As of 7 April 2020, there were some signs that the increase in infections and, eventually, deaths caused by CV-19, were increasing more slowly and that the curves may be flattening out. Although statistics are confused, even within an individual country, by variations in the number of people tested, reporting delays (and, in some countries, concealment) and different ways of reporting cause of death, the trend appears encouraging.
Eventually, the pandemic will run ‘backwards’ in time, with the epidemic curve being bought down by running out of people to infect: either there aren’t any more susceptible people to infect, or isolating and distancing measures will have reduced the contact rate sufficiently so that a given area and eventually all areas get down to zero new cases.
A particularly eye-catching example of this slightly slippery concept can be seen on these excellent graphs in an article by Danny Dorling of the University of Oxford, one of which appears on the header image. These concern themselves not with actual numbers of deaths attributable to CV-19 – which will always continue upward or plateau but, unless sources are revised, never fall – but with change. A movement up represents a higher number of deaths and movement to the right a higher increase. What is good, and is starting to happen, is movement down and to the right, which indicate the reverse.
Much of Europe seems about to go make this favourable transition – for now. However, the pandemic can re-boot itself simply by people starting to mingle or travel again. Our travel habits were the engine that have powered the pandemic so far. The ‘forward’ pandemic went from Wuhan to South Korea and Singapore; then to Italy and France; and, in parallel, to Japan and the USA. All this took no more than a month or two. It could happen again.
The consequent social restrictions will not have been experienced before by anyone – even those who remember WWII – who has not spent time in North Korea or a high-security prison. The pressure for their relaxation is strong. To do so without learning from what has happened will be to re-set the clock to 1 Jan and watch the whole nightmare play out again, sparing only the uncertain number of people who have developed an immunity.
A vaccine is the obvious holy grail. However vaccines at scale are 12 to 18 months off. It’s also not yet clear how long they would last and how many clades (different families) of Coronvirus they would be effective against. Much the same can be said about antibody tests, which will demonstrate if the person has had the virus. So – what else is there?
Some of what follows might seem a bit dystopian. The reality which we need to face is that when we get back to ‘normal’ it will be a new normal. One could pick half a dozen areas which this might apply. The one we’re going to look at concerns contact testing and, intimately connected with this, a revision of some of views about the nature and perhaps the purpose of data privacy.
In normal times, many of us might not choose to live in Singapore or South Korea, countries with a high level of centralised power and control. They are, however, the standard by which other counties’ responses have been judged (and fallen short). It’s no co-incidence that, in both, government agencies that require personal data for public interest purposes can collect and use data without the need to obtain consent. The last few years have seen plentiful examples of breaches of data security (and, in some cases, as with Facebook and Amazon, the growing realisation that the whole service is predicated on providing information about yourself in order that they can nudge you towards people and products that match your tastes). In most countries, there has so far been no compelling public-interest case for using data without consent. Indeed, the recent GDPR regulations have helped establish individual rights. CV-19 has changed that.
There has also been an assumption that anyone can at any time travel wherever they want to can can afford to go without, except in some specific cases (see below) proving that they are free of disease. CV-19 has changed that, too.
So what might we expect between, say, 1 May 2020 and 1 May 2021?
There needs to be in place a stringent regime of contact tracing, testing, and certificates of immunity (had the virus and had recovered/had antibodies) to allow travel to return to anything like its previous levels. There will need to be a widespread acceptance that certification is required. Tests are getting very cheap and increasingly accurate and fast, which will help with this.
There is form on this. There’s already a system for this for proof you’ve been vaccinated for yellow fever, without which you cannot enter some countries. Both WHO and national teeth are sharp on this – spreading a notifiable disease is a pretty serious crime. At some point one can imagine everyone entering a country being required to have a test, or produce a certificate – or even both, given immunity isn’t guaranteed to last. Even when there’s a vaccine, it won’t be 100% effective and it won’t last more than perhaps a few years. So, certificates and testing may be here to stay.
So too might contact tracing.
Contact-tracing apps now exist on smart phones. Combining these with testing for the virus and antibodies help form part of what might be termed a digital-social immune system. It will also require, as suggested above, a sea-change in the way GDPR and data-collection laws are enforced and public acceptance of the need for, and confidence in the execution of, these changes.
Using proximity detection via Bluetooth radio, or else carefully processing location service data, we can infer who was likely to have been co-located with someone and for how long. This can be accurate to within metres and seconds.
When the phone carrier reports systems, or tests positive for infection, the list of recent contacts, together with an interview to check for additional people who may not have been carrying phones, can notify a set of people that they may have been infected and to be alert for symptoms, or to get tests as soon as possible if they are not already immune.
Additionally, the distribution of contacts and timing of presentation with symptoms and immunity give public health experts more precise models to predict the progress of the outbreak, to measure the effectiveness of interventions such as social distancing advice or lockdown, by area or demographic, and more accurately to model how the relaxing of interventions will play out. All evidence is subject to false positives and false negatives, so designs should maximise effectiveness in correct notifications while minimising false contact-tracing workload.
This needs to be extremely cautious about subject privacy. The concern goes beyond the vague fear that someone might hold the kind of personal data that might be available on a business card but enters into the much more serious realm of medically confidential data. There are some vital principles that need to be agreed. At no point, for instance, should personal data be re-linkable, except with explicit (or in the case of hospitalisation, implicit) consent. Notifications to potential contacts should be privacy-preserving. There is much work being done on this and, at present, a broad acceptance of the need to marry the needs of personal privacy with the public interest. Crises have a way of providing consensus which normality, with the possibility of a decision deferred to another day, never can.
Such systems are only effective if they’re used in large numbers, which demands an equally high levels of public trust. However, the numbers involved are not as large as one might think. 40% of the population using whatever app was designed would be enough to make the exercise useful. The best is the enemy of the good. 60% or 99% might be ideal but the process has to start somewhere, and quickly. At present, a good deal of leg work by NHS staff and organisations like the Police is need to identify the people with whom an infected person might have come into contact on a given day. If you were asked whom you had met on 27 March, would you be able to remember? Digital evaluation of your movements would enable clusters and patterns to be established far more quickly.
Such an approach will also help with a lot of other public-health issues, prevention being in every case better than cure.
In the absence of universal and frequent testing, contact tracing has the potential to complement the coarse-grained statistics with which experts currently have to deal. Processing such real-time and detailed data will allow early lifting of this or any future lockdown with assurance that we can prevent the dead-cat bounce of the pandemic.
There’s a powerful argument that public-health considerations regarding controlling infectious diseases trump any views about personal privacy. It’s also the case that technology, which can pinpoint us and others in time and space, can provide the necessary information. Anything can be perverted to ends not anticipated by its creator. There are risks. Huge corporations and governments, for whom such data would be like gold, will be involved. Can they be trusted? Would the data be secure? As with so many decisions, there is no clear answer. As with so many things, it comes down to a balance of good and bad. Do the likely advantages outweigh the possible drawbacks?
There won’t be time for six-week consultations and public forums so the whole thing, initially at least, will need to be voluntary. There has never been a better time for this to happen. If we’re at all serious at surviving – the word is an exact one – without regular repetitions of what we’re going through now, we should seize it. We’re talking weeks, rather than months or years, for this to be started.
Written by Brian Quinn and Jon Crowcroft, Marconi Professor of Computer Science at the University of Cambridge and Researcher at Large for the Turing Institute.
Image: taken from Three Graphs that Show a Global Slowdown in COVID-19, Danny Dorling/Kirsten McClure, republished from The Conversation, first published April 7th 2020.