[sfvegandude]

Interesting to note:

> Variations in medical care among states may also result in differences in cancer rates. In states where higher percentages of people participate in cancer screenings, more cancers are diagnosed early when the prognosis is often better.

It seems like this data could be impacted such that better healthcare would result in worse cancer rate, since this study naturally counts know cancer rates. That is, it would not necessarily result in literally more incidents of cancer, just the rate as measured.

[nequo]

I would think that if lung cancer goes undiagnosed, it will eventually kill the patient, at which point it would be recorded as a lung cancer case. So I wouldn't expect cancer screening to bias measurement.

Would people who die from metastasized lung cancer not be recorded as having had lung cancer?

[adanto6840]

How often is an autopsy performed?

I don't know the answer but my guess is that the majority of deaths do not have an autopsy performed, especially in 65+ age range.

Edit -- I looked it up, sounds like ~7% of US deaths have an autopsy currently, so I'd guess that the data is influenced by healthcare quality (I didn't read it, perhaps they try to control for that?).