Could we not say this about ANY disease? Unless 100% of pop is tested?
This is why in statistics we use random sampling to predict actual results. Currently there is a very low % of the tested poplulation that returns a positive test. This gives us enough data samples to extrude accurate information.