Preliminary Reach of an Information Technology Approach to Support COVID-19 Testing in Schools

To test the effectiveness of TM and HN in reaching communities at risk for COVID-19, we are conducting a 3-armed randomized controlled trial (Scale Up COUNTS). Using a 40:40:20 allocation ratio, parents/guardians of students at participating schools were randomized to receive TM, TM + “resource-conserving” HN (TM + HN), or usual care. The overarching goal of the ongoing trial is to determine whether COVID-19 testing rates are higher among those in the TM or TM + HN group compared with those in the usual care group and whether health navigation improves testing rates among those who do not adhere to testing recommendations. SCALE-UP Counts additionally assesses implementation facilitators and barriers (through surveys and interviews with randomly selected parents/guardians and school staff) and analyzes the impact of sociodemographic variables (eg, race/ethnicity) as potential moderators of intervention effects. The SCALE-UP Counts protocol was approved by the University of Utah institutional review board and registered on ClinicalTrials.gov (#NCT05112900).

To engage stakeholders throughout the research process, we convened a Community and Scientific Advisory Committee (CSAC). The CSAC includes the principal investigators of SCALE-UP Counts (Wu, Wetter), the K–12 testing coordinator from the UDOH, an epidemiology supervisor from the Salt Lake County Health Department along with the emergency/COVID-19 coordinator, the director of the Community Learning Center, and 2 family engagement specialists from the participating school district. In addition to being in frequent contact via e-mail, the CSAC meets quarterly to receive updates on key processes, provide input on challenges, and discuss adaptations of the intervention in response to the current state of the COVID-19 pandemic.

SCALE-UP Counts uses several methods to invite K–12 schools in Utah to participate in the study. In school districts where a research application has been approved, the study team prioritizes connecting with the principals of Title 1 schools (to maximize reach to historically disadvantaged populations). Although Title 1 schools are prioritized, all schools in the approved district are contacted. Initial contact is attempted through an e-mail that provides a brief description about the potential for a COVID-19 testing partnership and a follow-up phone call. The study team also works with the UDOH to connect with schools that have requested free test kits from the health department, thus indicating potential interest in the SCALE-UP Counts project. These schools are invited via the same process previously described. For the schools that agreed to participate, all staff and 1 parent/guardian with a mobile phone capable of receiving texts were sent a welcome message, which includes the option to opt out of further contact. The parent/guardian who was listed as the student’s first priority contact was selected to participate. The intervention was launched in February 2022 and recruitment is ongoing.

Participants were randomized to 1 of 3 groups: usual care, TM, or TM + HN. All participating schools were provided with a supply of at-home rapid COVID-19 tests that was periodically replenished to meet demand. COVID-19 tests were available to all staff and parents/guardians, including those enrolled in the usual care group.

Usual care participants receive access to at-home COVID-19 tests through their school and periodic unidirectional TMs reminding them of test availability. Thus, similar to the TM and TM + HN participants, usual care participants have access to free, self-administered COVID-19 tests, but the messages they receive (approximately once a month) are a simple reminder rather than a bidirectional communication system that includes symptom screening.

SCALE-UP Count’s TM system consists of bidirectional texts that prompt for single-touch responses and support multiple levels of interaction on the basis of parent responses. Approximately once a month, an initial message screens for reported COVID-19 symptoms or exposure to identify whether the student or other household member should get tested. Participants who reply “yes” receive additional messages with a recommendation to undergo testing, including information on how to get tested by either picking up an at-home test at their school or visiting a community testing site. If a participant reports a positive test, they receive additional guidance on what actions to take on the basis of the latest Centers for Disease Control and Prevention recommendations. Those who reply “no” are provided with the option to later report symptoms or exposure. TM targets and content are adapted to changes in the status of the pandemic, testing criteria, and logistics (eg, schools closed for holidays/summer, eliminating the option of test kit pickup at the school). The TM infrastructure used for this study allows us to program and deploy new messages in a matter of minutes to hours depending on the extent of the adaptations.

Participants in the TM + HN arm receive the same text messages as those in the TM group except in the case that they are classified as “nonresponders” to the TM intervention. Specifically, nonresponders are those who report COVID-19 symptoms or exposure but then either do not undergo testing or do not respond to messages about test pickup/results. Once a participant meets these criteria, they qualify for contact by a health navigator and receive a message notifying them of an upcoming call with the opportunity to opt out of the call. Health navigators implement an empirically validated behavior change approach (Motivation And Problem Solving [MAPS]).^15–18  MAPS is a holistic, dynamic approach to behavior change that integrates 2 empirically validated approaches (motivational interviewing and practical problem-solving) with demonstrated efficacy for helping patients engage in target behaviors. MAPS is based on >20 years of extensive feedback on structure and content from low SES, predominantly minority populations, and has been demonstrated to be effective in increasing enrollment and enhancing and maintaining behavior change.^15–18  Using MAPS, health navigators provide motivational enhancement, practical advice, and connections to services, including addressing testing concerns (eg, worries about repercussions of a positive test, infecting family members, quarantining). Figure 1 provides an example of messages in the TM + HN intervention workflow.

For each participating school, descriptive and demographic data describing the student body during the 2021–2022 school year were collected from the National Center for Education Statistics,¹⁹  including the number of students, the percentage of non-white students, the percentage of Hispanic students, and the percentage of students receiving free lunches.

Reach indicators were collected from the PHM system that was used to deploy text messages and monitor responses. Reach indicators included TM Opt-Out (ie, participants opted out of receiving texts before any engagement) and TM Engagement (ie, participants responded to at least 1 text that was not an opt-out). Specific to the TM + HN group, additional indicators were HN Decline (ie, participants requested not to be contacted by a health navigator) and HN Engagement (ie, participant completed at least 1 call with a health navigator).

The SCALE-UP Counts study employs a rapid cycle research approach (Plan–Do–Study–Act framework)²⁰  in which interventions are tested on a small scale using short time frames and iterative evaluation cycles. Using this approach, interventions are continually designed, adapted, implemented, evaluated, and refined. In the Plan stage, the research team conducted meetings with stakeholders, including potential schools in which the intervention would be adopted/implemented, to design the initial interventions. In the Do stage, interventions were implemented within an initial school, allowing the research team to obtain and study descriptive data from interventions to guide refinements. The research team then acted on these data, along with stakeholder feedback and changes to the intervention context (eg, changes to testing recommendations/availability), by making any necessary refinement to the interventions and then disseminating the intervention to additional schools.

Parents/guardians who did not opt out of the study upon receiving the welcome message were randomized to the TM, TM + HN, or usual care group following a 40:40:20 allocation ratio. Randomization was stratified by school and was additionally constrained by family. Specifically, if a parent/guardian had any children who were siblings of another parent/guardian’s children in the intervention, they would be assigned to the same intervention group to reduce the potential for contamination. Randomization was automatically implemented through the flowable platform using randomization tables generated by 1 of the study statisticians (Chipman). Out of necessity, health navigators were not blind to the study group.

The research team was provided with administrative data on each student in a participating school, including siblings in the school and their primary contact’s phone number and e-mail. These data were processed by the research team to identify students who were part of the same family (ie, had siblings in common) and determine whether the listed phone number could receive text messages. Before sending the initial welcome message for the study, each participating school received a supply of at-home test kits with instructions on how to record distribution of the tests. Participating schools were also asked to send a message to all parents/guardians to inform them of participation in the SCALE-UP Counts program. Participants were grouped into cohorts (usually a single school) and typically received multiple iterations of their randomized intervention throughout the study period. The first intervention cycles began in February 2022; this article reports on intervention cycles that occurred through May 2022.

Descriptive statistics on demographics, reach, and engagement metrics were calculated both overall and, where possible, separately among the 3 study groups.

To date, 6 K–12 schools from a single, large school district are participating in the intervention. These schools serve a total of 4881 students and include 5 elementary schools and 1 high school. Consistent with the priority of the SCALE-UP Counts program to reach historically disadvantaged communities, over a third of the students in these schools (36.2%) are eligible for free or reduced lunches. These schools enrolled 32.9% non-white students and 15.4% Hispanic students.

As of May 2022, a total of 3506 potentially eligible parents/guardians were identified from participating schools. As shown in Fig 2, of these potential participants, 44 were excluded because they did not have a phone that could receive text messages and 132 opted out of the program before randomization (another 1007 had not yet been entered into a workflow because of a technical glitch). Among the 2323 parents/guardians included in at least 1 intervention workflow, 79.6% (n = 1849) were randomized to 1 of the 2 groups providing TM and 20.4% (n = 474) were randomized to usual care. At the time of these analyses, all enrolled participants had received at least 1 (and up to 4) intervention text messages that screened for COVID-19 symptoms or exposure. Participants received this message an average of 1.33 times, with most (67.7%; n = 1572) receiving it once. Of those randomized to a group that received text messages, 19.3% (n = 354) engaged with at least 1 message. Within the TM + HN group (40.1%, n = 932), 1.3% (n = 12) qualified for health navigation, of which 1 person declined to speak with a health navigator; 50% (n = 6) did not respond to contact attempts by a health navigator, and 41.7% (n = 5) talked to a health navigator.

In the initial 4 months of the SCALE-UP Counts project implementation, >2000 parents/guardians were reached by at least 1 program message. Less than 2% of parents/guardians messaged did not have a mobile phone that could receive text messages, only 3.8% opted out of the study, and almost 20% engaged with at least 1 text. These reach data support the feasibility of using TM as a PHM strategy to promote COVID-19 testing among historically disadvantaged communities.

To determine the potential of evidence-based interventions to impact public health, programs must be able to reach at-risk populations.^21,22  Given the disproportionate impact of the COVID-19 pandemic on low SES and racial/ethnic minority populations, it is especially important to ensure that resources and interventions reach these groups.²³  To overcome barriers to COVID-19 testing (eg, low income)²⁴  and provide more equitable access, the SCALE-UP Counts program prioritizes participation among Title 1 schools, which enroll a large proportion of members from historically disadvantaged groups. Our community-engaged approach, including the formation and active engagement of the CSAC, has also been critical for proactively identifying and addressing barriers to engagement. For instance, in response to their input, we developed a Spanish language version of the intervention and methods to rapidly identify those who would prefer intervention content and assessments in Spanish.

Rates of engagement (19.1%) for those receiving TM may seem low compared with other text-based public health programs, such as the Centers for Disease Control and Prevention v-safe vaccine safety monitoring program, which uses an opt-in model of recruitment (ie, potential participants chose to enroll), likely resulting in a more motivated study sample.²⁵  In contrast, SCALE-UP Counts used an opt-out model of recruitment where participants were automatically enrolled unless they actively chose to opt out of participating. Thus, participants may not have felt it necessary to respond to messages, particularly when they had no COVID-19 symptoms or exposure. Perhaps even more importantly, opt-out recruitment dramatically reduces the selection bias that accompanies the active consent/enrollment process, yielding a far more representative sample that approaches the entire eligible population. That is, only 3.8% of participants opted out of enrollment in SCALE-UP Counts.

Several additional factors should be considered when evaluating these rates of engagement. First, parents/guardians did not need to provide a response to a text message or talk with a health navigator to pick up a COVID-19 test at their school or to be tested at a community testing site. TM and health navigation may have served as cues to action,²⁶  even among participants who did not engage in these interventions directly. Importantly, the analysis plan for the overall project will allow us to evaluate whether testing rates differed between the TM and usual care groups, regardless of TM and HN engagement rates. Second, the extent of parent/guardian engagement may have been influenced by the current phase of the COVID-19 pandemic. When the intervention launched in February 2022, the rates of COVID-19 were declining from a peak.²⁷  To promote engagement, we continue to implement multiple strategies, such as enhanced communication about the program from participating schools (eg, announcements in parent–teacher association meetings) and social media postings.

Because the SCALE-UP Counts project is an ongoing study, the impact of the interventions on COVID-19 testing and vaccinations is not yet known. Therefore, we do not yet know the extent to which reach and engagement affected the primary study outcomes or whether the interventions contributed to reductions in COVID-19 health disparities. We also do not know how reach and engagement varied by demographic characteristics because the schools did not share individual-level demographic data with the research team. Further, our ability to remain in contact with parents/guardians relies upon the accuracy of the contact information received from participating schools. As the study progresses, changes to student enrollment and contact information could impact reach; we are proactively creating systems to update intervention contacts in response to these types of changes, and we are coordinating with school leadership to ensure timely communication of student directory updates. Lastly, the reach and engagement outcomes reported here should be regarded as preliminary; as the study progresses, we anticipate being able to evaluate engagement as a continuous variable that changes over time. Our monthly tracking of study outcome variables and continual monitoring of state and local testing/infection rates will allow us to evaluate how the evolving state of the pandemic impacts reach and engagement.

In the initial months of the SCALE-UP Counts project, intervention messages have reached >2000 parents/guardians. Our reach data indicate that TM and telephone-delivered health navigation are feasible ways to reach a diverse population of parents/guardians of K–12 students to provide health screening messages and offer access to at-home COVID-19 testing. Applying a rapid-cycle Plan–Do–Study–Act method, along with a community-engaged approach, the SCALE-UP Counts program will continue to be adapted to both respond to changes in the state of the pandemic and promote continued engagement.

The study team thanks our research staff and health navigators, including Malynne Cottam and Sarah DeSantis. We also thank the ongoing contributions of the Granite School District leadership that enabled the enrollment of the schools currently included in the SCALE-UP Counts project and have provided continuing input and support for intervention implementation. Brooke Walker, MS, Duke Clinical Research Institute, provided editorial review and submission. Ms Walker did not receive compensation for her contributions, apart from her employment at the institution where this study was conducted.

COVID-19

coronavirus disease 2019

CSAC

Community and Scientific Advisory Committee

HN

health navigation

K–12

kindergarten through 12th grade

MAPS

Motivation and Problem-Solving

PHM

population health management

SES

socioeconomic status

TM

text messaging

TM + HN

text messaging + health navigation

UDOH

Utah Department of Health