What they should do is randomly test 500 (or ideally more) people. Just random people. Then that can be broken down into what percentage have the disease, and e

What they should do is randomly test 500 (or ideally more) people. Just random people. Then that can be broken down into what percentage have the disease, and each can be followed up to have their eventual symptoms, if any, studied, including death. That would give the most accurate picture. Anything else is going to be heavily skewed.