They give two methodological nitpicks, which is not really "engaging the data directly". One is basically, there are other paradoxical results in with this paradoxical result, so obviously this paradoxical result can't be true! Really, even in the COPD patients they saw the same result, which means it's obviously wrong. Right? Right?
The other one is fair enough: self-reporting has its limits. But what are they really?
Let's do the thought experiment: if the converse result were being reported, i.e. non-smokers were having worse outcomes than expected, but growing evidence showed smokers developed serious cases more frequently than non-smokers, would that article be any different? I think you could just write it paint-by-numbers. Also, kids, don't start smoking no matter how cool you think it is, mkay?
The other one is fair enough: self-reporting has its limits. But what are they really?
Let's do the thought experiment: if the converse result were being reported, i.e. non-smokers were having worse outcomes than expected, but growing evidence showed smokers developed serious cases more frequently than non-smokers, would that article be any different? I think you could just write it paint-by-numbers. Also, kids, don't start smoking no matter how cool you think it is, mkay?