Portal de Periódicos da CAPES

Data journalism and the COVID-19 pandemic: opportunities and challenges

2021; Elsevier BV; Volume: 3; Issue: 10 Linguagem: Inglês

10.1016/s2589-7500(21)00178-3

2589-7500

Angel N. Desai, Pierre Nouvellet, Sangeeta Bhatia, Anne Cori, Britta Lassmann,

COVID-19 Digital Contact Tracing

Academics and public health practitioners studying communicable disease dynamics have long advocated for open-access data to better inform risk assessments. During any evolving outbreak, the collection, aggregation, visualisation, and analysis of granular data is paramount to developing appropriate public health interventions.1Desai AN Kraemer MUG Bhatia S et al.Real-time epidemic forecasting: challenges and opportunities.Health Secur. 2019; 17: 268-275Crossref PubMed Scopus (51) Google Scholar The COVID-19 pandemic has underscored the need for this type of information, especially in relation to context (eg, timing and intensity of interventions) and epidemiology (eg, spatially resolved and age-specific case counts). Non-traditional disease surveillance tools, including news media reporting, have disseminated event-based information during past disease outbreaks.2Brownstein JS Freifeld CC Madoff LC Digital disease detection—harnessing the Web for public health surveillance.N Engl J Med. 2009; 360: 2153-2157Crossref PubMed Scopus (512) Google Scholar The current global health crisis has highlighted the additional possibilities that so-called data journalism can offer. While the news media has traditionally reported on events of public health importance, media outlets over the course of the COVID-19 pandemic have also conducted data collation, including detailed summaries of case counts and deaths, data curation, and, in some instances, analysis (table). Although some of these data are available through public health department websites, relevant interpretation and data visualisations by media outlets have provided information on the COVID-19 pandemic to the public in near real time.TableExamples of COVID-19 data collection, visualisation, and analysisDescriptionData sourcesURLFinancial Times: coronavirus trackerThe Financial Times analyses the scale of the COVID-19 outbreak including the collection and analysis of data on excess mortality (ie, numbers of deaths higher than the historical average) across the globeWHO, the COVID Tracking Project, Johns Hopkins University, Our World in Data, US Centers for Disease Control and Prevention, and othershttps://www.ft.com/content/a2901ce8-5eb7-4633-b89c-cbdf5b386938The Economist: tracking COVID-19 excess deaths across countries globallyGlobal excess death trackerHuman Mortality Database, World Mortality Dataset, and EuroMOMOhttps://www.economist.com/graphic-detail/coronavirus-excess-deaths-trackerThe New York TimesThe pandemic's hidden tollExcess deaths during the COVID-19 pandemicData are compiled from official national and municipal data for 24 countrieshttps://www.nytimes.com/interactive/2020/04/21/world/coronavirus-missing-deaths.htmlTracking the coronavirus at US colleges and universitiesCOVID-19 tracker at US colleges and universities; with no national tracking system, and statewide data available only sporadically, colleges have been making their own rules for how to tally infectionsThe New York Times surveyed more than 1900 US colleges and universities for COVID-19 informationhttps://www.nytimes.com/interactive/2020/us/covid-college-cases-tracker.htmlWhat we know about coronavirus cases in K-12 schools so farReporting focused on district-level and statewide COVID-19 case totals for public schools in the USA; the numbers presented are minimums because of differences in reportingState and local health and education agencies or were identified by The Covid Monitor or the National Education Association and independently confirmed by The New York Times; The New York Times directly surveyed every school district in eight states: Colorado, Florida, Georgia, Illinois, Indiana, North Carolina, Texas, and Utahhttps://www.nytimes.com/interactive/2020/09/21/us/covid-schools.htmlHow full are hospital ICUs near you?Occupancy levels in US ICUsUS Department of Health and Human Services (hospital capacity data); US Department of Homeland Security (hospital locations); and the Covid-19 Hospitalization Tracking Project, University of Minnesota Carlson School of Managementhttps://www.nytimes.com/interactive/2020/us/covid-hospitals-near-you.htmlThe Atlantic: the COVID-19 tracking projectCollected, cross-checked, and published COVID-19 data from 56 US states and territories regarding testing, hospitalisation, and patient outcomes, providing ethnic demographic information and data on long-term-care facilities*As of March 7, 2021, The Atlantic is no longer collecting new data.COVID-19 data from websites of US state or territory public health authorities; a public data API provides access to all their data at a national and state levelhttps://covidtracking.comThe Hindu: coronavirus India trackerCollects, aggregates, analyses, and visualises state-level COVID-19 cases, deaths, and testing data from India and globallyBing and Johns Hopkins Universityhttps://www.thehindu.com/coronavirus/Zeit Online: coronavirus in Deutschland und bei Ihnen [coronavirus cases in Germany and globally]Collects, aggregates, analyses, and visualises COVID-19 cases, deaths, patients in ICUs, and vaccinationsRobert Koch Institute, websites of German counties and states, and Johns Hopkins Universityhttps://www.zeit.de/wissen/gesundheit/corona-zahlen-deutschland-neuinfektionen-inzidenz-aktuelle-karteICU=intensive care unit. API=application program interface.* As of March 7, 2021, The Atlantic is no longer collecting new data. Open table in a new tab ICU=intensive care unit. API=application program interface. Before COVID-19 was declared a Public Health Emergency of International Concern, news reports served as key data sources to further understand disease transmission and spread. Academic institutions and researchers assembled early epidemiological data scattered across various news articles to inform risk assessments, forecasts, and policy decisions.3Imai N Dorigatti I Cori A Riley S Ferguson NM Report 1: estimating the potential total number of novel Coronavirus cases in Wuhan City, China. Version 2. Imperial College COVID-19 Response Group.https://doi.org/10.25561/77149Date: Jan 17, 2020Date accessed: July 6, 2021Google Scholar The relative dearth of traditional public health data at the beginning of an epidemic is not a new phenomenon. During the 2014–15 west Africa Ebola outbreak, for example, early epidemiological data were often only available through local and international news media articles.4Cori A Donnelly CA Dorigatti I et al.Key data for outbreak evaluation: building on the Ebola experience.Philos Trans R Soc Lond B Biol Sci. 2017; 37220160371Crossref PubMed Scopus (52) Google Scholar As the COVID-19 pandemic evolved, and in response to epidemiological data gaps, news media outlets began to collect and synthesise data for scenarios involving congregate settings such as schools, large public events, and household transmission. In some cases, media have actively reached out and solicited case counts from their readers—a strategy known as participatory surveillance—effectively recruiting the public back into public health (table). News media outlets have also been among the first to systematically collect, aggregate, and analyse excess death counts. For example, the data behind the Financial Times tracker for COVID-19 excess deaths dates to April, 2020; the tracker is open access, and the code and methodology used to clean, analyse, and present the data are available on GitHub. The Economist and The New York Times have also provided their own analyses on excess deaths (table). As the COVID-19 pandemic has shown, there is an urgent need for real-time data that can inform risk assessments to guide public health interventions. While traditional data collection remains the cornerstone of outbreak response, public health programmes and information technology infrastructure are chronically underfunded in many countries and are not always well positioned to collect contextual information in a flexible manner. This is of particular concern during an outbreak when traditional data sources might lag in reporting cases early on. Another key need for epidemic forecasting and risk assessments is data surrounding non-pharmaceutical interventions such as physical distancing, school closures, and lockdowns.5Blavatnik School of GovernmentUniversity of OxfordCOVID-19 government response tracker.https://www.bsg.ox.ac.uk/research/research-projects/coronavirus-government-response-trackerDate accessed: July 6, 2021Google Scholar Interventions differ regionally and implementation timelines are not often readily disseminated. While non-traditional disease surveillance systems have begun to fill some of these gaps, more can be done. Partnerships between academic research centres and news media should be considered, given their complementary strengths; indeed, collaboration between these entities might mitigate their respective weaknesses as well. While news media can rapidly aggregate and disseminate information, they might be unable to sustain these efforts following the course of an outbreak. Likewise, research centres might be able to continue collating and analysing data long after an outbreak has ended, but might be unable to collect relevant information in a timely manner early in an outbreak. It is also important to note that news media data and data visualisations, while informative, differ from peer-reviewed literature. Divergent incentives, intended audiences, and analytic methodologies can result in very different outputs and conclusions. Supporting collaborations between news media outlets that can provide an expedient data stream and academic institutions that can support targeted analyses could be an important step towards improving outbreak response timeliness in the future. In response to the COVID-19 pandemic, several global epidemiological data collection and harmonisation efforts have been initiated to provide guidance on conforming case definitions, data formatting, and data sharing. As these efforts are further developed, data collected by media outlets could be integrated for use by researchers and policy makers, although regulatory issues surrounding data privacy will need to be addressed. Cross-collaborations between academic groups and the media should be encouraged and the role of the media in curating, analysing, and sharing epidemiological information that is otherwise hard to collect should be recognised. While these efforts should be considered complementary to traditional public health endeavours, the rapid dissemination of accurate, real-time information remains paramount in the face of current and future communicable disease outbreaks. We declare no competing interests.