The data-powered drive to defeat COVID-19
27 April 2020: statistician Rob Mastrodomenico tells ICAEW Insights about the unsung community using data, predictive analytics and machine learning to combat the coronavirus pandemic.
With much of the world’s economy in a policy-driven coma, its workforce stuck at home on laptops and the general populace in varying degrees of lockdown, there’s a data-driven push to help bring an end to the COVID-19 pandemic.
One recent story to emerge revealed how NHSX – the digital arm of the health service – is heading up an effort to work with a number of tech firms to create dashboards that will draw on data gathered via 111 calls and COVID-19 test results to help the government and the NHS better access a range of information to map the spread of the virus and improve the healthcare system's ability to deal with it.
It is hoped the dashboards can use the information as soon as it becomes available to help understand how the virus is spreading. Other anticipated benefits are identifying the risks to particularly vulnerable groups as well as proactively increase resources in emerging hot spots. Another hoped-for benefit is to ensure critical equipment is supplied to hospitals and other facilities in greatest need.
Rob Mastrodomenico is a key member of the ICAEW’s data analytics community. With a doctorate in statistics, and as the owner of a company specialising in the statistical modelling of sports, he is uniquely well placed to walk us through some of the great body of work being undertaken to help fight the pandemic – much of it unsung – and from a diverse community of people using data, predictive analytics and machine learning skills.
“It’s a massive piece of work”, says Mastrodomenico with individuals, small teams and “lots of institutions and academics within the government and beyond all putting various types of effort in”.
“We haven’t really had an event of this level before with the technology we now have. We have these mobile communication tools that tell us where we are at all times, can log where we are and what we’re doing, and all that information and data is so much bigger than we’ve ever had before, so the people using it now have so much more to play with.”
Mastrodomenico says that a number of people he knows that are working on the pandemic “will be modelling the disease to help manage the current situation” while others are “looking at vaccines to help us going forwards”.
“They’re all doing their bit individually to try and come up with something that can help either tell us where we are within it or help us get out of it. It’s a huge effort”.
But he’s quick to point out that many people are engaged on this sort of work on an ongoing basis and perhaps it is only now that the wider world is alive to it.
“A lot of this has been brought to the forefront because of where we are but a lot of great statisticians and epidemiologists who work in these areas are trying to find cures for diseases or model this type of thing all the time.”
He believes that “we’re incredibly lucky to have such a great pool of expertise from which to draw on and fight such challenges”.
“There’s a whole sector and industry that perhaps isn’t really known about because it's not often brought to the forefront, but when it’s needed we have great statisticians, great doctors and nurses, people who are doing these roles day in day out and they are looking to make a difference across a variety of ways.
“The call to arms is now and they’re all doing their bit but when this is over they’ll still be doing what they’re doing – everybody is trying to find cures for diseases and not just in a pandemic, trying to make everyone’s lives better.”
Mastrodemenico said he hoped wider society “realises that there are lots of people doing lots of good work with the use of data, statistics, modelling and technology” to make tangible and positive differences.
He cites the collective of UK-based Formula 1 teams, engine manufacturers and their technology divisions coming together to evaluate their support for the government’s drive to manufacture ventilators, as a case in point.
“In this country, we can do a lot of good when there’s a common goal and when we pull together our expertise it’s a really positive thing. Whilst people have died and it’s been terrible, the work everyone’s been doing should be commended.”
On the wider issue of privacy concerns surrounding the use of data, Mastrodomenico acknowledges that those using people’s data cannot be “cowboyish”, and there needs to be care taken around how it is used and collected, particularly with the introduction of GDPR.
But he’s also acutely aware that as users of mobile technology and apps “we give away so much data about ourselves”, admitting he’s unsure “how people will look back on this era and how much we gave away” given that “data is a huge commodity” which “allows companies to make loads of money from it”.
But with the right privacy checks and balances in place, it has the potential to be positively transformational at a societal level.
While delighted that the words epidemiology and epidemiologist have now fully entered the mainstream lexicon of the zeitgeist, he is, like most of us, uncertain when all these collective big data-driven efforts might lead to the beginning of the end of lockdown.
And despite talk of being near or passing the peak of the pandemic in the UK, the high level of “unknowns and variables” means the government has been “a little like a statistician” on this, Mastrodomenico believes.
“They have to be quite conservative on this. As a statistician, you’re never very bullish about the results you find, everything is evidence-based but it's not absolutes. So, we would never say something is significant, we would say there is evidence of significance, and the government must be conservative on this as there has to be a lot of evidence it’s the right thing to do. Getting it wrong and having that second wave would be a disaster.
“But the economic stuff is important too and they’re balancing a lot of things at once and doing the best they can. It’s not like we can look back on historical data on this. Generally if you’re looking at machine learning, regression prediction, any kind of modelling, you’re trying to look back on what you’ve seen before to help make a prediction in the future but we don’t have any data here on what’s the right thing to do – they’re making history up as they go along.
“My hope is that we can gently ease back in. My primary thing is sport, and I don’t see any live sport happening for quite a while yet, so I think it’ll be a slow easing back.”
For most people, that moment can’t come soon enough.
Rob Mastrodemenico will take questions on data preparation tips and techniques at 12.30pm on 26 May as part of an ICAEW data community event. Listen to the on-demand webinar first, then pose your questions live. For more details click here.
For the latest news and guidance on the ongoing impact of COVID-19 for businesses and accountants, visit ICAEW’s dedicated coronavirus hub.