'Second wave will have more infections but same number of deaths': expert

22 November 2020 - 13:48
Professor Nico Orce, who says no model can predict exactly how many people will die.
Professor Nico Orce, who says no model can predict exactly how many people will die.
Image: Supplied

SA’s second wave of Covid-19 is likely to produce many more infections than the first, but an equal number of deaths — about 25,000 — as the fatality rate drops.

That wave will probably start in February and peak in April and May, according to a new model published in the international journal Applied Mathematical Modelling.

Among its three authors is professor Nico Orce from the physics and astronomy department at the University of the Western Cape. This week he told the Sunday Times: “The first indication is that we may be expecting a mirror wave in terms of deaths but with infections being much higher.

“Our model infers that the rate of deaths is lower in the second wave because most people recover or become asymptomatic with time.”

Those remaining also benefit from the immunity that has been acquired in the community.

Orce said the new model showed a universal pattern: the virus hibernates in the hot summer then re-emerges as a second wave that infects even more people as countries open up and everyone relaxes protocols while “trying to go back to normal life”.

This unfolded in Europe, and is coming our way, according to Orce and his fellow researchers. But “the virus will eventually lose”, he said.

Orce worked with researchers from the University of Granada in Spain and the University of Lyon in France. Their paper is one of only a few related to Covid-19 to be published in the Applied Mathematical Modelling journal.

The more traditional mathematical model for the spread of infectious diseases, developed about a century ago, is the SIR model, which involves differential equations to calculate the number of people who are susceptible, those who are infected, and those who recover or die.

“It uses four dynamic equations that describe the pandemic evolution and spread. What we did was find a solution for this system of four and simplify it into one,” said Orce.

To achieve that, the new model is based on a “drastic assumption” that nobody is spared from the virus — everyone gets it.

“That allows us to solve the system of four equations, and by solving it analytically, we can look at the upcoming data from the government every day,” said Orce.

Their model can predict “the evolution of a pandemic at its different stages” and can then be extended to factor in “additional spatial-time effects such as the release of lockdown measures”.

It suggests that most people become infected but are asymptomatic and that there is a “common pandemic evolution” across the globe, with the virus behaving the same way in every country.

According to Orce, the model gave the same mathematical answer as an extended SIR model that they also developed, and that is “very surprising because it’s very hard to put these two approximations together — one that includes recoveries and one that doesn’t”.

He said that can only be explained by the fact that the vast majority of people got infected but remained asymptomatic.

When SA went to lockdown level 1 on September 21, it took between 15 to 20 days later for infected people to recover, or die.

“Surges in the Western Cape and Gauteng were brought on by level 1, but those only manifest 20 days later,” said Orce.

He cautioned, however, that “no model can predict exactly how many people will die because they depend on too many variables”.

Instead, “we have to be scientific and look at what is happening in Europe, which is ahead of us. In physics, we look for symmetries to make problems easier, and when we see symmetry, we say it looks like a pattern that can repeat itself.”

Based on that, SA is in for a “tough time” that will peak in the winter months.

On the brighter side, Orce said he and the team were also “converting something very grim into something hopeful”.


• 57.2m - Covid-19 cases worldwide by Friday

• 1.36m - Global death toll by Friday

He and a larger team of local and international researchers have developed user-friendly computational programs, Android apps and other science tools that use real data from the pandemic for outreach and education.

Orce said the programme encompassed epidemiology, modelling, computation, maths, biology and physics. Components of it are already being implemented at the universities of Pretoria, Western Cape, Nelson Mandela and Zululand.

“We are taking the grim subject of Covid-19 to create learning opportunities that can speed up transformation in STEM (science, tech, engineering and maths),” said Orce.

This article was corrected to reflect changes in the headline, the first headline was incorrect.

Would you like to comment on this article or view other readers' comments? Register (it’s quick and free) or sign in now.

Speech Bubbles

Please read our Comment Policy before commenting.