Population-based screening for colorectal cancer with faecal occult blood test--do we really have enough evidence?

Population-based randomised controlled trials (RCT) have shown that invitation to biennial screening with faecal occult blood testing (FOBT) during 10 or more years reduced colorectal cancer-specific mortality. These results have stimulated plans to introduce mass screening in various countries; however, none of these trials has been able to show any reduction of total mortality, which should be expected with reduced disease-specific mortality in a RCT.

The aim of this review is to analyse, in more detail, the findings in these trials. The results of the trials have, in this review, been systematised and discussed in the context of potential bias, validity and effectiveness.

It is found that the reduced cancer-specific mortality is modest and that the clinical significance may be discussed. The number of persons needed to be invited for multiple screening rounds to avoid one death in colorectal cancer (CRC) is high, ranging from about 600 to 1,200. A remarkable finding is that only one fourth of the carcinomas in those invited to the screening were actually detected by this intervention. The absence of reduced total mortality in all series is a serious problem and evokes questions about the validity in determination of cause of death. None of these trials showed any effect on incidence of CRC by removal of precancerous adenomas.

It seems reasonable to conclude that the scientific evidence to support introduction of population-based screening programmes with FOBT appears not yet strong enough. In addition, harm/benefit and cost/effectiveness ratios are not well determined.