Protecting Our Community: COVID-19 Testing

Overview

The COVID-19 pandemic has disproportionately affected American Indian (AI) and Latino communities, and these groups also have increased risk of poor prognosis due to high rates of chronic disease such as diabetes, cardiovascular disease, and cancer. In the northwestern United States, AI and Latino communities already face significant disparities in health care access, which have been further exacerbated by the COVID-19 pandemic. In the proposed study, Protecting Our Community: A Pragmatic Randomized Trial of Home-Based COVID Testing with Native American and Latino Communities, the investigators will leverage our long-term community-based participatory research partnerships to test the hypothesis that home-based testing will be feasible, impactful, and better-accepted using active delivery of test kits by trusted community health educators in two vulnerable, high-risk rural communities. Our two long-term partner communities are the Flathead Indian Reservation of the Confederated Salish and Kootenai Tribes in Montana, and the Yakima Valley of Washington, a large Latino community. The investigators will determine the cultural, social, behavioral, and economic barriers to home-based SARS-CoV-2 testing; culturally adapt and enhance home-testing educational materials and create home-testing instructional graphics and YouTube videos; conduct a 2-arm pragmatic randomized trial of active (delivered by community health educator) vs. passive (without community health educator) home-based testing kits (n = 200/community) for testing completion; and create a model for community-driven testing protocols that can have significant impact for increasing home-based testing uptake among AI and Latino communities nationally. This work will enable underserved AI and Latino communities to take full advantage of the coming wave of rapid point-of-care home tests and decrease the significant impact of COVID-19 in their communities.

This clinical trial utilizes community-based participatory research partnerships to test the hypothesis that home-based testing for COVID-19 will be feasible, impactful, and better accepted using active assistance with the testing kits by trusted community members vs. passive delivery in two vulnerable high-risk rural communities. Our two long-term partner communities are the Flathead Indian Reservation of the Confederated Salish and Kootenai Tribes in Montana and the Yakima Valley of Washington, a large Hispanic community. The investigators will implement a two-arm pragmatic randomized trial during February to June 2021. Two hundred (200) adults per community will be randomized and stratified by age (18-25, 26-59, and 60+) to active vs. passive delivery. Our primary outcome will be the proportion of subjects who return testing collection kits in the active vs. passive arm. The trial will enable us to determine best practices and issues with home-based testing delivery in American Indian and Latino communities. Participants randomized to both the active and passive arms will receive home testing kits augmented with materials developed by the study in both English and Spanish. In order to control for potential barriers to home testing kit delivery in this population (including internet access, digital literacy, English language proficiency) and limited access to mail delivery services (UPS, FedEx, etc.), participants in both study arms will be provided kit delivery and home collected sample return support coordinated by study staff. Coordination support options include (1) pick up / drop off at the research site, (2) shipping the home testing kit to the participant's mailing address by USPS or (3) home delivery/pick up to accommodate those with mobility concerns. The home testing kit selected in this study is the Everlywell COVID home-sampling kit. Participants randomized to the passive arm will receive a home test kit with instructions on how to self-register their kit online, and will be directed to contact Everywell for assistance, if needed. Participants randomized to the active arm will have their home collection kit registered on their behalf by community health workers while the participant remains on the phone. Study developed materials will direct participants to contact the study team for assistance. Participants in both arms will receive study developed instructions to access their test results from the Everlywell online portal. Positive test results will be verbally provided by Everlywell contracted physicians. Study staff will confirm all participants have retrieved their test results from the online portal and referred to appropriate public health or clinical support as needed. An initial survey of participant demographics and any current symptoms and medical conditions (including the RADx-UP Coordination and Data Collection Center common data \& evaluation metrics) will be done via phone with participants prior to kit delivery. Home testing kits returned and tested within 14 days of randomization will be considered "complete", while testing kits not returned or returned later than 14 days will be considered "not-complete". Within 14 days of randomization, will will conduct a phone delivered post-testing feedback survey with participants who completed their home testing kit to ask opinions on testing barriers and supports, interest in and barriers to self-testing before and after the study and usefulness of training materials. Participants who did not complete the home testing kit will be asked to provide reasons and barriers to completion. Post-testing feedback surveys will be brief and participants will receive a $35 gift card incentive for survey completion. The investigators will also conduct in-depth, semi-structured, phone interview with 40 participants, 20 at each site, to collect perceptions of and experience with the home testing kit, vaccinations and other issues related to the pandemic. To facilitate collection of data at multiple timepoints in the study, interviews will be staggered. Interviews will occur roughly every two weeks over a 10-week period. The investigators will attempt to match interviews by active and passive arm assignment, completer and non-completer status, and timeframe. Twenty interviews with participants randomized to the active arm (10 participants who completed their home testing kits and 10 who did not) will be conducted as well as 20 interviews with participants randomized to the passive arm, also split evenly between participants who completed and not-completed their home testing kit. Participants who complete the in-depth, semi-structured phone interview will receive an incentive equal to $25. Study consent, enrollment and study assessments will be conducted in the participant's preference of Spanish or English. Harms The investigators expect an extremely low occurrence of unfavorable events in this study and will be classifying negative study related harms to human subjects as either "social impacts", which for the purposes of this study are defined as an inadvertent and negative impact on an individual's standing in the community, within their family, their job, with the government, or with their finances as a result of participating in the study. Social impacts may also include worry (warranted or unwarranted), feeling upset or depressed, embarrassed, shameful, or guilty as a result of study participation or alternatively adverse events (AE) which are defined as any untoward medical occurrence in a subject during study participation. Identification of both social impacts and AE will primarily be through passive surveillance, such as survey responses and conversations with site staff, and will be documented following informed consent until the final study assessment is complete. All social impacts and adverse events will be reported to the Data Safety Monitoring Board. Adverse events will be reported to the Montana State University IRB within 24 hours of site awareness. Study staff at each site will be trained in techniques to help resolve distress and connect participants to appropriate resources Data Analysis and Sample Size Considerations Our prior work with passive strategies has produced a return rate of approximately 70%.The investigators view a relative increase of 20% (absolute increase of 14%) to be a meaningful impact associated with the effort involved in active outreach, and therefore the investigators selected a sample with adequate power to detect such a difference. Specifically, with 400 total subjects (stratified by community) the study will have 90% power to detect an effect size associated with a rate ratio of 1.20, or an absolute difference of 14% (70% in passive, 84% in active). The investigators evaluated the design for robustness to assumptions regarding the passive return rate, and they have \> 80% power to detect an absolute difference of 14% (relative rate of 1.23) if the passive return rate is 60%. No correction will be made for missing outcomes since the outcome is negative if no kit is returned. In addition, exploring the impact of using regression adjustment (logistic regression with an identity link) to account for community stratification, and the primary analysis maintains \> 80% power for a range of passive rate differences across communities (ranging from 5% to 10%). All power calculations used the R statistical package and the PWR library and simulation methods. The investigators will evaluate the net improvement in the rate of returning kits (primary outcome), comparing active vs. passive community outreach for uptake of home-based testing. Within each community, individuals will be randomized to active or passive delivery, and our primary analysis will compare the difference in the proportion of subjects who return collection kits while adjusting for community. Secondary analysis will compare active and passive rates across the communities to evaluate consistency of the intervention effect. Quantitative analysis of survey responses will focus on the creation of summary statistics that characterize the response to key survey items. An important statistical consideration is survey non-response and the potential for non-representative samples. The investigators will evaluate baseline factors that are potentially associated with non-response and then use inverse probability weighting (IPW) to correct for any potential bias associated with measured factors. All statistical tests will be carried out at the 0.05 significance level, and estimates, 95% confidence intervals, and associated p-values will be reported for all tests. Analyses will be conducted using SAS 9.4 statistical software.