> If you want to find an attenuated strain you have to look in people with mild symptoms

Shouldn't we be looking at people with no symptoms? I know this is not ideal as we're likely only testing symptomatic people.