Abstract Body

Antigen-detecting rapid diagnostic tests (Ag-RDT) for SARS-CoV-2 are an inexpensive diagnostic tool with fast turnaround times. Ag-RDTs in combination with measures to reduce contact rates after a positive test result can reduce the spread of SARS-CoV-2. Understanding when and in what settings Ag-RDTs can best be utilized to reduce transmission is critical for resource allocation. Here, we used a suite of mathematical models to quantify the impact of SARS-CoV-2 Ag-RDT testing strategies on COVID-19 outcomes in a variety of use-cases.

Our analysis synthesized the results from eight mathematical models from different modeling groups to assess the potential impact of Ag-RDT testing for SARS-CoV-2 infection across multiple use cases: (a) community testing, (b) mass gatherings, (c) K-12 schools (kindergarten to 12th grade/high school, or primary/secondary education), (d) universities, (e) border crossings, and (f) testing to exit quarantine. We calculated two outcomes relative to the status quo in each use case: (1) impact: the percent and number of infections averted and (2) efficiency: the number of tests required to avert one infection. We investigated the impact of different epidemic conditions including effective reproductive number (Rt) and COVID-19 prevalence, and the frequency of testing (for community testing, K-12 schools, and universities).

Different use cases require varying testing strategies to reduce infections most efficiently and effectively across a range of epidemic conditions, with some global trends. Overall, there were tradeoffs with impact and efficiency. Across use cases, increasing test frequency (and/or more testing) was associated with greater percentage of infections averted. However, lower testing frequency was generally more efficient. In the community testing and university use cases, testing was most effective and efficient when Rt and/or infection prevalence was low but for border crossings testing was most effective and efficient when Rt and/or infection prevalence were high (Table 1).

The optimal timing of the intervention depends on whether one is trying to maximize effectiveness or efficiency, and on the use case itself. For a robust understanding of total community-level impact and cost-effectiveness, future work should aim to assess the combined impact of interventions through a single model that can consider all use-cases.