At a time when the government of Quebec is preparing to reveal different projections of the spread of the coronavirus in the province, dr. Karl Weiss reminds us that they are not accurate forecasts.
“These mathematical equations are not robust enough to take into account the human complexity”, he says in an interview with Mario Dumont in LCN, on Tuesday morning.
In the case of the coronavirus, the models must take into account a host of variables that it is difficult to identify. For example, compliance with social distancing, the rate of hand washing in the society, the wearing of mask recommended or required, and the age of the population should all be considered.
“There are so many parameters in the equation that it becomes difficult,” explains Dr. Weiss.
Margin of error
The doctor considers that it would not be surprising that the projections are not accurate in the next few weeks, which would be quite normal.
“The statistics, it’s a bit like polls. These are forecasts, and you have all chances to be wrong on some aspects”, says he.
In addition, the projections serve first and foremost as a model. “It’s totally alright for a government to make statistical studies for projections. If it is said that the peak is April 19, and that the peak arrives on the 26th of April, it’s not worth it to say that they have poorly done their job.”
The publication of these projections is even more important for a question of transparency, considers the guest.
After the summit, a plateau
What can we expect of the projections that will soon be made public? For dr. Weiss, it will be seen that the ‘peak’ of which we have often talked about will, in fact, a “summit”, and that it will be followed by a plateau. “It is a mountain where you go for a walk on the summit plateau for a while, before going down again” shows there.
It is therefore possible to say that the number of cases, hospitalizations and deaths is expected to stabilize. The decline of the curve would not be completed instantly.