Oxford committee was concerned that trial might be unethical

Editor—Stone and Blogg call for a national research ethics committee and cite in support of their argument their recent experience with an application concerning multicentre research, which was submitted to central Oxford research ethics committee.1 They are perhaps unaware of the imminent establishment of regional multicentre research ethics committees, which will reduce the bureaucracy entailed in submitting proposals for multicentre research and may well provide some uniformity in decision making. We were surprised that they singled out their application to the central Oxford research ethics committee as an example of a local committee making a poor decision.

Stone and Blogg ask “what is peculiar about the Oxford local research ethics committee.” Local research ethics committees have been exhorted to require systematic reviews of existing research before approving new proposals.2 The central Oxford research ethics committee was perhaps “peculiar”because it (presumably unlike the other 120 committees) was aware of a particular unpublished systematic review. This review raised concerns that without a placebo group the proposed trial design might be unable to answer the scientific question and might therefore conceivably be unethical. Stone and Blogg were invited by letter to allay these concerns, but they failed to do so and the committee heard no more of their research proposal. The file on their application still remains open. It is therefore not correct to state that the research ethics committee would not allow them to take part in a large multicentre study without major revision of the protocol.

Local research ethics committees are on one hand asked to require systematic reviews before considering applications2 and now, it would seem, on the other hand to ignore them.1 Stone and Blogg could well have participated in this study if only they had addressed the concerns raised by the local research ethics committee and were able to provide satisfactory answers to a serious ethical question.

Authors' reply

Editor—It was not our intention to criticise the central Oxford research ethics committee specifically but merely to highlight some general problems in the conduct of multicentre trials. We welcome the recently announced intention to streamline the ethical review of multicentre research in Britain through the use of multicentre research ethics committees.12 However, prior approval by these committees will be an additional requirement for approval by the local research ethics committee and will not replace it. An extra layer of bureaucracy may result, but we trust that this will not be so and wait to see how the new arrangements develop in practice.

Research proposals must indeed be considered in the context of best current evidence, such as that assembled, appraised, and synthesised in a published peer reviewed systematic review.3 An unpublished review that is not in the public domain cannot, however, be accorded such authority.

In any conflict of opinion in a multicentre trial individual participants are inevitably caught between the study's principal investigators and their own local research ethics committee. If the study protocol cannot be strictly adhered to, or if a local requirement for extensive revision causes substantial delays, participation is not possible. In our particular area of study the use of placebo groups has been much discussed and criticised.45 In our view the denial of timely treatment of known efficacy to patients in a placebo group is a greater ethical concern than that posed by the lack of a placebo group causing a loss of scientific elegance. It would seem that the other 120 study centres were of the same opinion.

Research discovers the right thing to do; audit ensures that it is done right

Editor—I would like to resolve the apparent contradictions identified by Scott and Pinnock1 in the BMA's publication Ethical Issues in Clinical Audit.2 The authors stated that there was a contradiction between the two sentences “audit is intended to influence the activities of an individual or team, i.e. local; but research attempts to influence medical practice as a whole” and “there is no need to consult research ethics committees on matters which are appropriately classified as audit.” They are concerned that this confusion might mean local research ethics committees becoming overloaded with work.

The first issue concerns the distinction between research and clinical audit. Scott and Pinnock have concluded that, by publishing results of clinical audit nationally, authors are attempting to influence medical practice as a whole and the work must therefore be regarded as research. The level of dissemination of a piece of work does not determine whether it is audit or research. Although results of clinical audit have traditionally been restricted to local use, they now have increasing prominence in national and international journals. A succinct but generalised distinction is that “research is concerned with discovering the right thing to do; audit with ensuring that it is done right.”3 As systematic pieces of work, published clinical audit projects can inform others on best methods and be used as comparators for other clinical audit work. This does not equate with attempting to influence medical practice as a whole.

The second issue concerns the workload of local research ethics committees. As a general rule, any ethical issues of clinical audit, whether aiming to inform, educate, or influence an audience, are the responsibility of clinical audit committees. Only in exceptional cases (for instance, when a clinical audit committee cannot reach consensus on an ethical issue, or when it believes that the work is research, not audit) should it need to refer the matter to the local research ethics committee. As with all BMA publications, the book is kept under constant review. We will consider any revisions in the light of recent inquiries.

Monitoring body is needed for audit

Editor—Scott and Pinnock are right to point out that the BMA's advice on audit is confusing.1 Audit cannot be defined by reference to who reads the report (local =audit; national = research). Under the terms of the BMA's guidelines, audits need not be submitted to local research ethics committees.2 Those of us who advise on such matters, however, feel distinctly uneasy at having no equivalent body for clinical audit, acting with impartial expertise in the interests of patients.

Although the question of equipoise should never occur, the intrusion into the patient's personal feelings (by questionnaire, for example) may be every bit as great during an audit as during research. Distinguishing between audit and research, although usually straightforward, is a red herring in such cases. Surely we should be asking whether patients' participation could conceivably be harmful. If so, a monitoring body of some kind is required. Local research ethics committees would indeed be “bombarded with demands for prospective ethical approval,” but this indicates a need to overhaul the monitoring system rather than to tweak definitions.