Time variation in the probability of failing to detect a case of polymerase chain reaction testing for SARS-CoV-2 as estimated from a viral dynamics model

Designing isolation guidelines for COVID-19 patients with rapid antigen tests

Appropriate isolation guidelines for COVID-19 patients are warranted. Currently, isolating for fixed time is adopted in most countries. However, given the variability in viral dynamics between patients, some patients may no longer be infectious by the end of isolation, whereas others may still be infectious. Utilizing viral test results to determine isolation length would minimize both the risk of prematurely ending isolation of infectious patients and the unnecessary individual burden of redundant isolation of noninfectious patients. In this study, we develop a data-driven computational framework to compute the population-level risk and the burden of different isolation guidelines with rapid antigen tests (i.e., lateral flow tests). Here, we show that when the detection limit is higher than the infectiousness threshold values, additional consecutive negative results are needed to ascertain infectiousness status. Further, rapid antigen tests should be designed to have lower detection limits than infectiousness threshold values to minimize the length of prolonged isolation.

[Digital transformation of COVID-19 research]

In a current life sciences research, we are in an era in which advanced technology emerging and utilize big data. Data-driven approaches such as machine learnings play an important role to analyze these datasets. However, limited clinical (time-course) datasets are available for infectious diseases, cancer, and other diseases. Especially in the case of emerging infectious disease outbreaks, clinical data obtained from a limited number of cases must be used to develop treatment strategies and public health policies. This means that many clinical data are not big data, which often makes the application of data-driven approaches difficult. In this paper, we mainly apply a mathematical model-based approach to the clinical data of COVID-19 and discuss how biologically important information can be extracted from the limited data and how they can benefit society.