Sir

I read with interest and dismay the News article about the recommendations on informed consent and the ethical use of tissue samples to be issued by the US National Bioethics Advisory Commission2.

There is no doubt that the use of archival tissues for genetic epidemiological and pathological studies carries with it complicated ethical issues — for example, who should profit from the discoveries made using this material? One of the arguments put forward to restrict scientific access to these samples, however, is that the information creates job and insurance risks for the individuals or populations analysed. Clearly, job and insurance discrimination are potential dangers associated with this type of research and there are many reasons to argue that these should be avoided.

However, the approach of restricting scientific access to archived tissue samples is not the way to do it. In many ways this is like saying that we should prevent stores from selling sharp knives because they might be used to harm someone. Knives can be very useful or very dangerous, depending on the use. The recommendations and legislation should be aimed not at the scientists, but at the uses that we consider unethical. Many of us believe that insurance discrimination is one such area. Perhaps our efforts would be better spent considering the ethics of the insurance practices that will be affected by the new information.

Most scientists are using these archived samples to investigate disease pathogenesis. This knowledge can be used to develop better diagnostic and treatment strategies to help the individuals and populations afflicted by the disease. Indeed, the populations affected have the most to gain from the scientific use of their specimens.

The bottom line is that the scientific use of archived samples is good for all of society. This fact will not change no matter how many different communities are consulted. If this is considered unethical, maybe we should re-examine our ethics rather than hinder the progress of biomedical research.