Isn't the same bias present in ILI (influenza-like illness) statistics?
[1] https://arstechnica.com/science/2020/02/coronavirus-spread-i...
"The claim was quickly challenged by an infectious diseases expert who serves on a committee that advises the WHO’s health emergencies program.
Gary Kobinger, director of the Infectious Disease Research Center at Laval University in Quebec, said it would be highly unusual for there not to be mild or symptom-free cases that are being missed. He pointed to the fact that outbreaks have popped up in countries far from China — including Iran and Italy — because people with mild infections were not detected and traveled to other places.
'There are mild cases that are undetected. This is why it’s spreading. Otherwise it would not be spreading because we would know where those cases are and they would be contained and that would be the end of it,' said Kobinger, who insisted that mild, undetected infections cannot be ruled out until people who haven’t been diagnosed with the illness can be tested for antibodies to the virus.
'As long as we do not have good serology data, I think that it is completely speculative to say that there are no undetected cases,' Kobinger said."
What Dr. Aylward says makes no sense and reads more like a PR fluff piece designed to praise China to keep them happy, and as Kobinger points out he's probably wrong. From the same press conference he says this "If I had COVID-19, I’d want to be treated in China."
Which is obviously bullshit.
The 18% number is mostly useless because it's a textbook example of selection bias.