Portal de Periódicos da CAPES

Revisiting the county/city-level event risk assessment during the COVID-19 pandemic

2021; Elsevier BV; Volume: 82; Issue: 5 Linguagem: Inglês

10.1016/j.jinf.2020.12.031

1532-2742

Akihiro Nishi, Lily F. Lee, Hiroshi Tsuji, Yohsuke Takasaki, Sean D. Young,

Risk Perception and Management

Risk communication is vital in medicine and public health.1Lohiniva A.L. Sane J. Sibenberg K. Puumalainen T. Salminen M. Understanding coronavirus disease (COVID-19) risk perceptions among the public to enhance risk communication efforts: a practical approach for outbreaks.Euro Surveill. 2020; 25 (Finland, February 2020) (PubMed PMID:32265008Pubmed Central PMCID: PMC7140598. Epub 2020/04/09)Crossref PubMed Scopus (96) Google Scholar,2Abrams E.M. Greenhawt M. Risk communication during COVID-19.J Allergy Clin Immunol Pract. 2020; 8 (PubMed PMID:32304834Pubmed Central PMCID: PMC7158804. Epub 2020/04/19): 1791-1794Abstract Full Text Full Text PDF PubMed Scopus (149) Google Scholar Risk-related information can differentially affect people's attitude and behaviors depending on how the information is presented.3Young S. Oppenheimer D.M. Effect of communication strategy on personal risk perception and treatment adherence intentions.Psychol Health Med. 2009; 14 (PubMed PMID:19697253Pubmed Central PMCID: PMC2956070. Epub 2009/08/22): 430-442Crossref PubMed Scopus (26) Google Scholar Tools that properly assess and communicate health-related risks are urgently needed by health departments and governments to inform their decision-making. A recent paper in Journal of Infection4Furuse Y. Risk at mass-gathering events and the usefulness of complementary events during COVID-19 pandemic.J Infect. 2020; (PubMed PMID:33271175Epub 2020/12/04)Abstract Full Text Full Text PDF Scopus (4) Google Scholar (and others5Chande A. Lee S. Harris M. Nguyen Q. Beckett S.J. Hilley T. et al.Real-time, interactive website for US-county-level COVID-19 event risk assessment.Nat Hum Behav. 2020; 4 (PubMed PMID:33168955Epub 2020/11/11): 1313-1319Crossref PubMed Scopus (43) Google Scholar,6Eisenstein M. What's your risk of catching COVID? These tools help you to find out.Nature. 2021; 589: 158-159Crossref PubMed Scopus (5) Google Scholar) responded to such an urgent need and aimed to make the "invisible" risk of COVID-19 at mass-gatherings visible and available to stakeholders and the general public. This letter attempts to provide feedback to improve both the utility of the tool and the likelihood for its successful implementation. The author introduced a series of formulas and a R shiny app (https://yukifuruse.shinyapps.io/covid_eventrisk_en/), which inputs the daily number of newly reported cases of a region, the population size of the region, and expected attendees at an index event held in the region and outputs the probability of the event containing at least one infectious individual. The formulas could tell users the level of their "risk" in sharing the same environment (event) with one or more infectious individuals. This is analogous to a weather forecast, which informs users of the "risk" of rain. The formulas and the app can reflect continuous updates in the SARS-CoV-2-related evidence (e.g. asymptomatic ratio) and input region-specific parameter values (e.g. the fraction of the reported cases by testing among the actual infected individuals). The app has garnered growing media attention in Japan and other countries (e.g. covered by TV Tokyo, YouTube [https://www.youtube.com/watch?v=r86YO7FZWxs], and Wall Street Journal [https://www.wsj.com/livecoverage/covid-2020–12–03/card/VleP5zyCw8feYTIybuQf]), where self-restraint requests made by governments are not legally enforced; as a result, holding/suspending or attending/not attending an event largely depends on the hosts' and attendees' discretion. Without the app, people would have difficulty knowing the level of the local risk involved in their events and going-out decisions. Therefore, the app could improve people's risk perception and enhance the scientific communication on SARS-CoV-2. A potential pitfall in the app use relates to the limitations drawn by the lack of the micro-level network data and the assumptions of the formulas. The formulas assume that infectious individuals are randomly located in the social networks within the index region (of the event) and that the attendees of the event are randomly selected from the region. To illustrate this importance, we calculate the probability that there will be at least one infectious individual at an event of 50 attendees in a region of 100,000 individuals with 20 new cases per day using the app's default setting on these parameters as of Dec/28/2020, which is 35.2%. What is the origin of the 20 daily new cases there? They are typically secondary cases arising within the region. When that is the case, the infections do not occur at random locations, but rather spread like a snowball in the region's social networks.7Nishi A. Dewey G. Endo A. Neman S. Iwamoto S.K. Ni M.Y. et al.Network interventions for managing the COVID-19 pandemic and sustaining economy.Proc Natl Acad Sci U S A. 2020; 117 (PubMed PMID:33177237Pubmed Central PMCID: PMC7720236. Epub 2020/11/13): 30285-30294Crossref PubMed Scopus (51) Google Scholar,8Block P. Hoffman M. Raabe I.J. Dowd J.B. Rahal C. Kashyap R. et al.Social network-based distancing strategies to flatten the COVID-19 curve in a post-lockdown world.Nature Human Behaviour. 2020; 4 (2020/06/04): 588-596Crossref PubMed Scopus (335) Google Scholar Here, let's divide the region into two "communities": the first community contains a majority of the cases (e.g. 15 cases in 10,000 population), and the second community contains the rest (5 cases in 90,000 population). The divide of a region's singular social network may relate to geography, age, risk preference, occupation, or others. Most of the individuals in the first community have a smaller "degree of separation" from the infectious individuals compared to those in the second community. Then, let's recalculate the probability using the same formulas per community. The probability of the event of 50 attendees to contain at least one infectious individual in the first community is 96.5%. If the event is planned for the first community, the app's estimation of 35.2% could lead people to dramatically underestimate their true risk of attending the event (35.2 compared to 96.5). This underestimation may make people overconfident in the first community (more widespread communities) and contribute to further infection spread there. The recalculated risk in the second community is 11.3%. If the event is planned for the second community, the app's estimated risk of 35.2% could lead people to over-estimate their risk and avoid the event due to misperception of risk (35.2 compared to 11.3). This overestimation may make people in the second community (less widespread communities) less confident and result in unnecessarily sacrificing social and economic activities there, which may cause "quarantine fatigue".7Nishi A. Dewey G. Endo A. Neman S. Iwamoto S.K. Ni M.Y. et al.Network interventions for managing the COVID-19 pandemic and sustaining economy.Proc Natl Acad Sci U S A. 2020; 117 (PubMed PMID:33177237Pubmed Central PMCID: PMC7720236. Epub 2020/11/13): 30285-30294Crossref PubMed Scopus (51) Google Scholar,9Zhao J., Lee M., Ghader S., Younes H., Darzi A., Xiong C., et al. Quarantine Fatigue: first-ever decrease in social distancing measures after the COVID-19 outbreak before reopening United States.arXiv:200603716. 2020.Google Scholar In sum, as with any new tool or program, this app and its effects should be studied to better understand its potential implications and implementation-related issues, including people's willingness to attend events, and health departments' response to event risks. The issue presented in this letter does not appear to stem from the formulas itself, but from the lack of data availability on COVID-19 case data at the municipality level due to privacy protection in most countries. This issue, which is shared with comparable event risk assessment tools5Chande A. Lee S. Harris M. Nguyen Q. Beckett S.J. Hilley T. et al.Real-time, interactive website for US-county-level COVID-19 event risk assessment.Nat Hum Behav. 2020; 4 (PubMed PMID:33168955Epub 2020/11/11): 1313-1319Crossref PubMed Scopus (43) Google Scholar,6Eisenstein M. What's your risk of catching COVID? These tools help you to find out.Nature. 2021; 589: 158-159Crossref PubMed Scopus (5) Google Scholar (e.g. in ref,5Chande A. Lee S. Harris M. Nguyen Q. Beckett S.J. Hilley T. et al.Real-time, interactive website for US-county-level COVID-19 event risk assessment.Nat Hum Behav. 2020; 4 (PubMed PMID:33168955Epub 2020/11/11): 1313-1319Crossref PubMed Scopus (43) Google Scholar the event risk is provided at the county level in the US, UK, and other European countries), might be resolved if the microdata of communities (where we can assume random mixing) are made available. AN is a consultant to Urbanic & Associates. HT is the Founder and CEO, Corporate Health, Inc. in Japan. YT is the CEO of Decades Inc. in Japan, and an employee of SoftBank Corp. SDY is a consultant and advisor for digital health startups, and receives royalties from HarperCollins Publishers for the book, Stick with It. Support for this research was provided by a grant from the UCLA Fielding School of Public Health High-Impact Data Initiative (AN). Additional support for AN and SDY was provided by grant P30DA027828 from the National Institute on Drug Abuse, awarded to C. Hendricks Brown. SDY received support from the National Institute of Allergy and Infectious Diseases (NIAID) Young: 7R01AI132030; and National Institute of Mental Health (NIMH) 5R01MH106415. The opinions expressed herein are the views of the authors and do not necessarily reflect the official policy or position of the National Institute on Drug Abuse, or any other part of the US Department of Health and Human Services.