Bogus information? Have you looked at graphs of "deaths" and "cases" for different countries, from February? Do this to 10 countries at random (largish countries, so exclude tiny countries, and exclude China).

See how well the two graphs correlate.

One would expect that if tests are good indication of sickness, then there should be good correlation between the two graphs (albeit with some weeks delay between the two). That would be what one would expect.

Pick 10 countries at random and see for yourself.

What threshold? First, I would discard "cases" that are clearly bogus, and look at deaths. What we care about are people actually getting sick. Really sick. You could use some measure of sickness, but thats too subjective, so "deaths" is a good number to use".