In June 2013 the European parliament will start the postponed voting on the New Regulation on Data Protection. More than 4000 amendments have been proposed and the Regulation is one of the most lobbied texts in the EU history. Not because of the problems it may cause to epidemiology, but because it regulates important conditions for transferring data of any kind within the EU. Thanks to a number of epidemiologists and epidemiologic societies, including the IEA, the commission is now aware that the text will impact severely on our research options. Whether they know it will also impact on democracy and the principles of the open society is more questionable.
We may hope that our working conditions for doing epidemiologic research will not get much worse, but we have no guarantees. All research related to health data still requires informed consent and the exemptions are currently defined in a way that allows for very strict interpretations. This interpretation will be influenced by an EU office on data protection, and the interpretation will probably try to achieve a balance between a Hungarian point of view versus how this has been administered in the Nordic countries. The present text can be interpreted in a way that will prohibit use of health data in research unless there is a serious national health threat (Hungary), or existing data can be used in all research that provide valuable public health information (in some Nordic countries). Research conditions will probably be better in some countries and worse in others. Unfortunately, the situation will most likely be worse in countries that have data.
Why have we ended up in a situation where the use of personal data in order to improve the health of people is considered a hazard rather than an opportunity not to be missed? One of the reasons is that personal data have been misused to harm selected population groups. However, this harm has been done by governments, not by researchers. Open access to data by governments and governmental institutions therefore requires strong democracies—no question about that.
The other reason is related to the Helsinki Declaration where informed consent plays a central role. What has not been taken into account is that the Helsinki Declaration addresses experimental research and was an appropriate reaction from the medical societies in the 1960s to avoid the unethical experiments that happened e.g. during the second world war on prisoners of war or on what the military power considered unwanted ethnic groups.
The principle of informed consent makes good sense in experimental research and in research that involves a risk to participants. However, re-use of existing research data or use of administrative data can easily be done without any risk to the participants, not even the risk of unwanted disclosure of personal data. The only risk is for those who have something to hide; politicians who may make laws that hurt people, private or public companies that produce something that causes harm etc. For that reason we do not see much epidemiologic research from totalitarian countries or from countries that do not care much about the health of the people.
Informed consent is also a natural component of collecting data de novo. If you do interviews you ask for permission. If you collect biological samples, including urine or hair, you ask for permission and you explain how data is going to be used and stored. Providing a specific informed consent may, however, not be possible in an explorative GWAS study where thousands of SNP data are collected. Or how do you get a specific, explicit informed consent in the large cohort studies where part of the data are collected without having a specific hypothesis in mind; where the purpose is to establish a data source for the future.
Re-use of research data is also much wanted, partly because we should make as good use of the data as possible, partly because it is a sound principle to invite other researchers to see if they can replicate your results after careful and critical scrutiny of your data.
Use of data from administrative registers has also been very useful. Seminal occupational health research carried out in the UK could not have been done without the access researchers had to death certificates. Much of what we know about the causes of cancers rests upon the use of data from cancer registers. A growing number of countries now establish an increasing number of valuable registers that can be used to identify social and environmental health hazards and to recognize suboptimal treatment procedures. Making it more difficult or impossible to use these valuable information systems is a waste of important resources and opportunities and will cost lives.
In many of the situations where existing data are used in research obtaining informed consent is not practicable or even possible. Furthermore it is not necessary because use of the data need not carry any risk. Anonymized data can be analyzed in well protected domains where a trusted third party keeps personal identifiers. Ethic committees should be able to provide access to data on behalf of the participants, and a data inspectorate can assure privacy is well protected. It will also be reasonable to set up a procedure for active opting out of research for those who insist on the right to be forgotten.
Informed consent, explicit or broad, has an important role in health research, but it is not a global requirement that can only be set aside when national security is at stake. Much useful research can be done on data we already have, and in many situations the ethical concerns for not addressing a public health or clinical problem are much larger than for doing the research. Time has come to re-visit the rationality of the global use of informed consent in research.
The Rapid Response Committee:
Jørn Olsen, Neil Pearce, Shah Ebrahim and Cesar Victora