[hannob]

They're doing an intention to treat analysis, which can be a bit confusing, but statistically makes sense and is the gold standard for clinical trials.

You're basicall saying "our randomization at the beginning of the trial is key to avoid biases, so we can't reassign people from the treatment group to the other group, even if they practically don't get the treatment". The reason is if you allow people to switch, your assignment is no longer random. People who avoid the treatment may have different health properties than the ones who don't.

In essence, you need your trial to be robust and large enough that a few people not getting the treatment don't matter.

[hef19898]

Then the title is wrong, since colonoscopies actually do increase survival rates (everything would have surprised me a lot as someone who was saved by one), but rather that invitations for colonoscopies don't work. No idea why, but maybe people willing to do one go anyway while those unwilling are very unlikely to follow an invitation. Either way, the title is wrong, misleading and very clickbaity.

[elcomet]

What you're saying is not totally correct. If you look only at patients that did get a colonoscopy, you cannot be sure that it is the colonoscopy that helps, or another variable (those people might monitor their health closer). Your assignment is not random any more so you cannot make any causal statement.

[bryan0]

Thanks. That explanation on avoiding selection bias makes sense, but it seems like the study is saying those who choose to get a colonoscopy (whether invited or not) are 50% less likely to die of colon cancer.

Also I’m not sure if the article mentioned this, but the data seems to imply that the % of people who opted to get a colonoscopy was similar in the invited and control group.

[akvadrako]

That suggests something strange though – that those who were invited but non-tested were more likely to die of colon cancer than non-invitie non-testers.

[zerocrates]

I suppose that by refusing to get a colonoscopy even when specifically prompted to do so, you're maybe sorting yourself into a more "unhealthy" group than the general population of non-colonoscopy-havers.

[robertlagrant]

That would make the outcome less surprising.

[bluGill]

It would be unethical to refuse to invite someone with a family history of colon cancer. Thus I think it is perfectly reasonable to assume there is something different.