10 Jul 2020
What Big Data from Previous Virus Outbreaks Tells Us About Our Future Living with COVID-19
One of the most used terms around the coronavirus pandemic is ‘unprecedented’. Yet the last 100 years has seen three viral pandemics. Many news articles have referenced the Spanish Flu in 1918, but there was also an Asian Flu outbreak in 1957 and Hong Kong Flu in 1968-69, the latter of which is estimated to have killed more than one million people. It is difficult to gauge much insight from these pandemics because the world is a very different place now to what it was 50 or 100 years ago. Globalization and international travel have vastly increased the velocity at which a pandemic can take hold. Taking a deep dive into news archives can provide valuable insights, whether you conduct a simple search within Nexis® or use the Nexis® Data as a Service bulk API as a pipeline to power predictive analytics.
SARs in the Nexis news archive
A more helpful example is the SARS (Severe Acute Respiratory Syndrome) outbreak that occurred in 2002-03. Both SARS and coronavirus are thought to have originated in the animal kingdom before mutating and infecting humans. The viral loads for both diseases are transferred via respiratory droplets.
Like COVID-19, SARS emerged in China and rapidly spread, in this case to 26 countries. Its epidemic trajectory however was radically different. There were, in total, a little more than 8,000 reported cases of SARS with 774 deaths. SARS was also contained within eight months of the first reported case. The methods used to contain SARS do sound very familiar to anyone reading the news today. Key tools in containment were prompt isolation of anyone suffering symptoms, with wider quarantines where required, combined with quick and accurate contact tracing to prevent the virus spreading.
Reporting on the SARS virus’ outbreak became global during March and April 2003. Reports surfaced on Friday 28th March 2003 that at least 53 people had died from the virus from around 1,400 infections. At this point Singapore, Vietnam, Hong Kong, Beijing, and Toronto were hotspots, with the World Health Organization (WHO) advising against travel to these areas. Reports also referenced school closures and a fall in the share prices of airlines, tourism, and leisure companies but as is the case with COVID-19 technology stocks remained strong.
By Monday 31st March Agence France-Presse (AFP) was reporting the ramping up of SARS cases in Hong Kong, with 100 new cases reported and growing disruption to travel and business. Over the weekend six more people had died while the infection rate had risen by 200, small when compared to the growth in the coronavirus in 2020 but still an exponential rise. AFP reported on “a trail of empty restaurants and shopping malls” in Hong Kong.
Big data from media shows airlines particularly disrupted
The AFP report also highlighted a tough time for the airline industry. The SARS outbreak took place in the immediate aftermath of the US invasion of Iraq and just 18 months after the terror attacks of September 11th, 2001. This triple whammy had a sustained impact on the industry, as reported the next day by United Press International with the headline: ‘Air Canada seeks bankruptcy protection’. By 2nd April Australian Financial Review was reporting on drops in passenger traffic in Asia of up to 70 percent. The news outlets commented that ‘Malaysian Air yesterday joined Cathay Pacific Airways, Singapore Airlines and other carriers which are facing falling revenues as the virus …hurts air travel demand’.
Increasing numbers, familiar actions found in news archives
By 2nd April focus had once again settled on the business impact of the disease with USA Today reporting that ‘everyday acts such as shaking hands suddenly seem dangerous’. By that date SARS had infected more than 1,900 people and deaths had increased to 76 worldwide. USA Today reported on an entire Motorola factory in Asia closed, with impact on other companies including Xerox, Lucent, HP, Microsoft, AT&T and Eastman Kodak. Employees were working from home and cancelling business travel. On the same day another US publication, the Atlanta Journal Constitution, raised the possibility of quarantining US residents if SARS became as widespread as in other countries.
So much of what the media was reporting in March and April 2003 is like today’s news. However, SARS did not manifest itself globally and cases were kept to less than 10,000. At the time of writing COVID-19 has infected 5.6 million people with more than 348,000 deaths attributed to the disease - an entirely different level of magnitude.
The ability of the world to shut the disease down in a matter of months offers some hope for the procedures and practices being followed today. Social distancing, working from home, contact tracing and isolating individuals with symptoms were clearly effective against SARS.
Those airlines referenced in the 2003 articles all continue to do business, albeit having experienced significant restructuring along the way. It took nine months after SARS for international air travel to reach the same level it had previously been before the outbreak. It is probably fair to say the recovery time after the coronavirus is contained will be much longer.
As for the technology companies that broadly seemed to ride out the storm, of those mentioned in the quoted articles in this blog, both Lucent Technologies and Motorola no longer exist in the way they did in 2003, while Eastman Kodak experienced bankruptcy a decade later. This demonstrates that business risk is bigger than one single event.
Learning the lessons of the past with predictive analytics
Events like coronavirus are extremely rare and living through them can be shocking and scary. News reports of similar, albeit smaller, events from the past can offer us a semblance of context. In the case of SARS, it is reassuring that the disease was eradicated from humanity within a short timescale by applying some of the tactics we are seeing deployed by some governments globally today. Clearly the scale of these crises cannot be compared but the past can help us more clearly understand what the future might hold.
If we can extract some small insights from a quick look through our news archive, imagine the valuable intelligence you can uncover when you combine the breadth and depth of Nexis Data as a Service with machine learning, predictive analytics and other AI-driven research.
Find out how access to a 40+ year archive of media data supports predictive analytics to help organizations make informed decisions in disruptive times.