Optimizing COVID-19 Testing Resources use with Wearable Sensors
Giorgio Quer, Arinbjörn Kolbeinsson, Jennifer M. Radin, Luca Foschini, Jay Pandit
Abstract
The timely identification of infectious pre-symptomatic and asymptomatic cases is key towards preventing the spread of a viral illness like COVID-19. Early identification has been done through routine testing programs, which are indeed costly and potentially burdensome for individuals who should be tested with high frequency. A supplemental tool is represented by wearable technology that can passively monitor and identify individuals at high risk, alerting them to take a test.
We designed a Markov chain model and simulated a routine testing and a wearable testing strategy to estimate the number of tests required and the average number of days in which an individual is infectious and undetected. According to our model, with 2 test per month available, we have that the number of infectious and undetected days is 4.1 in the case of routine testing, while it decreases by 46% and 27% with a wearable testing strategy in the presence or absence of self-reported symptoms.
Introduction
Throughout the COVID-19 pandemic, rapid antigen testing supply has periodically been unable to keep up with demand, hindering public health authorities’ response to identify new cases and limit further community transmission. One of the challenges for preventing the spread of a viral illness is timely identification of infectious pre-symptomatic and asymptomatic cases. Robust contact tracing, paired with frequent testing of populations, is one way to identify these cases before they have the opportunity to transmit infection to others.
An aggressive way to identify such asymptomatic cases is to enact routine testing programs, already employed by a number of schools and businesses, where individuals were required to test typically 1–2 times per week. Although these programs are effective, they are also costly, time consuming and burdensome to individuals who must be tested at high frequency to identify most infections before they spread.
Results
We specified the available number of tests per month and investigate the performance of the two strategies in terms of number of undetected infectious days (UIDs). A reduction of the UIDs is related to a reduction of the number of individuals that will potentially be in contact to the infected individuals, thus it is related to a reduction in the spread of the disease. In the case of routine testing, if we fix the number of tests per month to 2, according to our model we expect the number of UIDs to be 4.1. Using the wearable sensor trigger strategy, in the absence of self-reported symptoms, the expected number of UIDs is 3.0 (27% decrease in UIDs with respect to routine testing), while in the case with self-reported symptoms it decreases to 2.2 (46% decrease in UIDs with respect to routine testing)
Discussion
Our results suggest that incorporating wearables to inform a testing strategy can decrease the number of tests required while minimizing the number of days an individual is at risk of exposing others. The model is a parametric model that can be tuned to different characteristics of the viral illness under examination, like the probability of developing symptoms or the length of time an individual remains infectious.
Although encouraging, these results are based on a choice of parameters derived from the delta wave of the COVID-19 pandemic and should be re-estimated for new COVID-19 variants, changes in infectiousness from vaccination, and prevalence of infection. Parameters in our model can be adapted to fit an ever-changing pandemic or even used for other viral illnesses in the future.
Citation: Quer G, Kolbeinsson A, Radin JM, Foschini L, Pandit J (2024) Optimizing COVID-19 testing resources use with wearable sensors. PLOS Digit Health 3(9): e0000584. https://doi.org/10.1371/journal.pdig.0000584
Editor: Sulaf Assi, Reader in Forensic Intelligent Data Analysis, UNITED KINGDOM OF GREAT BRITAIN AND NORTHERN IRELAND
Received: March 7, 2023; Accepted: July 16, 2024; Published: September 5, 2024
Copyright: © 2024 Quer et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: We make our model openly available at https://huggingface.co/spaces/arinbjorn/markov allowing the research community and decision makers to test and use the model extensively by varying the underlying assumptions. Additional data and code is available in the Supplement of the paper. We integrate in our analysis the results of other published papers, all accessible and properly cited. No other data beyond the presented data has been collected or used to produce this manuscript.
Funding: GQ and JP were supported by grant UM1TR004407 from the National Center for Advancing Translational Sciences (NCATS) at the National Institutes of Health (NIH). Evidation Health provided support in the form of salary for AK and LF. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: AK and LF were employees of Evidation Health at the time of the analysis. JR was employed by Scripps Research while the work for this manuscript was completed and now works for Moderna. LF is currently employed by Sage Bionetworks. The other authors have declared that no competing interests exist.
Source: https://journals.plos.org/digitalhealth/article?id=10.1371/journal.pdig.0000584#sec003
