Report in English with a Dutch summary (KCE reports 45A)

KCE reports vol.45 Screening for Colorectal Cancer 83
colonoscopic examination. Hospital files and the files of The Norwegian Cancer
Registry were searched to register any cases of CRC in the period 1983 - 1996.
In the first round (1983), 324 (81% of intervention group) individuals attended
endoscopic screening and 451 (71% of total group) in 1996. From 1983 to 1996,
altogether 10 individuals in the control group and 2 in the screening group were
registered to have developed CRC (RR 0,2 - 95% CI: 0,03 - 0,95; P = 0,02).
Strikingly, a higher overall mortality was observed in the screening group, with
55 (14%) deaths, compared with 35 (9%) in the control group (RR 1,57 - 95%
CI: 1,03 - 2,4; P = 0,03). However, before drawing possible conclusions on this,
the possible effect of screening on overall mortality should be addressed in
larger studies. Currently, a few larger trials are underway, but those are not
expected to report mortality results in the near future.
5.5.3.2 UK FS Screening Trial
In the UK FS Screening Trial 279 170.432 men and women aged 60 to 64 in
fourteen centers were sent a questionnaire by mail to ask if they would attend
for FS screening if invited. Of 354.262 people to whom this questionnaire was
sent, 194.726 (55%) agreed to participate. Interested respondents were
excluded if they informed the local trial unit of exclusion criteria missed by their
general practitioner, or if they had a strong family history of colorectal cancer
(at least two affected close relatives), a temporary health problem that would
prevent them from having the screening test, or a worrying bowel symptom
that required investigation. Individuals with a strong family history of bowel
cancer or suspicious symptoms were managed outside of the trial, because
randomisation would not have been in their interest. Finally, 170.432 eligible
subjects were randomized using a 2:1 ratio of controls (N = 113.178) to those
invited for screening (N= 57.254). The screening protocol involved a FS with
removal of all small polyps seen at the time of sigmoidoscopy with colonoscopy
reserved for those with high-risk polyps (three or more adenomas, an adenoma
greater than 1 cm in diameter, a villous or severely dysplastic adenoma) or
invasive cancers. Of the 57.254 individuals invited for screening 40.674 (71%)
attended. The attendance rate was higher in men than in women (20.519 of
28.097 (73%) vs. 20.155 of 29.157 (69%, p< 0·001). However, the men and
women who attended for screening showed similar age distributions:
proportions aged over 60 years: men 8.976 of 20.519 (44%), women 8.839 of
20.155 (44%). Of the 16.580 who did not attend, 7.541 (46%) provided a reason
to the unit: 3.324 no longer wanted the test, 547 said they had had a similar test
already, 794 were undergoing hospital treatment or awaiting an appointment,
265 had moved away, 97 had died, and 2.514 provided various other reasons.
It should be recognised that this study is essentially a volunteer study. The trial
used a two-stage recruitment procedure whereby eligible participants were
enrolled only if they responded positively to a questionnaire asking whether
they would be likely to accept the offer of screening. 55% of questioned people
responded positively, and 71% of those invited for screening (all of whom had
replied positively) actually attended. Therefore, as the researchers state, the
population coverage achieved was equivalent to 39%.
In the screening group, 2.131 (5,2%) were classified as high-risk and referred
straight to colonoscopy; of these 165 for reasons other than high-risk polyps
(safety of polypectomy: 31; family history of cancer: 20; suspicious symptoms:
16). 38.525 with no polyps or only low-risk polyps detected were discharged
after screening FS. Distal adenomas were detected in 4.931 (12%) and distal
cancer in 131 (0,3%). Proximal adenomas were detected in 386 (18%) of those
undergoing colonoscopy and proximal cancer in nine cases (0,4%). Of particular