there will be general agreement in practice over judgements of competence is wholly misplaced. Hence commonsense, which is perhaps an insecure guide even to the soundness of the decisions of normal people, may actually be misleading here. ii) Problems of understanding: The most obvious difficulties of understanding arise with disorders involving disturbances of cognitive or intellectual functioning (for instance, in mental handicap, dementia and confusional states). However, a wide variety of communication difficulties may be associated with other species of psychopathology. Anxiety, for example, is a barrier to communication of a general kind that may be much increased in a number of mental disorders. Similarly, depression may produce a slowing of intellectual processes that resembles dementia; and in hypomania there may be marked distractibility. Patients unable to give consent The difficulties just outlined may sometimes be so extreme as to invalidate consent altogether. The problem of consent has been most discussed in respect of difficulties arising from the patient's psychopathology - defective comprehension, for example, and severely impaired insight. However, coercion (overt or covert), material non-disclosure, and plain misunderstanding may all invalidate consent. This raises particular problems for research. In relation to ordinary clinical practice, the ethics of involuntary psychiatric treatment (although certainly contentious) can be derived from the Hippocratic principle. Involuntary treatment of a patient whose capacity for consent is impaired may be justified under the Hippocratic principle if it is either in that patient's interests or (a widely recognised although still contentious exception) necessary for the protection of others from serious harm (20). A research procedure, on the other hand, defined by the intention to advance knowledge, must be differently justified. Any benefits to the patient, direct or indirect, are (by the definition of research itself) secondary; and the principle of protecting others from harm is essentially a restrictive principle, the purpose of which is precisely to limit to exceptional cases any subordination of the patient's interests. It is because the patient's interests are subordinate in this way that the ethical requirements for consent to a research procedure are more rigorous than those for consent to treatment. With a research procedure it is the patient's assent, not merely the absence of their dissent, that must normally be considered valid, and this of course greatly extends the number of potentially problematic cases. The Mental Health Act 1983 and its companion volume, The Code of Practice (21), are both silent on the question of research. The Royal College of Psychiatrists' guidelines suggest that dissenting patients should always be excluded from research, whether or not they are considered competent (22). Even this restriction could raise difficulties - for example, in studies of involuntary admissions or treatment procedures under the Mental Health Act. However, the point is that in research, all incompetent patients, even those assenting, are ethically problematic. Furthermore, there is a medico-legal twist to the problem. For, at least under English law, the consent of a third party (of a relative, say) is legally effective only in the case of children. Hence, while involuntary psychiatric treatment is covered by the Mental Health Act, there is at least a possibility that research procedures with adult incompetent patients (whether assenting or not) may actually be illegal. Indeed, where such research involves touching (for example to take acopyright. on November 22, 2021 by guest. Protected by http://jme.bmj.com/ J Med Ethics: first published as 10.1136/jme.19.2.85 on 1 June 1993. Downloaded from 88 Ethics of research with psychiatric patients: pninciples, problems and the primary responsibilities of researchers sample of blood) it could conceivably result in an action in battery (23). It is important to keep these problems in perspective. In the first place, on the question of battery, legal opinion itself is divided (24). Then again, as the Royal College of Psychiatrists' guidelines emphasise (25),