internal
temperature, unless the turkey is consumed early in the
morning, or thoroughly refrigerated and reheated before
serving, there is a great potential for improper handling
associated with this practice.
The purpose of this study was to better understand among
a cross section of U.S. consumers how they handle a whole
turkey with respect to (i) washing raw turkey, (ii) stuffing a
whole turkey before cooking, (iii) cooking a turkey overnight,
and (iv) holding a turkey after cooking and before serving.
MATERIALS AND METHODS
This study was approved by Drexel University's
Institutional Review Board.
Survey development
The final survey consisted of 10 questions. Survey format
and most questions were adapted from a previously used
survey (10), with the addition of a question about stuffing
a turkey before cooking.
All survey questions were entered into the SurveyMonkey ™
platform. In this platform, each question was given a response
type (multiple choice or fill in the blank). In addition to
the response type, each question was coded to not allow
participants to skip any of the questions. The completed
survey was tested multiple times for errors and accuracy
before it was launched and shared with SurveyMonkey's™
panel of participants.
Recruitment and survey administration
Potential subjects were recruited using SurveyMonkey's™
contribute panel. This panel is made up of millions of people
from diverse backgrounds in the United States. The survey
was launched at the end of December 2017 immediately
following the holiday season. Participants from the contribute panel could select to take the survey from either their
SurveyMonkey™ dashboard or from an email or text message
sent by SurveyMonkey™ inviting them to complete the survey, depending upon the notification settings they selected
when they completed their SurveyMonkey™ profile. Responses were populated in SurveyMonkey™. Once participants
completed the survey, SurveyMonkey ™ automatically donated
money to the charity of their choice.
Data analysis
Data was analyzed using Microsoft Excel 2016. Analysis
of variance was used to determine statistical significance
(P < 0.05) between consumers of different demographics.

March/April Food Protection Trends

233

Food Protection Trends - March/April 2021

Table of Contents for the Digital Edition of Food Protection Trends - March/April 2021