Concerns about privacy in research may be exaggerated

Editor—The risk of misusing electronic patient records has been pointed out,¹ and there is also justified professional concern about privacy in medical research. We have enrolled 1.14 million people in a register based cancer study in the seven counties of Sweden that were most contaminated after the Chernobyl accident.

According to the permit from the responsible authority (the Data Inspection Board), we had to inform the public in these counties because of the principle of informed consent. This presented practical problems, so an advertisement was published in the two largest national newspapers, which together have a daily circulation of 538 000 (the Swedish population is 8.8 million). In the advertisement we stated that anyone could contact us to check whether they were included in the study and what data were registered. This advertising was picked up by journalists, with interviews in both national and local television stations as well as radio stations.

Seventeen local newspapers in the seven counties, with a total circulation of 392 300 copies a day, published articles about the study. After this broad media attention for a study that had not even started we feared a large response from the public, especially people who did not want to take part. Interestingly, however, the few responses we got were from people who wanted to join the study. Only one person wanted to use his legal rights to obtain the information in his record, but he did not want to quit the study.

As newspapers, radio, and television are widespread in Sweden, most people in the seven counties studied are likely to have been informed about the study. The reason why we met no objections might be that the population believes the study to be important. It was obvious that they did not consider that the study was such a violation of privacy as is often implied by the scientific community or legislators.

Our experiences might be unusual, but they might also be applicable to other environmental (or even occupational) cohort studies, especially if based on registers. Interestingly, to our knowledge, not a single case of individual harm has occurred worldwide because of inclusion in an epidemiological study. Our experiences provide another argument when similar study proposals are considered by ethics committees or authorities that want to restrict the use of registers for epidemiological purposes.