To fight COVID-19, the world needs science, and science needs data

It may seem an eternity to us all now, but it has been less than four months since COVID-19 emerged as a frightening threat to global health. The death toll, globally, is now well over 52,000 and rising. How high this number will go depends on the effectiveness of a global machine of science and science-informed policy. Engineering this machine, driving it forward even as it is being built, is a partnership between governments and the global scientific community. The world’s political leaders must make difficult decisions quickly, under considerable pressure, relying on incomplete data and uncertain models. While governments cannot wait for complete information, what information they can gather must be used to fuel decisions and action. To accelerate the narrowing of this policy and data gap, the Canadian Institute for Advanced Research recently convened a virtual roundtable of more than 70 scientists from six countries to discuss how advances in AI and machine learning, many developed right here in Canada, can help. Topics included the usefulness of AI to improve epidemiological modelling, drug and vaccine development, clinical trials, and health-system capacity and resilience. The resulting report makes it clear that AI can indeed offer solutions, but it requires large amounts of data, much of it in real time. Barriers to data mean barriers to effective treatments, to faster and better health system mobilization, to economic recovery, and to a vaccine. Some Canadian data barriers exist for good reasons, some due to the right to privacy or jurisdictional issues, and some due to technology and process challenges. But during a pandemic, the protection of individual data privacy must be balanced against the public good. Governments around the world recognize this. The World Health Organization, the Organisation for Economic Co-operation and Development, and a recent statement from the Wellcome Trust co-signed by numerous global organizations have all affirmed this position. Indeed, the past week brought news of the Pan-European Privacy Preserving Proximity Tracing, a project from 130 researchers in eight countries using phone apps for contact tracing. German Chancellor Angela Merkel, herself in isolation after a COVID-19 contact, has stated her support for such an app if it is effective and voluntary. We know that Europe has some of the world’s most stringent privacy laws, but the terrible lessons of Italy and Spain have made their mark. No doubt Europe is looking to South Korea’s success in early containment of the pandemic, by following the advice of the WHO: test, trace, and treat. As Dr. Tedros Adhanom, director general of the WHO has said, “You cannot fight a fire blindfolded. We cannot stop this pandemic if we don’t know who is infected.” Everywhere in this interconnected world, and right here in Canada, we must launch widespread testing that is inexpensive, easy to use at home, and acceptably accurate, possibly by combining PCR-based and serology-based technologies. But beyond testing, we must also ramp up tracing and notification of possible contacts. As we heard at the CIFAR Roundtable, the South Korean government has undertaken comprehensive measures to this end, with authorities using all means at their disposal, including tools that analyze credit card transactions, CCTV recordings, and GPS data. The data streams created by mobile devices and apps present many opportunities for better understanding, modelling and controlling disease outbreaks here in Canada, too. Two such mobile apps under development, by CIFAR Fellows and Canada CIFAR AI Chairs Alán Aspuru-Guzik in Toronto and Yoshua Bengio in Montreal, propose peer-to-peer systems, with each person’s phone continuously updating infection risk based on proximity to infected individuals. With location tracking and daily prompts to record symptoms, the apps create a contact matrix, identifying possible points of contact. Both projects propose decentralized networks to protect privacy, where no single node has all of the information, but each device has a local AI model to accurately assess risk. Solutions such as these, and the data required to drive them, must be co-ordinated responsibly by scientists and policymakers to fight this pandemic, here in Canada and around the world. We must be clear about data ownership and the purposes for which the data can be used. We must respect and appropriately balance individual and collective interests. But to maintain barriers to the data that are needed by scientists and decision makers is to prolong and expand the tragedies that will be wrought by COVID-19. It is science that will put an end to the threat of COVID-19, and all citizens of the world have a right to its promise of relief. Dr. Alan Bernstein is president and CEO of CIFAR, a Canadian-based global charitable organization. Dr. Elissa Strome is interim vice-president of research at CIFAR and executive director of the Pan-Canadian AI Strategy.

To fight COVID-19, the world needs science, and science needs data

Beyond testing, we must also ramp up tracing and notification of possible contacts.

It’s time for Canada to be the global research leader of tomorrow

Here’s what’s still needed to drive innovation

Medical professionals’ discomfort with digital tech is stumbling block to spread of AI in health care, say experts

It’s time for Canada to be the global research leader of tomorrow

How polls help parties plan policy

Success of the Stephenville project is an important test case for Canada

Good corporate citizenship isn’t a matter of choice

International development: more important than ever for the public and private sectors

How to avoid a war with China

Deep fakes and big data: the next level of cyber warfare

Canada’s life sciences sector: its time has come

How AI can help Canada’s post-pandemic recovery

Why a robust health data infrastructure is so important

Contact-tracing app just ‘one layer’ of COVID response, says top doc, as Quebec shelves plans for rollout

News

Get free news updates

Opinion

Submit a guest opinion article

resources

Advertise

subscribe

get help

To fight COVID-19, the world needs science, and science needs data

Beyond testing, we must also ramp up tracing and notification of possible contacts.

Get The Weekend Point of View Newsletter