An Innovative Non-Pharmaceutical Intervention to Mitigate SARS-CoV02 Spread: Probability Sampling to Identify and Isolate Asymptomatic Cases (preprint)

Studies estimate that a substantial proportion of SARS-CoV-2 transmission occurs through individuals who do not exhibit symptoms. Mitigation strategies test only those who are moderately to severely symptomatic, excluding the substantial portion of cases that are asymptomatic yet still infectious and likely responsible for a large proportion of the virus spread (1-8). While isolating asymptomatic cases will be necessary to effectively control viral spread, these cases are functionally invisible and there is no current method to identify them for isolation. To address this major omission in COVID-19 control, we develop a strategy, Sampling-Testing-Quarantine (STQ), for identifying and isolating individuals with asymptomatic SARS-CoV-2 in order to mitigate the epidemic. STQ uses probability sampling in the general population, regardless of symptoms, then isolates the individuals who test positive along with their household members who are high probability for asymptomatic infections. To test the potential efficacy of STQ, we use an agent-based model, designed to computationally simulate the epidemic in the Seattle with infection parameters, like R0 and asymptomatic fraction, derived from population data. Our results suggest that STQ can substantially slow and decrease the spread of COVID-19, even in the absence of school and work shutdowns. Results also recommend which sampling techniques, frequency of implementation, and population subject to isolation are most efficient in reducing spread with limited numbers of tests.

Machine Learning Maps Research Needs in COVID-19 Literature (preprint)

SummaryManually assessing the scope of the thousands of publications on the COVID-19 (coronavirus disease 2019) pandemic is an overwhelming task. Shortcuts through metadata analysis (e.g., keywords) assume that studies are properly tagged. However, machine learning approaches can rapidly survey the actual text of coronavirus abstracts to identify research overlap between COVID-19 and other coronavirus diseases, research hotspots, and areas warranting exploration. We propose a fast, scalable, and reusable framework to parse novel disease literature. When applied to the COVID-19 Open Research Dataset (CORD-19), dimensionality reduction suggested that COVID-19 studies to date are primarily clinical-, modeling- or field-based, in contrast to the vast quantity of laboratory-driven research for other (non-COVID-19) coronavirus diseases. Topic modeling also indicated that COVID-19 publications have thus far focused primarily on public health, outbreak reporting, clinical care, and testing for coronaviruses, as opposed to the more limited number focused on basic microbiology, including pathogenesis and transmission.