[kgwgk]

> cannot be used to evaluate the effect of colonoscopies. Seems the study does provide data that allows this so, and shows clear differences in cancer rates and survival rates for those that did get colonoscopies.

Really?

You cannot simple compare the outcomes for the people who got an invitation AND got the procedure done with the outcomes for the people who didn’t get an invitation.

You need the compare them with the people who didn’t get an invitation BUT would have got the procedure done if they had been invited.