Question about flu stats

[Sausagepants]

How does the WHO accurately measure (or estimate) the number of people affected by a virus?

I was looking yesterday at the stats for people infected by coronavirus vs swine flu (H1N1). I got swine flu back in 2010 when it was active. It was dreadful, and I knew almost immediately it was flu BUT I never bothered a doctor with it because I was young(ish), otherwise healthy and could manage the symptoms at home.

So was my case of flu never recorded, or would it have been estimated? Just trying to work out how they know the rate of infection if most people stay at home for a week and gradually get better without every consulting a medical professional. Most flu viruses are not notifiable illnesses.

The statistics would be generated from models (computer generated estimates based on known cases) in instances like influenza.

That said, if you didn’t go to the doctor then, you don’t actually know if you had H1N1. There are lots of strains of influenza, so no way to really know without testing on an individual level.

On the population level though, it’s done using statistical software and big fancy computers.

True, there was no way of knowing for certain it was H1N1, but it took out about 40% of the entire office in Canary Wharf at a time swine flu was making headlines, and we all had very similar symptoms (notably a weird intense sweating at night which lasted for about 2 weeks after most other symptoms had disappeared!).

So how do you model spread rate and know if you're accurate, if you don't have exact numbers? I mean, of course you can extrapolate if you know that one in every 5 people will catch it - but how do you arrive at the infection rate in the first place?