Ethics approval for papers submitted to the BMJ

The editors aim to ensure that all articles published in the BMJ report on work that is morally acceptable. To achieve this, we aim to appraise the ethical aspects of any submitted work that involves human participants, whatever descriptive label is given to that work including research, audit, and sometimes debate. This policy also applies on the very rare occasions that we publish work done with animal participants.

Many people consider that studies referred to as audit do not need any consideration of ethics, whereas all research must be approved by a formally constituted research ethics committee or, in the USA, an institutional review board. But the distinction between audit and research is unclear, and the assumption that audit or analysing previously collected data is never unethical may not be justified. Furthermore, review by an ethics committee cannot necessarily guarantee that work is morally sound.

For these reasons journals have a duty to consider the ethical aspects of both submitted and published work. The BMJ’s policy on these issues has been developed with the help and advice of the BMJ ethics committee and its key elements are explained here.

Editorial appraisal of a study’s ethics is not always easy because the standard format for presenting original papers does not give room to report the ethical aspects of research. We welcome, therefore, detailed explanations of how investigators and authors have considered and justified the ethical and moral basis of their work. If such detail does not easily fit into the manuscript please provide it in the covering letter or upload it as a supplemental file when submitting the article. We will also be pleased to see copies of explanatory information given to participants. Even if we do not include such detailed information in a final published version, we may make it available to peer reviewers and editorial committees. We already ask peer reviewers to consider and comment on the ethics of submitted work.

Editorial appraisal of ethical issues goes beyond simply deciding whether participants in a study gave informed consent although this is, of course, one very important issue to consider. Editors should judge whether the overall design and conduct of each piece of work is morally justifiable, as summed up by the following questions:

• How much does this deviate from current normal (accepted, local) clinical practice?
• What is the (additional) burden imposed on the patients (or others)?
• What (additional) risks are posed to the patients (or others)?
• What benefit might accrue to the patients (or others)? 
• What are the potential benefits to society (future patients)?

Even when a study has been approved by a research ethics committee or institutional review board, editors may be worried about the ethics of the work. Editors may then ask authors for more detailed information and ask them how they justified the ethical and moral basis of the work. Editors may also ask authors to provide the contact details of the research ethics committee that reviewed the work, so that the journal can request further information and justification from that committee. For studies that have not been reviewed by research ethics committees or institutional review boards editors may ask authors to explain what ethical issues they considered and how they justified their work.

Editors may ask other editorial colleagues to evaluate the ethical aspects of an article, the authors’ comments, and the response of the relevant research ethics committee to the BMJ’s queries about ethics approval. This consultation may be informal, between the BMJ’s editors, or more formal, through seeking the advice of the BMJ’s ethics committee or the Committee on Publication Ethics (COPE). Problems referred to COPE or the BMJ ethics committee will be considered as anonymised summaries of the relevant articles, written by the editors concerned.

What happens when the BMJ considers a study to be unethical? We believe that editors have a duty to take on issues of unethical audit or research, not to seek punishment for the authors, but to prevent unethical practice and to protect patients.

If the BMJ, with or without the advice of its ethics committee and/or COPE, considers the work in a submitted article to be ethically unsound the editor may seek further advice or recommend investigation or action. The fact that the article would have been rejected anyway for other scientific or editorial reasons would not prevent the editor from taking such further action on serious ethics problems.

In the first instance the editor would usually contact the head of the department where the work was done to explain the BMJ’s concerns and recommend a local investigation. Secondly, the editor might write to the professional registration body of the paper’s guarantor or principal investigator. For a doctor in the UK, this body would be the General Medical Council.

In rare instances the BMJ might publish an article despite ethics problems in the work it reported. The usual reason would be that work done in one setting might not reach the ethical standard of work done in another setting, because of differing local resources and standards for health care and research. In deciding to publish such an article, we would consider carefully the context of the study and aim to balance the overall benefit to society against the possible harm to the research participants.

Editors' duty of confidentiality to authors

BMJ editors treat all submitted manuscripts as confidential documents, which means they will not divulge information about a manuscript to anyone without the authors' permission. During the process of manuscript review the following people may also have access to manuscripts:

• editors and editorial staff at the BMJ, including medical students on placement and occasional overseas visitors - usually doctors or editors from other journals
• external reviewers, including statisticians and experts in trial methods
• members of the journal's editorial committees, comprising the final stage in our peer review process for original research articles
• the only occasion when details about a manuscript might be passed to a third party without the authors’ permission is if the editor suspects serious research misconduct—see below.

What we do if we suspect research or professional misconduct

• if an editor has concerns that a submitted article describes something that might be considered to constitute research or professional misconduct, the case may be discussed with the BMJ Ethics Committee.
• if the case cannot be resolved by discussion with the author(s), and the editor still has concerns, s/he may report the case to the appropriate authorities. If, during the course of reviewing an article, an editor is alerted to possible problems (for example, fraudulent data) in another publication, the BMJ editor may contact the journal in which the previous publication appeared to raise the concern
• the BMJ is a member of the Committee on Publication Ethics (COPE). Cases of research or publication misconduct may also be referred to COPE in an anonymised format.

Guidelines on publishing articles critical of doctors or other health professionals

Trial by media does not constitute due process, but the BMJ has an interest in exposing wrongdoing. The BMJ has guidelines on articles in which the health professionals are clearly identified and those in which they are not, and some elements are common to both:

• those accused of wrongdoing have the right to a proper investigation using due process
• "trial by media" — including the BMJ — cannot constitute due process
• the media have a job to expose wrongdoing and the failure to deal adequately with wrongdoing
• professional journals have an interest to raise wrongdoing; not just to expose individuals but also to illustrate professional problems
• doctors and other health professionals are not entitled to any more protection from accusations of wrongdoing than anyone else.

Indeed, being a professional implies operating to a higher ethical standard than the general population. Thus the BMJ, when balancing the interest of patients and the profession against those of an individual professional, has an obligation to give more weight to the interests of patients and the profession than might be the case if a publication was balancing the interest of a lay individual against those of the general community.

Because of the sensitivity of articles that make accusations against individuals then the editor should be involved in all decisions on whether to publish and he or she should consult with at least two other editorial colleagues. Sometimes it will be necessary to consult more widely.

Articles in which the doctors or other health professionals are clearly identified

The article must clearly avoid libel. The facts being "true" will not be enough on its own. It will be necessary to be able to convince a court that they were "true". In other words, we will need to satisfy ourselves that we have the evidence. If in any doubt over libel (and there is usually doubt), then take legal advice.

Major accusations of unprofessional conduct should be investigated by employers, the General Medical Council, or other institutions that can ensure due process. We will want to report the results of such investigations if they have a broad importance. We may want to include comments from those who have made the accusation, the accused, those who held the investigation, or other commentators — on, for instance, the broader significance of the investigation. It is not essential after an inquiry to get comments from everybody or necessarily from the accused. But "balance" should be considered. For example, if you have a quote from those who made the accusations you should consider getting a quote from the accused — and if you decide not to you should be prepared to justify your decision.

We will not make substantial criticisms of named doctors and other health professionals in the journal except in the following circumstances:

• an investigation has been held and the results are publicly available
• there is evidence that the investigation was inadequate
• there seems no possibility of an investigation being held
• when an investigation has been held we must aim for balance. This will almost always mean getting a quote from the "accused".

Articles in which the doctors or health professionals are not identified

The BMJ often wants to publish articles, including scientific papers, which expose poor performance by doctors and other health professionals. If the aim is to expose wrongdoing by particular individuals, then individuals should be named and the guidelines above followed. If, however, the aim is to make a general point about professional performance then the following guidelines should apply:

We must be clear that the aim is not to expose the wrongdoing of an individual but to make a general point. The general point should clearly be important. The importance of the general point must be weighed against the possible damage to an individual or individuals.

We must ensure that the individuals cannot be identified. This will usually mean that the article will have to be anonymous. If, however, a considerable number of individuals (say more than 15) are being criticised at once — for instance general practitioners in a particular city — then the article may not have to be anonymous.

We must recognise, however, that anonymity is hard to achieve. With information that emerges from the doctor patient relationship we have set a standard that even the patients themselves should not be able to recognise that the information is about them. This means, in effect, that informed consent must be obtained from patients in all cases. It is not necessary to achieve this degree of anonymity with papers that describe the wrongdoing of individual professionals and are published because of their general importance. In other words, it will not be necessary to get consent for publication from the accused individuals even if the individuals, their immediate colleagues, and those making the accusation and some of their advisors, family, and friends may recognise the individuals from anonymous articles. If a wider circle than this can recognise the individuals then the guidelines for when individuals can be recognised should be followed.

If all the previous guidelines are followed then it will not be necessary to include quotes from the accused.

Libel

The United Kingdom has very strict libel laws. You can be sued for libel “if you lower someone’s standing in the eyes of his or her peers.”

To defend itself against an accusation of libel, a publication has to prove that the statement it published was true, that it was published “without malice,” and, where possible, was in the public interest.

If an allegation turns out to be false (ie based on incorrect facts), we will find it hard to defend, so fact checking is imperative. But we may have a small chance of defending ourselves, if the allegation has been shared fully with the “accused”, and that he or she has had a chance to respond, and if that response has been forwarded unedited to us.

 So here are a few “musts” for authors.

 1)    Ensure that you check all your facts

2)    Ensure that all articles are balanced. If you are publishing an allegation against someone, you must give the accused a chance to reply.

3)    When you approach the accused, you must reveal in detail what your allegations are, so that he or she can have a chance to answer them in full. If, for example, you are going to claim that a hospital employed a doctor who was not properly qualified, and it did not investigate complaints against that doctor, you must put all the allegations in full to the hospital management, so that it has the chance to answer each and every one of the allegations.

4)    It is no defence to say that an allegation has already been published elsewhere. If an allegation about a doctor or a drug company has appeared in a newspaper in Spain, Italy, or the US, for example, we cannot rely on that fact to defend ourselves. Firstly, that local newspaper might have got the facts wrong; secondly, the libel laws might be different in that country. So although the doctor/company might not have sued in that location, he or she could come after the BMJ in the UK, because our libel laws are so tight.

5)    If you are worried about an article, please alert the news editor to any doubts you have, so we can discuss it. The BMJ also retains a lawyer who is paid to give libel advice, and the news editor can discuss it with him.

6)    Always be prepared to ask for more time (and more money) if you think that an issue is complicated, and requires more investigation than you originally thought.

7)    Wherever possible, please obtain documentary evidence for your claim. 

It is also important to remember that words can be interpreted in different ways. You may think that your article is harmless because you have not laid an allegation clearly at one person’s door (you might have just alleged that something dubious “occurred”), but it is not “you” that matters here, it is how an ordinary, disinterested member of the public would interpret it. If an ordinary person would infer from the story that someone in the story was responsible for the dubious occurrence, you can be considered to have libelled them.

If in doubt, please ask for advice.   

Informing workforces about the results of research in which they have participated before publication in mass media

Research undertaken with workforces can take place only with their full cooperation. Understandably, they expect to hear from the researchers about the results of the research and its implications for them before publication in the mass media.

These guidelines advise researchers how to do this without allowing the results to leak into the mass media before the full peer reviewed research has been published:

• the onus is on researchers to meet with the workforce
• they should make arrangements but keep journals aware of what is happening
• the researchers will meet with the workforce early in the week of publication. In the case of the BMJ, which is embargoed until 00.01am on Friday morning London time, this will usually be a Monday
• the journals will not put out press releases until after the meeting between researchers and workforces has taken place

The researchers and workforce managers and representatives will emphasise the importance of maintaining the embargo, but workforce managers and representatives will be able to speak to the media on the understanding that the embargo is respected.