SCALE-UP Utah II: Community-Academic Partnership to Address COVID-19 Conversational Agent Study

Racial/ethnic minority, low socioeconomic status (SES), and rural populations suffer profound health inequities across a wide variety of diseases and conditions, including COVID-19. For example, as of June 2021, the cumulative COVID case rate in Utah per 100,000 was 10,803 among Whites vs. 17,541 among Latinos. The positivity rate was 14% among Whites vs. 24% among Latinos. Similar disparities persist across the nation for vaccination rates between urban vs. rural, high vs. low SES, and White vs. non-White populations. Low vaccination rates leave underserved populations at risk for local outbreaks, and more contagious and severe variants. Thus, interventions targeting these populations at the interplay between testing and vaccination among underserved populations are critical for pandemic control.

Not only do underserved populations experience profound health inequities, but there is also a critical digital divide between high and low resource healthcare systems. Low resource settings are far less likely to adopt Health Information Technology approaches, and often do not have the capacity to implement large scale population health management (PHM) efforts utilizing data analytics and automated patient outreach. As such, research is needed utilizing targeted PHM approaches that proactively identify, reach, and navigate vulnerable patients to both increase opportunities to engage in vaccination and testing, and to address barriers to engagement. Community Health Centers (CHCs) are optimal settings for implementation of PHM interventions to increase the uptake of COVID-19 testing and vaccination among underserved populations. Eleven Utah CHC systems are participating in SCALE-UP II. Their 38 primary care clinics serve over 112,000 unique patients annually (36% Latino, 10% Native American, 63% <100% poverty level, 57% uninsured, and 42% of clinics are in rural/frontier areas).

SCALE UP II is comprised of two distinct studies, the Text Message (TM) study and the Conversational Agent (CA) study. Patients will be triaged into one of two studies based on self-reported ownership of a smart phone with internet access. Patients who report not owning a smart phone with internet access will be included in the TM study. Additionally, patients who do not respond to the question regarding smart phone ownership will be included in the TM study. Patients who self-report ownership of a smart phone with internet access will be included in the CA study.

SCALE-UP II: CA study will implement and evaluate practical, accessible, and scalable PHM interventions to increase COVID-19 testing and vaccine uptake based on the best evidence available, patients' specific barriers and hesitancy factors, and extensive collaboration with CHCs, AUCH, and UDHHS. This study is a 2x2 design with patients being initially randomized between receiving either text messages or the link to a conversational agent as well as into two distinct types of available patient navigation, Request-PN or No PN.

Text Messaging (TM): bidirectional text messaging to connect patients to vaccination or mailed at-home rapid test kits for use, as needed.

Conversational Agent (CA): automated, scripted and interactive agent used to mimic human interaction to: 1) elicit specific hesitancy factors and barriers to testing; 2) provide tailored information to address each individual's hesitancy factors and barriers to testing; and 3) offer access to at-home rapid test kits.

Patient Navigation (PN): phone call from a community health worker to help address hesitancy and barriers, and to offer at-home rapid test kits. This study will examine two distinct forms of Patient Navigation: Request PN and No PN. Each patient will be randomized to receive either Request PN or No PN. Request PN allows patients to request patient navigation by responding PERSON to a text message/conversational agent. Patients who are randomized to receive No PN will not be provided the opportunity to speak with a patient navigator.

The primary outcome, Reach-Accept Testing, captures whether patients requests an at-home test kit and confirmed their mailing address. Secondary outcomes include: Testing (i.e. the proportion of participants who self-report having used an at-home test at least once during follow-up) as well as several implementation outcomes including Reach-Engage Testing (proportion of patients that reply to an offer to receive an at-home rapid test kit), PN-Request (proportion in PN who requested patient navigation), PN-Engage (proportion in PB who talked to a patient navigator), and Opt-Out. The outcomes and compared groups provided for the Smartphone study are main effects for delivery mechanism (Chatbot \& TM), main effects for PN (PN \& No PN), and main effects for message frequency (10-day \& 30-day).

SCALE-UP II will include a Consortium Data Reporting Unit (CDRU) consisting of a Data Manager and one member the project's biomedical informatics team. The unit will attend regular meetings and dissemination activities organized by the CDCC. The CDRU will seek guidance from the CDCC with regard to data acquisition and consent for data sharing. As required by the NIH, SCALE-UP II will collect RADx-UP Tier 1 Common Data Elements for study participants who receive an at-home COVID test through the project. These data will be collected through surveys administered one month after the participant receives their at-home test. Data will be standardized according to the data dictionary provided by the CDCC. Our CDRU will work closely with the CDCC to establish a protocol for frequency, format, and exchange of data. SCALE-UP II will share identifiable data with the CDCC and NIH for the Data Hub as well as future research. Participants who complete the survey data will first complete an informed consent process. The informed consent will be administered to the patient at the time of survey collection.

Lighthouse Research and Development will conduct phone surveys to assess patient reported use of COVID-19 at-home testing (~2,300 participants) among patients who received test kits and do no respond to the online survey request. Lighthouse will send notifications (e.g., postcards, text messages, voice messages, etc.) to participants to alert them of the opportunity to complete the survey online or over the phone. Interviewers will complete up to 15 call attempts across weekday, evening, and weekend calling shifts over a one-month period to each participant. Patients will be compensated with a gift card for completing the survey.