When is it safe to ease COVID-19 restrictions? How does that relate to the “peak” of the epidemic? And when might restrictions need to be reimposed? Here’s a metric that can help answer those questions.

COVID-19 is highly contagious. Government restrictions to impose physical distancing such as business and school closures have curbed its contagion and new cases are down from their peak across the country.
Provinces across Canada have started easing those restrictions, although each in its own way and its own pace. Some have also already said restrictions may need to be reimposed if the outbreak worsens.
But when is it really safe to ease those restrictions? How does that relate to the peak of the epidemic? And when might restrictions need to be reimposed?
Tracking the diseases changing contagion is a key, researchers say. Heres how its done and what that means.
The contagiousness of diseases is represented by a seemingly simple number: the number of other people a single infected person infects. This is known as the reproduction number, commonly abbreviated with the letter R.
The basic reproduction number, R0, pronounced R-naught, where naught means subscript zero, is the fundamental infectiousness of a new disease, when no one has any immunity and no interventions have been imposed to curb its spread.
What R0 means
This diagram illustrates a disease with an R0 of two as it spreads from an initial infection through four generations. Each dot represents an infected person.
PlayPauseRestart simulation
What R0 means
Generation 2
Generation 1
Generation 3
Generation 4
For COVID-19, the R0 averages around 2.6 to 2.7 based on data from China and South Korea, researchers from the Centre for Evidence-Based Medicine at the University of Oxford report. That means in the absence of interventions, the average infected person gave the disease to between two and three people. That makes COVID-19 about twice as contagious as the flu, more contagious than Ebola, only half as contagious as smallpox and a lot less contagious than measles.
R0 and contagiousness
These diagrams show how quickly each of five viruses spreads over four generations, depending on its R0, which represents how contagious it is.
Gen. 11st generationGen. 22nd generationGen. 33rd generationGen. 44th generation
R0 = 2.6
1 total infection
(Spanish flu)
R0 = 1.8
1 total infection
R0 = 2
1 total infection
R0 = 6
1 total infection
R0 = 15
1 total infection
R0 and contagiousness for smallpox
R0 = 15
1 total infection
R0 and contagiousness for measles
David Fisman, a professor of epidemiology at the University of Toronto, says R0 tells us four things:
How herd immunity works
The diagram shows that if R0 = two, then once half the population is immune, there arent enough susceptible people left for the infection to increase in the population.
PlayPauseRestart simulation
R depends on the:
So any intervention that changes any of those things can change R a vaccine or increased immunity in the population, a therapy that speeds recovery or measures such as physical distancing that reduce the contact rate between people. Fisman and his colleague Ashleigh Tuite, also an epidemiology professor at the University of Toronto, have posted an interactive tool that shows how that works.
Beyond the initial pandemic, R is referred to as the effective reproduction number, Re or Rt (where t represents a point in time).
By now, weve heard that if youre tracking cases, hospitalizations or deaths during an epidemic, they roughly follow a bell-shaped curve. The more people infected by a single person, the steeper the curve will climb. If each person infects more than one person even just 1.1 people the epidemic will keep growing exponentially.
At the point where new cases stop increasing, R is one each infected person infects only one other person. If R remains one, we get a plateau.
But if R falls below one, the curve will start to go downhill and eventually reach zero.
When R is below or above 1
These diagrams show what happens if R is less than one, equal to one or more than one respectively, the number of new cases declines exponentially, stays the same or increases exponentially.
PlayPauseRestart simulation
Very important, say both Fisman and Jasmina Panovska-Griffifths, a senior research fellow at University College London who specializes in mathematical modelling of diseases.
If R is below one, Fisman said, the disease is at a slow burn without epidemic spread.
That suggests it may be safe for governments to ease restrictions.
However, because R fluctuates and is based on case and hospitalization data that may not be accurate, its a good guide but isnt very precise, Panovska-Griffifths said.
For example, in the U.K., during the first week of May, it was estimated to be between 0.6 and 0.9. Thats a big difference, Panovska-Griffifths said. It also may not take into account hot spots of infection somewhere in the system.
Because of that, she recommends easing measures slowly and keeping an eye on R, along with related indicators such as hospitalizations and deaths in conjunction with widespread testing and contact tracing.
Fisman recommends waiting until R remains below one for a couple of weeks before loosening things up.
In the context of reopening, we are going to hopefully keep Re at or below one because then we don’t have exponential epidemic growth, he said.
He added that if R goes back above one after easing restrictions, governments may need to reimpose measures that increase physical distancing.
I think its almost inevitable that we will need to go through the winter, he said during an interview on CBCs Metro Morning in late May.
He warned that while Ontarios R was previously below one, as of May 25, it was 1.1. That means the diseases spread has, at least temporarily, returned to growing exponentially if slowly and that could represent a setback.
Panovska-Griffifth notes that if R does climb above one, that could lead to a second wave of infection (although thats not whats happening right now, as the World Health Organization says the first wave isnt yet over).
Historically speaking, she said, the second pandemic wave in all the existing epidemics today has been more severe than the first.
How fast the epidemic declines
The rate of decline depends on R. The further below one it is, the more quickly the number of new cases declines.
PlayPauseRestart simulationHow fast the epidemic declines
How fast the epidemic declines
Panovska-Griffifths emphasized that because of the uncertainties in R, its important to track other metrics such as hospitalizations and deaths and compare them.
She said that government physical distancing measures need to be lifted slowly, but even then, we could be getting social bubbles where we have infections.
Because of that, she said another important component is widespread testing and making sure contacts of those who test positive are quickly traced and isolated to prevent further spread of the epidemic.
It’s only once we have an effective contact tracing and isolation strategy in combination with easing the lockdown [that] you can possibly get to the end, said Panovska-Griffifths.