Data can change your life; it probably already has
What is data? How can it improve and save lives? And why does it matter during the COVID-19 crisis and beyond?
Mention the word “data” and some people’s eyes glaze over. They imagine endless numbers, statistics, characters, observations and details too big and complex to understand. It’s true; the essence of data are facts and figures; a set of qualitative and ….
Uh-oh. Are we losing you, too?
Let’s break it down to something more personal. In everyday life, data can help cities track vehicle and accident statistics, map alternate routes, and literally get you out of a jam. Data can help your doctor identify problems faster, and personalise more effective treatments for diseases like cancer. You can even get personalised eyeglasses, thanks to different eye tests and many data points.
When we combine the enormous volume of facts and figures called data together into a meaningful and understandable way, we can recognise patterns. Those patterns help us understand more about diseases and what causes them. The insights we gain can improve diagnosis, or help develop new medicines and prevent disease. They might help us pinpoint the next COVID-19 hotspot or predict the next pandemic – and even prevent it.
See? Data isn’t boring or meaningless at all! In fact, when you talk with Fanny Sie at Roche Canada, it’s difficult to be anything but excited about the potential of data and the ability of artificial intelligence, known as AI, to collect, curate and analyse data and turn it into powerful information that can improve and save people’s lives. Her group, along with others in an integrated, global approach across Roche, is harnessing the power of data to advance healthcare now and the future.
A powerful example of data
In October, Roche became the first company to deliver clinical trial results to the International COVID-19 Alliance and Workbench, a coordinated platform to help researchers access data and speed up development of treatments. Roche became a pioneer in sharing aggregated data from phase 3 trial patients with COVID-19 pneumonia, because during a pandemic, timely and transparent sharing of information – in the case of clinical studies data – is of the essence. “Sharing clinical trial data enables all COVID-19 clinicians and researchers to learn together in real time to benefit patients suffering now. This is why we’ve shared our clinical trial data,” says Michelle Rohrer, Global Head of Product Development Regulatory at Roche.
“There is a lot of confusion about what artificial intelligence is and what it can do, but at its essence, it’s a way to help us make decisions,” Fanny says. “There is so much information out there, in so many different formats, particularly now in regard to the pandemic. The problem is, for humans alone, it’s too much information. We can’t possibly analyse or understand it all manually. And that’s where artificial intelligence and machine learning come in. We can use the power of computers to sort and make sense of that information in a fraction of the time it would take a human to do it.”
Like a universal translator with superpowers
Roche believes the world can win the fight against COVID-19 by sharing knowledge and healthcare data to better inform patient care and health system decision making. Roche Canada assembled the Data Science Coalition, a group of like-minded public and private organisations with a common mission and vision to bring actionable COVID-19 intelligence to patients, frontline healthcare providers, institutions, supply chains and governments around the globe.
The Roche Data Science Coalition includes public and private organisations and groups across Roche, unified in a mission to collect, combine, organise and share COVID-19 information around the globe.
“Coming together from academia, private industry, multi-market, multidisciplinary and global focus in the Coalition, we ended up being 10 years faster than if we had to go through all of the contracts and provisions,” Fanny says. “The urgency was COVID: this was a common challenge, and we cannot solve it without each other and we have to do it quickly. It accelerated us into the future.”
In simplest terms, think of the Coalition as a universal translator with superpowers.
“Data scientists say ‘look; great data,’ Fanny says. “But other stakeholders, like governments, may look at that data and say ‘I do not get it. I do not understand what this graph says; how does this help to open up my economy?’ The Coalition is like a bridge that overcomes a unique hurdle of people who have different needs and see things from a different viewpoint, and it connects the two.”
The Coalition can quickly gather and process massive amounts of global information, then deliver it back in formats that are meaningful and useful in the native “language” of scientists, epidemiologists, health agencies, statisticians, governments and more.
There is a lot of confusion about what artificial intelligence is and what it can do, but at its essence, it’s a way to help us make decisions.”
“Some people are data scientists creating a solution, some are aggregating and curating data in a safe place, some people are all about analytics,” Fanny says. “And if people know where they live in the value chain, it becomes a perfect supply chain for data. And eventually we are translating data into healthcare solutions that impact people’s lives for the better.”
The power of partnering
Fanny says when the pandemic hit, there was an abundance of data available on the larger population, but virtually none at the individual patient level. Further, there were few ways to link the two data sets into a cohesive picture.
“We have so many amazing scientists around the world, but they did not have the chance to exercise what they are really good at,” Fanny says. “Now they see how they can use all of that knowledge to help inform government decisions, help hospitals predict and prepare, and be part of the solution. Also, it’s Roche coming to the plate and saying ‘we want to be part of the solution, too.’”
The Coalition teamed with Roche experts and asked politicians, hospitals and frontline workers around the globe: what questions can we help you answer that will help have an immediate impact? For example: why are people in a specific area not feeling well, and could this area become the next pandemic hotspot?
In addition to public data, the Coalition partnered with companies big and small to collect proprietary data in a safe, secure and unidentifiable way, to create a more comprehensive overview of what is happening worldwide.
“Early on, if you looked at developing markets, all you saw was the number of confirmed cases – but that doesn't tell you much,” says Yuvi Gill, who works in Global Access at Roche.
The teams developed a global report that helps governments and healthcare providers manage and prepare for patients, and identify community risk and plan ahead.
“The report tells you: here is exactly the number of people we feel are infected within the country,” Yuvi says. “Here is your testing rate. Are you testing too little? Are you testing a lot? Here are your confirmed cases. How does that compare with another country? We also bring in Google mobility data to see if people are still moving around, even under shelter-in-place orders. That can tell you how effective the shelter-in-place order is.”
Global data, local answers
As the world enters the second wave of the pandemic, the report helps governments monitor infection rates in near real time. As more and more data are added, the report provides greater insight for governments and health agencies on how to respond.
For example, the Coalition talked with officials in Pakistan, who were trying to determine the consequences of Ramadan in April, and whether they should ease lockdown restrictions. The benefit of the Coalition report? The Pakistani prime minister’s task force talked with industry and government, saw similarities between the data it already had bolstered by the Coalition data, and was able to deploy a strategy.
Similarly, the Roche team in Algeria partnered with the Minister of Pharmaceutical Industry early in the pandemic. Soon, they were asked to work with the National Institute of Public Health Wellness to show how data could help the Algerian government going forward.
“We had very open conversations about data, the need of having data and the need for sharing it,” says Khalil Kadaoui of Roche Algeria. “It’s a partnership that is fit for their purpose, and for the global community as well. It’s truly an example of co-creation with an institution, brought on by COVID and the need of having data. But it’s just the beginning.”
Fanny agrees, and says these new solutions will help shape emergency planning now and in the future.
“This collaborative approach of sharing data is a necessity for COVID, but it will be equally necessary in any other global emergency we may face,” Fanny says.
Roche Canada launched the Roche Artificial Intelligence Centre of Excellence in late 2020. It is the first collaborative centre to combine expertise of all three national AI institutes in Canada, with the goal of driving digital transformation in health to optimize and reduce the cost of healthcare delivery, improve health outcomes and respond quickly to opportunities and challenges.
Data can be life-saving! Learn more about the data tools available:
A central location for curated, publicly available data from sources across the globe. The scientific and research community, as well as patients, clinicians, infectious disease specialists, epidemiologists and data scientists can use this to accelerate research, evaluate solutions and share knowledge globally. It’s run on ThinkData Works’ Namara platform.
Calling all those who geek out on data for the greater good! The Data Science Coalition issued a challenge administered by Kaggle, an online community of 4.3 million data scientists and machine learners. The United Network for COVID Data Exploration and Research, or UNCOVER, is available to the public and focuses on healthcare capacity management, research, and research and capacity limitations.
A self-assessment tool for anyone who thinks they may have COVID-19 or has been diagnosed (being treated or recovered) with the disease. The tool runs through the free Health Storylines app. Patients report their data and it is anonymised and shared with the Data Science Coalition respositoritory, where it’s free to the research community. Patients can also print or display their self assessment and share it with their healthcare provider.
Population Level Analytics (PLAn)
This free tool, powered by Roche and run by provider Aetion, analyses real-world data to inform critical policy making and resource planning at the regional level in the US.
Join the Data Science Coalition
Like-minded organizations can contribute expertise, datasets and communication channels. The Coalition seeks epidemiological data such as population census data, morbidity and mortality records, medical records, insurance claims, billing records, travel records, surveys of individuals or households. Send an email.