Commentary: ethics of clinical research without patients' consent

BMJ 1996; 312 doi: (Published 30 March 1996) Cite this as: BMJ 1996;312:818
  1. Claire Foster, research fellowa
  1. a Centre of Medical Law and Ethics, King's College, London WC2R 2LS

    What makes a research project ethical? A series of criteria need to be satisfied. Difficulties arise when only some of them can be satisfied or, as in the case of this trial, when two of them are mutually exclusive. The criteria are that the research is aimed at obtaining a desirable end, that the research subjects will not be harmed by being in the trial, and that the research subjects are consulted before being enrolled into the research.

    Desirable aim

    In this research project the aim was towards a result that would be widely recognised as desirable. There will be no discussion or dispute of this claim here, and we will assume that the first criterion for ethical acceptability is fulfilled by this research project.

    Duty of care

    A doctor running a trial owes a duty of care to his research subjects. Such a duty means at least not harming the subjects and, in a therapeutic trial such as this one, should mean benefiting them. It is important to appreciate that in a trial of new treatments the fact that one treatment turns out to be better than another does not mean that the researcher has failed in his duty to benefit his subjects. At the time of starting the trial it was not known which treatment was superior. Indeed, this is the very justification for the research. This means that, for the patients who are the subjects of the trial, the benefits from treatment in the trial are to be at least as good as benefits received from standard treatment outside the trial. Thus, the duty of care was honoured in this trial, and on two counts the trial may be considered to be ethical.

    Subjects' consent

    For human research to be ethical, researchers must ensure that the people who are used for the purposes of research are happy so to be used. It is a matter of respecting people as ends in themselves and not merely means to other people's ends. This requirement is generally answered by seeking the consent of the subjects before enrolling them in a trial. Some might argue that, since the research would not harm and might benefit the subjects, it is not necessary to seek their consent, just as some might argue that in ordinary treatment it is not necessary to obtain consent—on the grounds that it is more important that benefit is offered than that autonomy is respected if the two conflict. But this does not take account of the special status people have by virtue of their faculty of reason. Reasoning people do not necessarily seek that thing which is most likely to benefit themselves. They may, for example, forgo personal benefit for the sake of some greater good. It is therefore important to consult reasoning people rather than to assume that they want good to be done to themselves.

    In this trial, however, the argument that the researchers raised against seeking consent was that obtaining the research subjects' consent would have had the effect of skewing the results. In other words the conflict was not between the duty to benefit the subjects and the duty to seek consent, but between the duty to conduct good research and the duty to seek consent. The researchers had the choice of fulfilling the first ethical criterion, to conduct good research, or the third, to seek consent, but they could not fulfil both.

    Ethics of not seeking consent

    Does this make the trial unethical? The answer to this question depends on the balance of apparently competing moral claims. If you regard the research question as paramount you might argue that the need to do the research outweighed the moral requirement to consult the subjects of the research. If you regard the subjects' welfare as paramount then you would be happy in this case since the subjects would suffer no harm. But if your overriding concern is that the wishes of people being used as a means to someone else's end should first be consulted then you would not consider this research to be ethical. By extension, of course, any research which needed to deceive the research subjects in order to obtain a result would not be regarded as ethical.

    There is the rider that, even if you decide that this research was ethical, there is the problem that the women who were in the trial might read this article and feel themselves to have been, if not harmed, then certainly wronged by not having had their wishes consulted. One can only hope the faculty of reason that they are deemed to have will lead them to conclude that the good done by the research is more important than their personal feelings. If not, then the researchers must answer their charges.

    View Abstract