Some might find the number of coronavirus cases coming out of China weird, and I’ve heard the accusation of possible “central fabrication.” In case you have similar doubts, I’ll try to provide more information for you to decide if this argument is justified.

Below is a figure of reported cases in mainland China in the past two months. It doesn’t look like a bell-shaped curve as expected, but it’s too soon to jump to the conclusion of data fabrication. First, the column in the middle that stands out is an artifact due to a change in diagnostic criteria. Since February 13th, the diagnostic criterion for confirmed cases no longer requires positive PCR test results and can be based purely on clinical symptoms and CT images. The lower criterion greatly reduces the chance of false negatives, but at the same time it introduces more false positives that would be corrected later.  Besides, if this curve doesn’t look natural, you might want to factor in those surging travel activities both ways around the Spring Festival, strict lockdown even in low-risk regions, and the once collapsed medical system in Wuhan and Hubei Province. The curve can’t possibly be perfect with these interventions. Lastly, as it applies to all countries right now, no data is accurate, and it’s limited by the availability of testing kits. If you have doubts in China’s data, remain skeptical in all other countries’ as well. Continue reading →