Cost-effectiveness of sars-cov-2 rapid antigen testing in lowresource settings

Background: The World Health Organization (WHO) has called for increased testing to help arrest the transmission of coronavirus disease 19 (COVID-19). Molecular testing (PCR) is the recommended method for the diagnosis of COVID-19. In low-resource settings (LRS) however, the availability and public health impact of these tests is constrained by availability of testing capacity, shortages of reagents/supplies, lack of skilled personnel, long turnaround times (TAT), and costs. Despite lower sensitivity, antigen detection rapid diagnostic tests (AgRDTs) could provide improved access at lower costs and quicker TAT. We evaluated the optimal use of AgRDTs to increase testing access within TAT and reduce the cost and the number of cases missed in LRS. Methods: We modeled estimated COVID-19 testing demand coverage based on current PCR capacity in three different epidemic phases across five African countries (Strategy 1). We then modelled five additional testing strategies that utilized a combination of PCR and AgRDT: 2)replacing current PCR coverage with AgRDT;3)saturating testing demand with AgRDT only;4)saturating testing demand first with PCR then the remainder with AgRDT;5)saturating testing demand with AgRDT and reflex testing with PCR for patients at risk of severe disease;6)constrained by budget of Scenario 1, using a mix of PCR and AgRDT. We estimated the total number of correct test results expected within a 48hr TAT, corresponding costs (assuming $12/PCR and $6/AgRDT), and the incremental cost-effectiveness ratios for each strategy and epidemic phase by country. Results: Across all countries and phase of epidemic, there was insufficient PCR capacity to meet the calculated required testing demand within a 48hr TAT (ranging from 0-20%) (Figure). In no instance was the base case strategy that was limited to current PCR capacity considered cost-effective (CE). Strategy 3, in which testing demand was saturated with AgRDT, was considered robustly CE in every epidemic phase ($4-$7 per additional person with a correct test result within 48hr TAT), and would require both a large increase in budget and wide AgRDT availability. Additional strategies on the CE frontier were country and epidemic-phase specific. Conclusion: Inclusion of AgRDT in testing strategies is CE and critical in increasing timely testing access in countries with low PCR capacity. Given the importance of timely results for epidemic control, future work should quantify the epidemic impact of saturating testing demand in LRS.