Ethics behind closed doors: do research ethics committees need secrecy?BMJ 2001; 322 doi: https://doi.org/10.1136/bmj.322.7297.1294 (Published 26 May 2001) Cite this as: BMJ 2001;322:1294
- Richard Ashcroft, lecturer in medical ethics ()a,
- Naomi Pfeffer, professor of social and historical studies of healthb
- a Imperial College School of Medicine, Department of Primary Health Care and General Practice, London W2 1PG
- b School of Community Health, Psychology and Social Work, University of North London, London N5 2AD
- Correspondence to: R Ashcroft
- Accepted 19 February 2001
Biomedical research involving human subjects must be reviewed by an independent ethics committee.1–3 In the United Kingdom, such committees are normally NHS local or multicentre research ethics committees.4–6 There is currently widespread frustration in the research community, in the regulatory authorities, and in the pharmaceutical industry with the efficiency of such committees. 7 8 At the same time there are doubts about the effectiveness of these independent committees in protecting patients.9–11 Doubts have arisen in the aftermath of some high profile scandals in the ethics of research in the United Kingdom and United States.12 13 Guidance on research governance within the public sector to parallel governance within the pharmaceutical industry is therefore welcome. 2 6 14
An important reason for the contention surrounding research ethics committees is the secrecy that surrounds them and their decision making. We argue that this secrecy is rarely justified. To investigate this, we reviewed the guidelines regulating research ethics committees in common law jurisdictions and searched the literature on research ethics committees.
Research ethics committees have a vital role in protecting the public and facilitating useful research but are increasingly under attack
In many countries, including the United Kingdom, research ethics committees meet confidentially, but no justification for this is normally offered
The possible justifications (protection of patients', researchers', commercial, or academic interests and the committee's independence) are weak and unconvincing
The minutes and meetings of research ethics committees should be open to the public except in a few special circumstances
Independence of review is an important element of good research governance. This is partly protected by lay membership of committees, clear lines of accountability to public authorities, and strict rules relating to conflicts of interest. However, this protection of patients' interests may be insufficient, especially in situations where institutional pressure to ensure vigorous research activity is a strong collective influence on the committee. Several US institutional review boards have had such problems. 10 11
There is no evidence that review within the NHS does not follow the highest standards of independence. However, the work of research ethics committees is not widely known outside health research circles. In the United Kingdom, their work is not routinely audited. Local research ethics committees are required to file annual reports to the director of public health of the relevant local health authority, and multicentre committees must report to the secretary of state for health. These reports are public documents that can be seen on request by any interested person. 4 5
This is a very reactive or even passive way of informing the public about committees' activities. Moreover, because of pressure of work, reports are often filed late. Most importantly, these reports contain only minimal information about the studies considered (names of committee members, number of meetings, short title of the projects with date submitted, current status of the project—for example, approved subject to amendments, approved and ongoing, or rejected) and nothing about discussions that took place over them.1 15 16
Although the responsible authority can ask to see the minutes of review meetings, they remain confidential documents. UK guidelines require that these meetings are convened in private and that their proceedings are confidential. 4 5 17–19 The only exceptions to this rule are if the committee requires external advisors to supplement its knowledge and expertise or when the committee requests a researcher to attend to present a project or to answer questions about it.
This confidentiality of proceedings is not unique to the United Kingdom. Some countries explicitly require it. 20 21 Others leave it to a committee's discretion, while implying that confidentiality is a good idea (subject to the above exceptions). 22 23 Possible explanations for confidentiality include a culture of secrecy and the predominance of self regulation in medicine.24
The increasing demands for mechanisms of governance or audit of the work of research ethics committees will provide other means of monitoring their independence and quality of review. 25 26 This may mean that public access to the meetings and records of committees will be less important as means of ensuring quality. But the general trend within the NHS and UK politics (as elsewhere) is towards greater openness and accountability.27–29 For instance, the recent WHO guidelines on establishing research ethics review committees do not mention confidentiality but refer only to ensuring “that their tasks can be executed free from bias and influence that could affect their independence.”3 The question we raise is: never mind the explanations for confidentiality of ethical review, can it be justified?
Possible justifications of confidentiality
The published guidelines in the United Kingdom and elsewhere rarely explain the grounds on which committee confidentiality is justified. There are, however, four main arguments that could be used.
Doctors have a duty of confidentiality to their patients, but this is normally irrelevant to research ethics committees. Patient information will be relevant in only a few cases. Some research ethics committees operate ad hoc as clinical ethics committees, which typically discuss the cases of particular patients.30 In addition, committees must be notified of serious adverse events, and although these are reported anonymously, it may sometimes be possible, circumstantially, to identify individual patients. And in some trials in rare or serious conditions, it may be possible for patients to be identified. Confidentiality is also required for trials of gene therapy, where UK official guidelines require that ethics committees discuss patient information in confidence in the patients' interests.31 Very occasionally, a patient may complain to the committee about a trial or other research, and their complaint may need to be discussed confidentially.
It is hard to document how much of a research ethics committee's business is taken up with discussion or information about individual patients. However, except for committees that have a dual research and clinical ethics function, it must be a very small amount. We recommend that research ethics committees should not handle clinical ethics cases, but that if they decide to discuss individual cases, this should be done under “reserved business” at the end of the meeting.
Commercial and academic confidentiality
The obligation of ethics committees to “facilitate useful research” could be argued to include not exposing researchers to the risk of having their ideas imitated or stolen. There may therefore be a duty of commercial confidentiality to protect researchers' innovations. The Medicines Act makes it a criminal offence to reveal evidence submitted to, or discussions resulting from, approaches to drugs regulatory agencies, and this may be relevant to phase I clinical trials and other early stage clinical research.26 However, most research reviewed by committees involves drugs, devices, and protocols about which much information is already known and available publicly. For example, patent registrations require a large amount of technical information to be lodged and publicly recorded. Most “new” drugs are variations on themes, and much will be known about a given class of drugs, even if particular members of this class are better or worse understood. It is therefore doubtful whether most of the information carefully protected by committee confidentiality is in any real sense confidential outside the four walls of the committee room.
If the raw information is not secret, the ideas involved in putting it to a novel use may be. Researchers may worry that predatory and efficient competitors could “look over their shoulder” and beat them in the race to publish or market. We are sceptical about this. Data and methods are nothing without the skill, experience, and ingenuity to exploit them; research on the replication of scientific experiments has shown just how difficult this is if the replicators lack the skills and resources possessed by the originating group.32
Another flaw with this justification is that most of the information required by the committees on the common application forms is entirely public information and is the sort required by patients for informed consent. If there is no duty on patients to keep confidences about the research they participate in, it would be odd for a research ethics committee to have such a duty in respect of the same information.
We conclude that there is no risk to researchers' commercial confidentiality. In the climate of the Cochrane Collaboration and the NHS Executive's requirement that researchers register all non-commercial research with the national research register, we contend that there is a strong pressure to make research widely known within the NHS in the interests of patients.33
Privacy of researchers
There might be a case for holding committee meetings in private to protect researchers from attack or satire, especially for student projects and controversial research. However, even now researchers can be asked to attend meetings, and the lead investigator's name is recorded in the committee's annual report. In some cases there may be doubts over a researcher's competence or trustworthiness, but this could be discussed under “reserved business.” The decision about competence would, nevertheless, have to be recorded, and an appeals mechanism must be established to ensure the researcher has a right of reply. If the topic (rather than the researcher) is difficult and controversial, there may be a strong public interest. This would be a good case for making the process more open rather than less.
Privacy of committees
The clearest argument for committee confidentiality is to protect the committee's ability to review protocols independently and without coercion from the researcher or public interested parties. We are not clear, however, that confidentiality is the right kind of guarantee of independence; it could be used to hide inadequacies in the reviewing process or in the reviewers themselves. Since the integrity of the committee is not for the sake of the committee but that of patients, the veil of confidentiality requires stronger justification.
One analogy that could be drawn is with a jury. However, the analogy is neither exact nor persuasive. Juries sit in secret to promote free and frank discussion; to protect them from defendants and their associates; to preserve the finality of verdicts; and to avoid any temptation on the part of jurors to capitalise on disclosure.34 No one suggests that researchers threaten committees or that members want to make a profit through membership. It is important to distinguish a review from a verdict; there should be an appeals process if the researcher feels hard done by, and committees should be required to give reasons for their decisions. This would be easier if discussions were properly minuted and publicly witnessed. The main argument here is that confidentiality protects the freedom of discussion.
That meetings held in public do not permit free discussion is arguably special pleading. Parliament and parliamentary select committees normally meet in public, or have full publicly accessible minutes; the same is largely true of local council meetings, public inquiries, court proceedings, community health councils, and some meetings of health authorities and trust boards.
Research ethics committees serve several important public functions, including the demystification of medical research and provision of a public forum for the accountability of researchers in the public sector. The assumption that committees should meet in private and give minimal accounts of their decisions can be challenged on the grounds that it lacks clear justification; the grounds normally cited in justification are insufficient to the purpose. Given that the culture of the NHS is shifting towards one of openness and accountability, and that researchers increasingly argue for public registration and reporting of clinical trials, the confidential position of committees is anomalous. In democratic societies, public law, common sense constitutions, and standing orders will protect the integrity of committee review better than confidentiality. There are some circumstances when particular cases or agenda items may require private discussion, but these are the exception, not the rule. When such exceptions are claimed, clear appeal mechanisms and specific standing order provisions are required. The presumption has been for confidentiality, without justification; we argue the presumption should be for openness.
We thank Allyson Pollock, Gary Allen, Mark Wilson, Brian Hurwitz, Jennifer Blunt, and Richard Lilford for their help and advice.
Competing interests NP is a member of the management committee of Consumers for Ethics in Research.