Statistics from Altmetric.com
Zorzi et al 1 are to be commended for their timely analysis of the impact of using a faecal immunochemical test (FIT) on colorectal cancer mortality in their screening programme. The 24% colorectal cancer mortality reduction they report will surely give impetus to efforts to introduce FIT into other screening programmes. As they rightly observe, the impact of these newer screening tests is likely only to be determined from observational studies such as their study rather than further randomised trials.
Nevertheless, some of their findings are quite surprising. It is remarkable that the mortality reduction they report should have been evident so early, within 5 years of starting to screen, when in the randomised trials a lower mortality from colorectal cancer was not apparent until over 5 years after screening started and for the reduction to be substantially greater in women than men when in these same trials the mortality reduction was similar.2–4
It is also remarkable that their colorectal cancer incidence, after peaking in the 2002–2004 roll-out period, should already have fallen below the prescreening incidence level when uptake of their screening programme was about 50% and only reached in 2006. While removal of advanced colorectal neoplasia might be expected to have some effect on incidence, it is difficult to conceive that such an effect would be clearly detectable within 5 years considering the fall in incidence of distal cancers in three of the flexible sigmoidoscopy trials was only just discernible at 5–6 years.5–7
As the authors recognise, ecological analyses are prone to ecological fallacies and confounding from other unrecognised differences between areas. In their analysis, is it possible that there are differences in colorectal incidence and mortality in the early-screening areas and late-screening areas affecting all age groups?
The findings in their analysis would be more convincing if the colorectal cancer mortality and incidence trends for the age groups not invited for screening in the early-screening areas had also been reported. Inclusion of these data would address the critical issue of the comparability of the early-screening and late-screening areas. If Zorzi et al can show that the decline in colorectal cancer mortality in the early-screening areas is only evident in the age group offered screening (60–74) and not in the older or younger age groups in these areas, this would surely accelerate efforts to introduce FIT-based screening elsewhere.
Contributors RFL and SPH both contributed to the drafting of this letter.
Competing interests None.
Provenance and peer review Not commissioned; internally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.