Understanding disease transmission through mathematics
Marc Brisson uses mathematical models to help change health policy on infectious diseases. Sound far-fetched? It may, but that’s exactly what the associate professor at Université Laval does.
“The goal of my research is to help inform policy decisions to reduce the burden of infectious diseases,” Brisson says. “I concentrate most of my efforts on sexually transmitted infections and vaccine-preventable diseases.”
To do that, he works with decision makers to build models that take into account transmission dynamics of infectious diseases to assess the effectiveness and cost-effectiveness of different interventions. Ultimately, that information helps those policy makers decide how best to optimize the prevention of infectious diseases.
And to get them that information, he uses mathematics.
“We build mathematical models that predict the spread of infectious diseases and then we apply interventions to see how much of the disease we can prevent,” he says. “To do this, we build complex models that often require super-computers.”
Those super-computers come from Compute Canada, the NGO responsible for Canada’s digital infrastructure.
“We model individuals within a whole population so we need a lot of computing power for our calculations,” Brisson says. “We create a fictitious population that represents the country we are modelling— let’s say Canada. Within this population, we model the behaviour of individuals and their healthcare use. If we’re modelling a sexually transmitted disease (STI), individuals in our model are given risk factors for STIs. For STIs, the main risk factor is sexual activity, so basically, we stratify our population to men and women, and according to their age, they are given specific sexual behaviours, such as when they start having sex and how many partners per year, based on Canadian statistics.”
Think of it is creating a mathematical soap opera: He models many individuals with their characteristics and their partnerships along with the partnerships’ formation and dissolution. And then he models how these STIs are transmitted between individuals over time. Take Human papillomavirus (HPV). He creates his models to reproduce HPV transmission and the incidence of its related diseases (cervical cancer, genital warts) in Canada without vaccination. With his models, he then predicts what happens with vaccination, using certain assumptions about who is vaccinated and how many are vaccinated.
“We make predictions on the impact of intervention in terms of reduction of infection and disease and the cost saved to the healthcare system, then we can look at whether it’s cost-effective to vaccinate.”
Basically, he said, what he does for governments is model vaccinations and then provide them with information about whether a vaccination program is cost-effective, and who should be vaccinated and at what age.
Policy makers formulate a plan based on many factors and criteria, including ethical considerations. As well, the vaccine has to be safe and effective and there has to be a burden of disease.
“What we do is really help them with getting a better idea of what would be the actual impact in terms of reduction of burden and whether it’s cost-effective,” Brisson says. “They’re model predictions so there’s always uncertainty. I never say that what we predict will happen in the next 10 to 30 years is reality, but we’re giving them a best estimate based on very extensive and complex methodology.”
Brisson started out working with chicken pox vaccinations, pertussis (whooping cough) and pneumococcal infections. Today, in addition to working on HPV, he and his team of 12, are working on c. difficile.
Brisson said it’s rewarding to be able to do work that guides policy makers into educated decisions.
“We can provide them with the results in ways that they understand and can incorporate into their decisions,” he says. “It’s always nice to see that they can base some of their decisions, at least partly, on our work.”
When the World Health Organization, for example, made recommendations for countries to go from three to two doses of HPV vaccine, Brisson’s team was heavily cited in the recommendation.
He said the work simply couldn’t be done without Compute Canada. “Basically, we wouldn’t be able to run our models without Compute Canada. If Compute Canada didn’t exist, we wouldn’t be able to do what we’re doing.”