Ethics of research with psychiatric patients: principles, problems and the primary responsibilities of researchers K W M Fulford and K Howse Warneford Hospital, Oxford and St George's Hospital Medical School, London respectively Authors' abstract In this paper some of the general issues surrounding recently published guidelines for the practice of research ethics committees are outlined, concentrating in particular on the difficulties raised by research with psychiatric patients. Research is distinguished from ordinary clinical practice by the intention to advance knowledge. So defined, research with psychiatric patients should be governed by the same four principles as research with any other group - knowledge, necessity, benefit and consent. In applying these principles, however, particularly the principle of consent, many acute difficulties are raised by psychiatric patients. A number ofproposals for addressing these difficulties are discussed. It is suggested that, notwithstanding the value ofpublished guidelines, and the help that may be available from research ethics committees, the primary responsibility for maintaining high standards ofpractice in research rests with research workers themselves. Introduction 1990 was a good year for research ethics in the United Kingdom. The Royal College of Physicians produced a revised and updated version of its Guidelines on the Practice of Ethical Committees in Research involving Human Subjects (1), together with a companion report, Research involving Patients (2). The Royal College of Psychiatrists prepared a detailed supplement to this document (3). The Department of Health issued their own draft guidelines for comment (4). These documents provide a valuable framework of practical recommendations for all those responsible for medical research involving patients. They leave unresolved, however, a number of issues and difficulties. It is with some of these, especially as raised by psychiatric patients, that this paper is concerned. The common theme that will emerge is the central importance of individual research workers in the development and maintenance of high standards of practice in medical research. Key words Psychiatric ethics; research. Definition and principles Research, whether therapeutic or non-therapeutic, is most usefully defined in relation to ethical questions by its intention to advance knowledge. This definition provides a direct link with the ethics of research. The ethics of ordinary clinical practice are governed by the broadly Hippocratic principle that the patient's interests should normally come first. It is the potential subordination of this principle to the advancement of knowledge that creates the need for ethical principles specific to research. These principles are the same for psychiatric patients as for any other group. They may be formulated thus: 1. KNOWLEDGE: there should be a reasonable expectation that the research will produce an increase in knowledge that is directly or indirectly relevant to patient care. 2. NECESSITY: there should be no practical possibility that the same increase in knowledge could be achieved other than by working with patients as research subjects. 3. BENEFIT: the potential benefits (to the research subjects or others) arising from the expected increase in knowledge should be of sufficient importance to outweigh any risks of harm inherent in the research. 4. CONSENT: patients should give valid consent to their participation in research; that is, they should assent freely and on the basis of appropriate information. Difficulties in applying these principles Although the principles governing good practice in research are largely uncontentious, difficulties often arise when it comes to applying them in practice. The difficulties may be especially acute in the case of psychiatric patients. Thus: KNOWLEDGE: