Background and Rationale

While there is much we do not know about the differential effects of COVID-19, early data
shows that minority communities are disproportionally effected by the virus. There are many
factors that may increase the likelihood of contracting COVID-19 including: 1) community
spread because of lack of access to testing in low-income communities, 2) work environments
that may increase exposure, 3) more densely populated housing that reduces the ability to
social distance, 4) limited understanding of preventive measures due to literacy and language
barriers, 5) constrained financial resources to stay home and not work, and/or 6) lack of
trust in the health care system. The Marshallese are a Pacific Islander population
experiencing significant health disparities with some of the highest documented rates of type
2 diabetes mellitus (T2DM) of any population in the world. Estimated T2DM rates among
Marshallese in the U.S. range from 25%-50%, much higher than the general U.S. population.
People with T2DM are more likely to experience severe symptoms and complications when
infected with COVID-19; however, those that manage their T2DM well are less likely to become
extremely ill from the virus. In order to reduce the disparities caused by COVID-19 and
ultimately compare the effectiveness of prevention interventions among the low-income
minority communities most effected by COVID-19, it is critical to understand minority
populations': 1) risk exposure, 2) knowledge of preventive recommendations, 3) barriers and
facilitators to implementing preventative recommendations, 4) barriers and facilitators to
COVID-19 testing when appropriate, and self-care behaviors during COVID-19.

Specific Aims

Aim 1: Document COVID-19 risk exposure for Marshallese community members.

Aim 2: Document Marshallese community members' knowledge of preventive recommendations.

Aim 3: Document barriers and facilitators to implementing preventative recommendations.

Aim 4: Document Marshallese community members' barriers and facilitators to COVID-19 testing
when appropriate.

Aim 5: Document self-care behaviors during COVID-19.

Aim 6: Explore COVID-19 pandemic related barriers and facilitators to T2DM self-management
among Marshallese adults with T2DM.

Aim 7: Assess the effect of the COVID-19 pandemic on T2DM self-management activities among
Marshallese adults with T2DM.

Study Design and Procedures

Participants will complete an online or telephone survey. Up to 100 participants will also be
invited to complete an in-depth qualitative interview in addition to the survey (these
results are not presented here). Community-based recruitment for the online survey will be
conducted by posting an informational flyer on social media. The flyer will contain a link to
the study information sheet and survey documents. Study team members will also reach out to
study participants via phone or electronic mediums (e.g. email, text, or messenger). Study
staff will send potential participants a link to an electronic survey using REDcap. The
REDcap system will include electronic documentation of consent prior to completing the
survey. Access to study data will be limited to only those personnel who need it to complete
relevant job duties. All data, regardless of whether it is identifiable or not, will be
stored in a locked file cabinet in a locked room, or on a secure University of Arkansas for
Medical Sciences (UAMS) server that requires two-factor authentication.Each participant who
completes the survey will receive a $20 Walmart gift card as remuneration.

Measures/outcomes

Our survey will be based on CDC risk assessments and utilize the NIH funded COVID-19 items
that are part of the PhenX Toolkit, as well as other standardized survey items/scales. Those
participants with T2DM will be asked to complete questions pertaining to T2DM self-management
during COVID-19.

Data Analysis

Given the descriptive nature of the study aims, the analytic strategy will focus on
presenting results of item-level descriptive analyses, with an emphasis on frequencies and
proportions. There will be no attempt to impute missing responses for any items. For each
analysis, the number of included responses will be reported.