Alumna Marie Betsy Varughese (’17 PhD) is a senior modeller with Alberta Health, at the centre of the race to understand the spread of COVID-19 in Alberta. Photo credit: John Ulan
For most people, time became a blur during the early days of the global pandemic, but for Marie Betsy Varughese (’17 PhD), the weeks took on a pattern as she worked tirelessly alongside a talented interdisciplinary team at Alberta Health to understand and prevent the spread of COVID-19 across Alberta.
Every Monday, Varughese delivers the new case data and the top priority questions and concerns to the research team in the Department of Mathematics and Statistics at the University of Alberta.
On Tuesday, the research team gathers via Google Hangouts to discuss the week’s objectives and methods. From here, Varughese and her fellow scientists go their separate ways, each responsible for a specific task.
On Wednesday, the team congregates again to discuss and interpret results.
On Thursday, Varughese summarizes the results and reports back to her team at Alberta Health.
And by early Friday morning, the final results are prepared along with other epidemiological evidence for communication to health officials, including the office of Alberta’s Chief Medical Officer of Health. And so it has been since early March 2020.
The data in question is something that every Albertan—indeed, nearly every human on the planet—has watched closely since the COVID-19 pandemic brought the world to its knees, causing daily life to grind to a halt across the country.
But for Varughese, whose PhD studies focused on modelling the spread of tuberculosis in Indigenous communities, building a mathematical model to understand disease is just another day in the office.
Vital statistic(ian)
Varughese, now a senior modeller with Alberta Health, came to the University of Alberta’s Faculty of Science after finishing her undergraduate degree in physical science and master’s studies in epidemiology at the University of Guelph in Ontario.
Her area of focus? Mathematical biology.
“To be honest, I’d never heard of this field before,” explains Varughese. “I was blown away that there was this other way of understanding diseases and disease transmission that I’d never even considered, using calculus and mathematical principles. I looked at various programs across Canada and supervisors with expertise in this area of study. And one of those was the University of Alberta—specifically with Michael Li in the Department of Mathematical and Statistical Sciences.”
Upon her arrival in Edmonton, Varughese enrolled in a master’s program to get the foundational skills she needed in mathematics and statistics. But seeing her quickly progress, her supervisor suggested that she apply for an interdisciplinary PhD program with her home in the Department of Mathematical and Statistical Sciences and a secondary focus in the School of Public Health.
“The choice of an interdisciplinary PhD was a natural one for Betsy, though the program was quite new at the time,” says Li ('93 PhD). “I was inspired by her determination, and I worked with my colleague Dr. Richard Long in the Faculty of Medicine & Dentistry to create a PhD project on using mathematical modelling to study the transmission of tuberculosis within Indigenous communities in northern Alberta and Saskatchewan. Betsy was thrilled by the research project, since Indigenous health is one of her passions. As a result, she had to work twice as hard as other graduate students, because she had to fulfil core requirements from both departments.”
Varughese’s program centred on two key areas: mathematical epidemiology and the social and demographic elements of health, allowing her to build her expertise in the fundamentals of calculus and statistics while honing her understanding of the practical, human elements of health and disease.
“This program and its emphasis on mathematics and statistics helped me understand the underpinnings of where some fundamental concepts in epidemiology came from,” she says. “It was so valuable for me. I was so happy that I ended up doing this. It really gave me a chance to understand the fundamentals, the methods, and the context in which we examine diseases and disease modelling.”
Building a model
So what exactly is the link between math and disease?
“Imagine that you have a pail of water,” explains Varughese. “Water is being poured into the pail, but there is also a small hole at the bottom of the pail through which water is running out. Those types of changes—how much water is entering the pail versus water being lost and how quickly—can be captured using these principles. That is how diseases fit into math.”
This type of model, known as a deterministic mathematical model, helps scientists understand what is happening with a disease in a larger population by understanding the rate of infection, recovery, and susceptibility in the population.
“As time progresses, we see new infections and people recovering, all at the same time,” says Varughese. “This is why understanding how variables change over time is so important—which is a fundamental component of calculus. How does each factor affect one another? For instance, as people are getting sick from infected people, some infected people are also recovering and are no longer able to transmit infection. The rates where people get sick could be different from rates of people getting better. Calculus in particular looks at changes of a variable in time and it is possible to also look at multiple variables changing over time. This is one of the links between mathematics and diseases.”
During her PhD studies, Varughese joined Alberta Health part time as a senior epidemiologist. And since graduating in 2017, she continued in that role before becoming a senior modeller. Throughout her time working for the Alberta government, Varughese maintained a strong partnership with her former supervisor Li and his research lab, including graduate students Donglin Han (’19 BSc) and Weston Roda (’16 MSc).
Each year, the research team works together both informally and formally to model influenza in Alberta. So when COVID-19 began to become increasingly serious in Wuhan, China, working together to develop a model to understand this new disease felt like a natural progression for the Alberta-based team.
Enter COVID-19
“Michael started to look at modelling of COVID-19 in January using data from Wuhan, China,” explains Varughese, noting that this kind of foresight is typical in her former supervisor. “We used a simpler model called a Susceptible Infectious Recovered (SIR) model to examine what was happening with the novel coronavirus and to see if we could predict what would happen with the disease.
“I was excited to help out with this project in January outside of my work hours, as a volunteer researcher. Michael was the one that suggested this project, and I am so glad that he did. It really did make us more prepared when we were doing the COVID-19 modelling in Alberta.”
When Canada began to see its first cases of COVID-19, the work that Varughese was doing informally with Li’s lab became a critical part of her daily work with Alberta Health—from answering questions and responding to data requests to analyzing numbers and contributing information toward briefing notes or presentations for Dr. Deena Hinshaw (’97 BSc, ’04 MD, ’08 MPH), Alberta's chief medical officer of health and a UAlberta alumna.
“The Analytics and Performance Reporting Branch that I am so fortunate to be a part of is made up of many different disciplines, from analysts to epidemiologists to health economists to policy-makers. Modelling is just one part of the story and one part of the information that is provided to people like Dr. Hinshaw,” Varughese explains.
And while modelling isn’t a crystal ball, it does provide important insight for policy-makers and everyday people alike as they learn about and make decisions based on public health orders. “Math modelling is also important for analyzing interventions—such as physical distancing or mask usage—before implementing them out on an actual population of people,” says Varughese. “And the ability to simulate different scenarios can have huge cost savings. Using these models, we’re able to identify potential outcomes from different interventions.”
Looking forward
Examining these potential outcomes along with epidemiological evidence has allowed Alberta to build a relaunch strategy for citizens to follow as we learn to live in the new normal, in a world with COVID-19. It will also help as both policy-makers and scientists prepare for the potential of a second wave of this deadly disease.
“Our next modelling priority is to prepare for the potential second wave in the fall,” explains Li. “To better prepare for it, we need to have a deeper understanding of the risk factors that contributed to the much larger epidemic in Calgary, as well as to many outbreaks at long-term care facilities and meat-packing plants. Betsy has been and continues to be an integral part of the COVID-19 modelling effort every step of the way, and she has played a leadership role. I am very proud of her.”