Use of a Digital Assistant to Report COVID-19 Rapid Antigen Self-test Results to Health Departments in 6 US Communities

This cohort study investigates the proportion of beneficiaries of a COVID-19 home test kit program who used a digital assistant to report results.


Introduction
Rapid antigen home tests for COVID-19 are an important part of the federal government strategy to expand COVID-19 testing access and availability throughout the United States. 1 However, the distribution and scale-up of rapid home tests for COVID-19 have been inconsistently accompanied by standard public health reporting mechanisms, challenging the ability to monitor rates of  testing. It is important to understand more about individual reporting choices to create an optimal system for self-testing and surveillance. This study characterized how often individuals in 6 communities logged their home test results through a digital platform and patterns of result reporting state departments of health (DoH).

This cohort study received nonresearch determination by the University of Massachusetts Chan
Medical School Institutional Review Board and so was determined to be exempt from review and informed consent. The study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

Test Kit Program Intervention Communities and Procedures
The Say Yes! Covid Test program, a partnership between the National Institutes of Health and the Centers for Disease Control and Prevention, distributed more than 3 000 000 self-tests to 6 communities across the United States from April to October 2021. 2,3 More details about the intervention can be found elsewhere. 4

Data Collection
An optional online platform and accompanying application, developed by CareEvolution, was launched with the test kit intervention as a platform for DTC orders, logging test results, and reporting results to the state DoH (Figure 1). Digital tool features were freely available and stored without personally identifiable information. Log and reporting features were available indefinitely in each site starting at the beginning of each respective distribution period. The log feature allowed households to document their test dates and results for their records. Households were also given the option to report each logged test to the state DoH through the digital assistant. For logged tests, test date, result (positive, negative, or invalid), and reporting decision (report or no report) were included in a data feed accessible to CareEvolution. For this analysis, reported tests included those reported with personally identifiable information or anonymously. A $25 gift card incentive was offered to participants in Indiana and Kentucky if they reported at least 1 test result per household to their state DoH through the digital assistant. The incentive was also offered in Georgia and Hawaii starting on October 4, and these locations were termed partially incentivized sites. No incentive was offered in Tennessee or Michigan for reporting test results to the state DoH. Beneficiaries were able to report tests at any point during and after the distribution period. Tests logged in the digital assistant from April 1, 2021, to January 12, 2022, were included in the analyses. Residents of Tennessee were unable to report tests to the DoH until June 24, 2021, so data before this point were excluded from reporting analyses for Tennessee.  Mean is over the distribution period.

Statistical Analysis
Total numbers of DTC orders and digital assistant users were calculated by community and incentivization status (full incentive, partial incentive, and no incentive). For partially incentivized sites, the percentage of reported results was analyzed before and after the onset of incentivization.

Distinct Users for Logging Test Results
Of 313  Kentucky suggests that application-based reporting systems may be associated with an improved reporting process when paired with incentives. However, the challenge remains in drawing people to use the digital assistant, as suggested by the low uptake of the digital assistant for testing purposes.
Symptom-based participatory surveillance through digital applications has been used successfully for monitoring influenza-like illness, among other infectious diseases, and rapid testing offers great opportunity to build on these technologies to rapidly ascertain changes in community prevalence of infection. 7,8 Other means of improving uptake of the digital assistant or other reporting mechanisms should be explored further to maximize the value of these interventions.

Limitations
This study offers a unique look into COVID-19 test reporting behaviors of nearly 15 000 digital assistant users throughout the United States. However, there are limitations to this data. The number of digital assistant users was small compared with all intervention participants, and with the current data, we were unable to assess demographics or socioeconomic status of digital assistant users or how digital assistant users compared with nonusers. Additionally, DoH reporting using the digital assistant was available to individuals who received their testes from community sites, in addition to those who used DTC ordering, and we were unable to assess whether the test distribution modality was associated with uptake of the digital assistant for logging and reporting test results. Additionally, the incidence of COVID-19 over the distribution period differed by community, with Hawaii having nearly 4-fold higher incidence of COVID-19 during the distribution period than Michigan. However, rates of COVID-19 may have changed drastically daily, weekly, or monthly, which was not reflected in these point estimates, and all sites were permitted to log and report rapid antigen tests through January 12, 2021, rather than solely during the distribution period. Further investigation is warranted to examine the association of community transmission with reporting behaviors.