Abstract
Dr Braillon calls for greater accountability in acting on scientific evidence and on reports showing that the effect can be achieved within normal public health care. In fact, Finland was among the first countries in the world to establish organized population-based screening programmes for breast and colorectal cancer after the general evidence from trials was available. We were among the first to know whether screening for breast cancer works as a public health policy (particularistic evidence on decreased mortality from breast cancer) and this will probably also be the case in colorectal cancer screening. Public health policies should be based on both general and particularistic evidence.
There are no studies yet available showing that the effect from trials on colorectal cancer screening can be achieved within normal public health care. Some evidence to the contrary does exist, for example, the programme in France to which Dr Braillon refers. We have not yet seen reports with proper evaluation on the effect on mortality in France.
Dr Braillon is correct-several official bodies have made recommendations, but these are based on different interpretations of the available evidence and on different principles and values in relation to screening programmes. They are not uniform regarding methods of screening and the recommendations per se are neither evidence nor proof of effectiveness. There is no direct evidence at present of effectiveness in reducing mortality from colorectal cancer by routine screening with any modality, and only limited evidence on harms (including cost).
Dr Braillon is also correct that we had no intervention in the control arm. An intervention like health education in this arm would potentially leave open the question of effectiveness. If there were no difference in colorectal cancer mortality between the arms, one would not know whether this was due to similar effectiveness of the interventions or no effect of FOBT (and education). A design which cannot answer the question would be unethical.
We do not think the Helsinki declaration point 32 on inferior treatment applies here. FOBT is generally available and in routine use but promoting indiscriminate use may lead to problems. Limited resources and variation in healthcare policies mean that trial results may not always be reproducible, and there are always potential harms from screening. Screening may, for example, promote unfavourable lifestyle changes and these could have a bigger harmful health effect than the small benefit (lower mortality from colorectal cancer) from screening. In several trials the overall mortality was higher in the screening arm than the control arm. No-screening is therefore not an inferior intervention.
To our knowledge there are no organized screening programmes in the US for any cancer site. In much of Northern Europe, by contrast, the public sector provides health care and cancer screening programmes are national and cost-free for the attendee, and based on invitation and particularistic evidence on mortality. The threshold to launch such a programme is higher than the threshold to give a recommendation on indiscriminate screening. This difference explains some, but not all, of the ethical issues.
Comparing Europe with the US is not valid, and raises ethical problems. A public health policy should maintain equity of access, cover the population with more than a theoretical risk (age 49 is far too low for CRC screening), and should be evidence-based. Virtual colonoscopy (or any colonoscopy) has not been studied and the information on effectiveness and harms (like overdiagnosis of preinvasive lesions) is lacking.
The design for colorectal cancer screening programme in Finland with individual randomization was based on several factors: a new programme could not be immediately launched with 100% coverage (not logistically possible, lack of colonoscopy resources etc.). Random allocation of screenees and use of all the available resources maintains equity in the society because everybody has the same a priori chance to be screened, and if the programme is not effective or turns out to be harmful, the chance of not being screened is the same for everyone.
We designed the programme, given the available resources, to provide the correct answers to relevant questions as quickly as possible. The programme will show whether screening works and whether it should be continued. It can be modified or stopped if it is not effective or if it is harmful. The general evidence on the expected effect indicates a relatively small reduction in mortality. Most programmes cannot demonstrate such a marginal effect if not randomized at the implementation phase. This is also true in France, where it will probably never be possible to know whether to continue with screening, to modify the programme, or to stop screening.
We therefore believe that our programme maximizes the benefits and minimizes the harms and costs, can be properly evaluated, and is ethically justified.