CIOMS has recently made their revised version of these guidelines available on the net but it is unclear if they are willing to accept amendments in the light of comments from the international epidemiological community. This Rapid Response is intended to provide IEA members with the encouragement to discuss the documents with the CIOMS and with their local ethic committees. The provisional text is available at www.cioms.ch/080221feb_2008.pdf. Please contact CIOMS quickly if you have comments.
This revised version is much improved on earlier versions. The document details several important areas that are badly needed for epidemiologists:
Guidelines concerning ethical committee review of multicentre studies
Research on anonymous data or data that has been encrypted
Use of incentives for participation in studies
Internet data collection
Long-term prospective studies and renewal of consent to participate
Informing participants about study findings
The general statements CIOMS make in the documents are often good but the detail is missing in many areas. There is a clear recognition that scientific quality is an ethical issue but no good solution about how to deal with this in practice. Informed consent is acknowledged not necessarily requiring written informed consent but it is still unclear what the participants should be informed about. Concealing the true purpose of an investigation (e.g. in case control studies relying on recall of exposure information) appears to be unacceptable although such a practice may well render such studies invalid (such a recommendation was present in an older version of the CIOMS document). There is clear acceptance of the principle that ethical reviews must take into consideration the risk or lack of risk the study imposes. There is no acceptance of making it financially attractive for participants to take part in studies that involve no risk. For example, studies on semen samples usually recruit less than 30% of those invited making it impossible to set up a monitoring system that can capture subtle changes in semen quality over time although such data are crucial. Renewal of informed consent at regular intervals as a principle is commendable but no discussion is presented on how and how often this should be done, how the costs of such process will be met, or the impact on cohort attrition by failure to renew consents. Consents to whole genome studies with thousands, soon hundred of thousands, of SNPs on thousands of individuals leave many committees in the dark on how informed consent to this research should be given. Even if the information to participants may skip the hypotheses, or lack of hypotheses, and deal with the risk of disclosure of current and future knowledge many uncertainties remain. Among them is the question of how participants are to be informed about these results both now and when we may know more about the functional importance of some of their mutations? Local ethics committees will in vain look for more specific guidelines on how to address these difficult issues.
Large population based repositories are being collected at present in order to provide the necessary biomarkers for environmental exposures or genetic traits for future research in genetic as well as in environmental epidemiology. These data sources are usually based, properly, upon informed consent, but the CIOMS group raises the unwelcome question of whether consent must be renewed each time the repositories are used. This would be impractical, expensive, may invalidate the research and undermine the overall purpose of establishing the repository in the first place, wasting the time and involvement of the participants and the scientists. A solution would be for an ethics committee to decide on scientific use of the data of behalf of the participants. The CIOMS group is of the opinion that this will not be acceptable in many countries and avoids making a stand to guide national committees. The main problem is related to the concept of “being informed” in the informed consent principle. The CIOMS group has wrongly applied the principle that is intended to be concerned with risks of participation, which makes little sense in terms of hypotheses that have yet to be developed.
Ethics committees are now expected to take scientific quality into consideration if they follow the CIOMS guidelines but it is not clear how they can do that without considerably more expertise in epidemiologic methods. It will be insufficient to rely on evaluations by medical research funding agencies as more and more money is now allocated to individuals based upon their previous track record in order to make it possible for these people to work on large programmes of research rather than writing applications for specific projects. Further, quality is an ill-defined concept and “rules of thumb” may not work and may cause more harm than good. Much is at stake here since ethics committees have censuring privileges.
CIOMS fails to acknowledge that much of the time it is failure to conduct epidemiological research, or do it too late, that is the ethical problem. Research on exposures that may cause harm should be conducted without delay. In this respect it is good that CIOMS recognizes that research on outbreaks often cannot wait until the project can be approved by an ethics committee. CIOMS might have provided guidelines for when ethics committees should accept or reject proposals outright. Presently, much money and resources are wasted on waiting for ethical approvals.
CIOMS discipline epidemiology according to its subject matter and not as a set of methods. They include both clinical and public health epidemiology and experimental as well as non-experimental studies. The text would be clearer if the guidelines were concerned solely with non-experimental research and perhaps even non-invasive research that only possesses a risk to participants by virtue of disclosure of individually identifiable information to others. Research where the procedures involve a direct risk to those involved (and some observational epidemiologic studies fall into this category) should be dealt with in a separate paper. Did CIOMS view epidemiology as an observational science only and then later realize that they were wrong? The document appears to be wrestling with the consequences of this mistake.
The guidelines state that ethical reviews should be done independently of the people who suggested the research but not necessary independently of the research organization that is involved and they allow influence from community groups that may violate principles of scientific freedom and the right to criticize existing bodies of power based upon sound scientific documentation. This area seems to be very superficially thought through but may be of importance for people who do research in countries with weak democratic traditions. Research has an important role to play in providing independent investigations in how laws work, who they benefit and who draws the short straw in terms of health outcomes. It is not good idea to let those who have the responsibility for law and order be those who, in part, regulate the research.
In summary, these guidelines will probably allow epidemiology to continue as a scientific discipline that can, in most cases, provide information on determinants of health and disease by using population experiences. In some cases life may even be easier for the researchers, especially those doing multicentre studies, and this can be done without violating any of the ethical principles we all agree upon. The ethics review committees should not make it more difficult to do research of good quality since quality has an ethical dimension as the CIOMS committee rightly recognizes.
Jorn Olsen, Shah Ebrahim, Chitr Sitthi-amorn