Abstract
The current COVID-19 pandemic contains an unprecedented amount of uncertainty and variability and thus, there is a critical need for understanding of the variation documented in the biological, policy, sociological, and infrastructure responses during an epidemic to support decisions at all levels. With the significant asymptomatic spread of the virus and without an immediate vaccine and pharmaceuticals available, the best feasible strategies for testing and diagnostics, contact tracing, and quarantine need to be optimized. With potentially high false negative test results, infected people would not be enrolled in contact-trace programs and thus, may not be quarantined. Similarly, without broad testing, asymptomatic people are not identified and quarantined. Interconnected system dynamics models can be used to optimize strategies for mitigations for decision support during a pandemic. We use a systems dynamics epidemiology model along with other interconnected system models within public health including hospitals, intensive care units, masks, contact tracing, social distancing, and a newly developed testing and diagnostics model to investigate the uncertainties with testing and to optimize strategies for detecting and diagnosing infected people. Using an orthogonal array Latin Hypercube experimental design, we ran 54 simulations each for two scenarios of 10% and 30% asymptomatic people, varying important inputs for testing and social distancing. Systems dynamics modeling, coupled with computer experimental design and statistical analysis can provide rapid and quantitative results for decision support. Our results show that widespread testing, contacting tracing and quarantine can curtail the pandemic through identifying asymptomatic people in the population.