Colorectal cancer screening: uptake of sigmoidoscopy Atkin et al. recently showed that a once-only sigmoidoscopy screening strategy can -in time -result in a colorectal cancer mortality reduction. 1 However, the design of this study did not allow a full interpretation of the actual uptake of sigmoidoscopy. Therefore, in the Journal of Medical Screening, a manuscript and, linked to that, an editorial were published about the uptake of sigmoidoscopy in colorectal cancer screening.2,3 The authors show an uptake of 45%, which might not seem very high, but actually is relatively high for screening with sigmoidoscopy. 4 -6 In the editorial the authors try to relate this uptake of sigmoidoscopy to the uptake of faecal occult blood testing (FOBT) with a guaiac base (gFOBT) in the United Kingdom and come to the conclusion that the uptake is at best only a little lower. However, some remarks should be made in reference to that comparison.First, some method of bias analysis should always be performed in comparative studies -and preferably the best available analysis. A limited bias analysis can be performed qualitatively by comparing both populations separately in their composition and presence of confounders (e.g. gender, age, deprivation). However, the level of scientific certainty of such a qualitative analysis is speculative at best. The best way to compare the uptake of two alternative tests is in a randomized controlled trial.Second, it is usual, and most informative, to compare with the best available other alternative, which in the case of colorectal cancer screening is an immunochemical FOBT. In the Netherlands a three-arm randomized comparison of gFOBT, immunochemical FOBT (FIT) and sigmoidoscopy, with 5,000 people invited in each arm, was performed.6 The (for ineligibility corrected) uptake percentages with confidence intervals (95% CI) were 50% (95% CI 48-51) for gFOBT, 62% (95% CI 60-63) for FIT and 32% (95% CI 31-34) for sigmoidoscopy.There is some debate about the performance of the sigmoidoscopy compared with the FIT where detection rates for colorectal cancer (and precancerous lesions) are concerned (at least in the subjects that do participate). However, the sigmoidoscopy screening is performed as a once-only strategy, where the FIT, similar to the UK gFOBT strategy, is usually repeated biennially, resulting in a cumulative sensitivity and the ability to detect incident cases as well as prevalent cases.
Leo GM van Rossum
Response to van RossumWe recently published results of a community-based study in which unselected men and women aged 58 or 59 years, living in the ethnically mixed London boroughs of Brent and Harrow, were invited to have flexible sigmoidoscopy (FS) screening performed by a nurse. 1 Uptake rates were around 45%, which we noted was similar to uptake of the guaiac faecal occult blood test (gFOBT) used in the English Bowel Cancer Screening Programme (BCSP) in the same areas of London. Van Rossum 2 questions the validity of our comparison of uptake rates since the data were not acquire...