Get best of the week

Self-isolation: how likely is it to get coronavirus if you go for a walk

Today, in Novosibirsk, self-isolation was introduced, and it became interesting for me to calculate the likelihood of getting coronavirus if a person went outside for a walk.

Source data from open sources:

  • The average number of days from infection to the onset of the first symptoms: 11.5 days.
  • The average number of days how many people spend “on their feet” after the first symptoms: 3 days.
  • Base reproductive number (this is how many people get infected from one infected person): 1.4-2.5 people.
  • The number of identified patients with coronavirus in Novosibirsk on 04/01/2020: 10 people.
  • The coefficient of undetected cases to the identified: 10.
  • The number of contacts per person per day is closer than 2 meters: 200 people (approximate estimate).

We will calculate how many days a sick person infects others: 11.5 days waiting for the first symptoms and another 3 days he suffers an illness on his feet. In total, in 14.5 days it will infect from 1.4 to 2.5 people (for further calculations, we take a rounded value of 3). During these days, he interacts with 2,900 (200 * 14.5) people. Thus, we can calculate the probability of infection if we meet a coronavirus patient: 0.1% (3/2900 * 100%).

Now we’ll calculate what is the probability of getting sick if we go to Novosibirsk for a walk these days. 1,500,000 people live in Novosibirsk. The number of undetected patients, according to doctors, is 5-10 times greater than those identified. It turns out that in the crowd in Novosibirsk the probability of meeting a patient is 0.007%((10 + 10 * 10) / 1500000 * 100%). We multiply the probabilities and get that if you went out for a walk in Novosibirsk, then according to the most pessimistic calculations, with a probability of 0.0007% . (0.007 * 0.1) you will become infected with a coronavirus. In conditions of self-isolation, this number will be even less.

CONCLUSION . It turns out that the probability of getting coronavirus on a walk for Novosibirsk is only 0.0007%. For Moscow, 0.018%. For Cyprus, 0.025%. For Italy, 0.2%.

For your city, you can calculate the probabilities yourself, the link is: Google Sheets

Immediately there are questions why such calculations are not published in the media, and is it not too early to introduce self-isolation in regions where the number of cases is in the tens?

UPD. I decided to check the prediction of this model on real data. Since the incubation period is an average of 11.5 days, in order to get the number of new cases for today, you need to calculate the number of people infected on March 22. The number of people infected that day in Russia was 367 and the probability of catching the virus was 0,0003% . We multiply the number of residents of the Russian Federation by the probability of infection. We receive 417 new patients in 11 days. We look at today's roundup of news - 440 new patients. It turns out a fairly accurate prognosis. But the model only works for cases without restricting the movement of people. In the case of quarantine, it must be adjusted.

More articles:


All Articles