My hypothesis is that having more information cannot result in a worse average outcome, given a rational response to the information (correctly accounting for relative risk). As a doctor, I do not have to recommend follow up scans if the cost is greater than the benefit, but I should still seek to use the information from those once they've been done.
I suggest that the above holds true even if there is some known chance that the information is junk.
To put it more simply, early intervention is always better.
Rational response to shadows on the lung in a CT scan is more scanning and more testing. The results of the further tests carry a risk. The benefits are zero for anyone without a disease, and low for anyone with a disease. That's a worse than average outcome for almost everyone taking those extra un-needed tests.
> No, early intervention is not always better.
There are many men who've had traumatic treatment for prostate cancer because of the results of early intervention style treatments. Many of those men would have died with, not of, prostate cancer. Thus, their lives have suffered because of a rational response to an early intervention.
But the "risk" only exists because of the CT scan, which is returning junk data.
Because a doctor has seen the junk scan, has seen the shadows on the junk scan, they now have to recommend follow up scans to rule out any disease. That subjects the patient to risk - hospitals are not safe places to be in. (Hospital acquired infections kill many people! Clinician errors kill many people! Traffic accidents getting to hospital kill many people!) Just being worried about the tests is going to cause a deterioration in most people's lives. And there's zero benefit for almost all of those people.
Early intervention is not always better.