Colombo et al. 10.3389/fpubh.2022.942795
leveraged to deliver complementary information about COVID-
19 to a wide audience.
Text messages (SMS)
Three to five SMS per week were determined to be ide al to
avoid undue burden on enrollees (
10). One weekly SMS, sent
to all users, served a reminder to complete the brief weekly
survey. This SMS contained a specific-participant link that was
associated with the individual user’s prior longitudinal data.
Additional informative SMS were released with the goal of
disseminating educational material to users. To receive this
message type, subscribers opted-in based on their preferences
during the AZCOVIDTXT enrollment process. Content for
informational SMS was intended to mirror and augment
website article material, and address challenges and topics of
particular interest to Arizonans. One informative SMS sent to
all subscribers was intended to spark interest and encourage
subscribers to visit the weekly updated website articles. Messages
that directed subscribers to view the website articles included a
URL as part of the character count limit. A second informative
SMS s ent to all subscribers provided an opportunity to send
more general information, such as current events and changes
in official guidance, local Arizona supportive services, and social
support reminders. All SMS were translated into Spanish by
staff at the University of Arizona Cancer Center’s Behavioral
Measurements and Shared Resource (BMISR) and were released
to subscribers who indicated a Spanish preference for receiving
their informative SMS.
Data provided by the surveillance component of
AZCOVIDTXT were reviewed to allow responsive messaging
back to the users. This design was in accordance with rese arch
demonstrating higher efficacy for text messaging-based health
promotion interventions when applying personalized messaging
(
11). Thus, these general SMS were often tailored, especially
during the early months of the pandemic, to address the
challenges that enrollees were reporting on their weekly
surveillance surveys (e.g., accessing cleaning and hygiene
supplies, medical and mental health care, and basic human
services such as housing, food, and water). A third informative
SMS was sent only to a specific group of subscribers: those
who reported that someone in their household that week was
sick. This gr oup of messages included information specifically
tailored to minimizing spread within households, how and
when to isolate and quarantine, and when and how to seek
urgent care.
AZCOVIDTXT offers real-time response. In the event
of a critical or emergent issue, additional, one-time SMS
could be sent. Examples of critical or emergent issues of
public health and safety, included wildfire evacuations during
lockdown, executive orders such as the Stay-at-Home order,
and to address widespread and life-threatening misinformation.
Urgent messages were only sent on six instances during
2020 (
Table 1).
SMS were sent via the AZCOVIDTXT communications
platform: a Django-powered application developed by
programmers and data scientists at the University of Arizona’s
Data Science Instit ute. The AZCOVIDTXT platform interfaced
with Twilio, a cloud-based communic ation service, in order to
send and receive SMS with subscribers (
12). The AZCOVIDTXT
communications platform enabled advanced scheduling of SMS
for a spe cific date and time and to unique subscriber groups,
such as Spanish SMS sent to Spanish-speaking enrollees. Further
filtering was possible (e.g., by zip code or specific flags such as
subscribers who reported a sick household member on a weekly
survey), allowing for messages to be tailored as needed, (e.g.,
sending messages regarding wildfire evacuations to only the
zip codes affected). The AZCOVIDTXT SMS communication
platform also logged and tracked all incoming and outgoing
SMS for reporting purposes. SMS released through this platform
had a 160-character limit.
Website articles
An average of t hree articles regarding pressing COVID-
19 related information were developed and published on the
system’s website each week. The articles were complementary
to the service’s other communication features and served to
provide comprehensive COVID-19 health information. The
intent was to clarify current science and policies and provide
tips and information for resilience during the pandemic.
Specific topics included in the articles discussed an array
of pandemic-related information including evolving COVID-
19 science, state-issued executive orders, available Arizona
resources, and interviews with field experts (e.g., university
researchers, healthcare professionals, etc.). Articles typically
ranged between 300 and 600 words and were written in a manner
that could be easily interpreted by the lay public.
Decisions about the content included in each article
were informed through multiple channels including (1)
surveillance data input from users, (2) community feedback
forms administered periodically, (3) feedback forum,
email, and incoming SMS responses, and (4) as guided by
AZCOVIDTXT team leads who had extensive backgrounds
in health communication, public health, nursing, and health
education prioritizing and ranking important emerging
topics. The articles were mainly researched and written by
AZCOVIDTXT st aff who have expertise in public health and
health education. Periodically, external health professionals
and experts were consulted for additional content matter.
After the material was drafted, any health-related content was
carefully vetted by public health and nursing professionals as
well as a health communication expert who refined language
Frontiers in Public Health 03 frontiersin.org