A group in a developing country decided to do a survey of maternal mortality in relation to the available local facilities in the various regions of their country. They set up a small committee to look at the incidence and causes. They obtained the permission of the local authorities, including the local elders, community leaders, and local health care providers, to scrutinise records and assess aetiology and quality of care. In the paper, aggregated results are described. However, the group did not seek formal permission from the ethics committee of their associated university as they regarded the survey as an audit.
Under these circumstances, do the committee consider that formal ethics committee approval was essential and, as this was not obtained, would this preclude publication?
The discussion centred on what is an audit and the distinction between an audit and research. In the case presented here, if we are talking about an audit in terms of aggregated data, the consensus was that the editor should publish. Also, as long as every reasonable step is taken to protect the identity of the individuals, the committee believed that formal ethics committee approval was not necessary.
The paper has been accepted and will be published.
The journal received a case report for a patient presenting with a particular syndrome in which patients give approximate answers to simple questions. This syndrome has been considered as a dissociative condition but others have argued that it reflects simulation of psychiatric symptoms. The case report was an individual who had crashed his car and, following that, developed complaints of memory impairment and at psychiatric review this syndrome was diagnosed. The patient was reviewed by the authors three years later and showed a number of inconsistencies on examination. Furthermore, it was discovered that, despite the fact that he was unable to recall the day of the week or his own birthday date, he had successfully trained as an official sports umpire. The case thus provided good evidence that this syndrome may be associated with malingering.
The report was reviewed by the Editorial Committee and publication offered as a letter. It was noted, however, that consent had not been provided and this was sought from the authors. The authors explained that this was not available and the Editor explored the possibilities of publishing without patient’s consent, applying the
whether it was in the public interest to publish
whether all attempts had been made to anonymise
that attempts had been made to contact the patient
one could assume that the patient would provide consent if it had been possible to contact him.
It was certainly felt that the paper was sufficiently novel to be published but, despite attempts to anonymise, the Editor had concerns that the individual could be identified. Moreover, the authors indicated that they had not approached the patient as they felt that he would almost certainly refuse consent to publication as the authors had established that he was malingering.
The Editor felt that since it could be assumed that the patient would not consent to publication, then the paper would need to be withrawn. The authors pointed out that this would inevitably produce a publication bias as patients who simulated disease would rarely ever give consent; a view that the Editor certainly shared.
Was this the correct decision not to publish?
The authors have withdrawn their paper, pending discussion at COPE.
While it was recognised that a case such as this, reporting for the first time an important finding, should be published for the greater good, any decision needs to be balanced against the possible harm to the subject of the case report if he or she is identified. It was suggested that one way around the problem would be to remove all forms of identification, including the names of the authors and their institution, but this would present problems in terms of ensuring an available audit trail. An editorial on the subject might be an option, but without using any of the data this would be difficult to produce. A more general editorial on the problems of lack of patient consent in certain circumstances such as in this case where individuals feigned disease, and how the failure to obtain consent could result in publication bias, would certainly be worthwhile.
It was generally agreed that the editor made a reasonable decision not to publish without the patient’s consent but it was a shame that he had to make that decision.
The editor fed back the views of the committee and the paper was withdrawn.
Information on competitors participating in a popular sporting activity was obtained from a website in the public domain. The authors used this data to see if the competitors' personal characteristics (height, weight etc.) affected their chances of winning. The editor asked the authors how they obtained consent from the competitors for this study. The authors responded saying that this data was in the public domain and therefore exempt from any requirement for ethics review.
By making their personal data available on a public website have the competitors also accepted that it will be available for medical research?
In this study, the authors were only reporting on basic information, but if they had included conditions/illnesses that these people were known to have, is that a problem?
The committee thought there is a lot of medical information in the public domain without consent. For instance if there was a claim that Kate Moss was anorexic and on drugs would we be shocked if it was included in a discussion on dietary habits? The committee thought that this paper was more for the lay press. If the information is available on a public website there is also the question of how robust the data is. It sounds like a dull paper that is not medical.
This is not an issue. Standards are set by the press complaints commission. The research sounds dull and it should be checked to see if it is robust.
The editor contacted the author and asked him to explain the fact that given that readers may be able to recognise the competitors from the information used, had he obtained consent from them? Alternatively, he was asked if he could provide proof that the study was approved by a recognised ethics committee who explicitly considered and addressed this issue. The editor also questioned the existence of the “website in the public domain” as it could not be accessed.
The author replied that at the time of data collection the website was viable, and the result of each match was posted publically. Since then the site has changed and these data are no longer posted on the website, likely because they are outdated. However, similar data are available from another site containing the same elements as the data used in the author’s study. The author also stated that as these data were in the public domain, the project was exempt from the requirement for ethics review as per the following two guidelines for ethics in human research:
(1) Tri-Council Policy Statement, Ethical Conduct for Research Involving Humans, Canada, 1998, Article 1.1(c) "Research about a living individual involved in the public arena, or about an artist, based exclusively on publicly available information, documents, records, works, performances, archival material or third-party interviews, is not required to undergo ethics review."
(2) OHRP-Code of Federal Regulations 46.101(b)(4): (b) ... the following categories are exempt from this policy: (4) Research involving the collection or study of existing data, documents, records, pathological specimens, or diagnostic specimens, if these sources are publicly available..."
The manuscript was then reviewed. Both referees recommended that the manuscript be rejected as it had has no valid biological question.
Some authors tested the effect of a food on the menstrual cycle. The manuscript included patient identifiable information, but the authors did not provide formal confirmation that the patients consented to publication of the study. Information was sent to the corresponding author, outlining legal obligations in respect of patients' consent to publication. But the authors stated that they considered themselves exempt from these regulations because they are research scientists and not healthcare providers. Furthermore, under the guidelines of their university's privacy office, all personal information relating to study subjects must remain confidential. To provide the journal with the appropriate authorisation from the patients would therefore contravene these guidelines. Under certain circumstances, the privacy review board may waive authorisation and allow publication without explicit consent to publish from patients. - Should the regulations apply equally to healthcare providers and research scientists? - In what circumstances would the privacy review board's waiver of authorisation be acceptable?
- The journal should set guidelines as to the standard of consent to publication that it required. To publish, authors would have to meet this standard. - These regulations do not allow for named consent, and they work prospectively in that patients give their consent to any future publication. The publisher must then keep that information confidential. - Were the subjects of the study patients or healthy volunteers as there is a different framework if the research subjects are not patients? - Ultimately, it is not possible to provide an answer to the editor’s questions. - The editor should contact the agency responsible for administering the regulations.
A paper was submitted that attempted to evaluate a new tool for diagnosing an acute symptom. This symptom is one that could be linked with various medical conditions—some causing little harm and some life threatening. The researchers recruited (and continue to recruit) patients into the study through a website devoted to this symptom. Patients viewing the website are asked if they would like to take part in an online survey. They are then asked to describe their symptoms in detail, and the tool (a computer algorithm) tells them their likely diagnosis. There is no mention of ethics committee approval having been granted in the paper. The website does have a standard disclaimer that patients are asked to sign, which states, among other things that the information contained in the website, and its associated websites, is provided as “a service to the Internet community, and does not constitute medical diagnosis or advice.“ It also states: “You grant us a non-exclusive worldwide royalty-free license to use or exploit in any manner, sell or transfer all data and information provided by you to us so long as such data and information is not identifiable to you.” - Does such a disclaimer constitute sufficiently informed consent?
- The website seems to tap into the problematic area of symptom checkers and diagnosis over the Internet. - Such websites are often used to feed trends and details to background commercial interests. - The people using the tool are often vulnerable with high expectations of these diagnostic tools. - The disclaimer as provided is meaningless—it essentially says that the result is not diagnostic at all. - Despite the claim that no relationship was formed, there is a de facto therapeutic relationship. - The paper indicated that the website was carrying out research on “patients” and so it could be argued that ethics approval should be sought. - The editor should engage the authors in a debate about the fact that there is a de facto relationship that could potentially harm vulnerable people.
A submitted paper detailing the negative experiences of overseas doctors applying for a training post in a district general hospital was poorly presented and scientifically weak, but on a topic of great interest and importance. The study consisted of an analysis of the CVs of the applicants and an analysis of responses to questionnaires sent to them with their rejection letters. Over a third of the questionnaires were returned. The editors were concerned that the authors had used the applicants’ CVs without their permission, and that the applicants were unaware that they would be used for reasons other than a job application. The CVs were not anonymised before the authors analysed them. And the study did not seem to have been submitted for ethical review. The thinking was that the overseas doctors might have been subtly pressurised to complete the questionnaire. In response to these concerns, the authors said that the study had been approved by the human resources department, who "own" the CVs, and the postgraduate tutor, who is responsible for pre-registration house officers. The study had also been informally submitted to the local research ethics committee whose view was that as it was an audit formal ethical approval was not warranted. The authors added that they did not ask for "permission" from the doctors as they felt implicit in the questionnaire was the fact that their answers would be used as part of the audit. The questionnaire was completely anonymous, they said, so to have asked the doctors “to sign a statement would have destroyed feelings of confidentiality which we felt were so important to the study.” The authors also enclosed a copy of a letter from their director of personnel and development in support of the above. The editors also wrote to the postgraduate dean and the chairman of the research ethics committee. The postgraduate dean replied that he was not aware of any formal approach for approval of the study, although he acknowledged that the postgraduate clinical tutors operate with a high degree of autonomy. The dean also stated that, in his opinion, the work was research and not audit. The chairman of the local research ethics committee wrote: “The part of the study involving sending questionnaires to the unsuccessful applicants was essentially research. …If consulted, we would probably have suggested that this part of the study required ethical approval. Individual consent would not have been required. However, the collection of individual comments may well have met with a different response. It is likely that we would have requested that further information outlining the purpose of the study and details of potential dissemination would have been required. We are not in a position to give retrospective ethical approval for research. We felt it was important that this information, having been obtained in good faith, should not be wasted. There was, however, concern that such a retrospective study, without adequate scientific scrutiny, may have introduced biases.”
- One way to rectify the use of the applicants’ data without permission would be to write to them and ask for retrospective permission. - The attitude of local research ethics committees varies and the authors might have found it difficult to obtain a review because the research did not concern patients. Nevertheless, the authors should have attempted it. - An ethics committee could have helped them draft a statement to potential applicants. - The human resources department does not “own” the CVs. - The likely harm from sending out the questionnaire was minimal in this instance, but the seemingly casual approach to the maintenance of confidential personal information at the trust was a cause for concern. - The editor should write to the authors to inform them of COPE’s views. - The editor should also ask the trust’s chief executive to look into the use of confidential personal information in the trust.
A paper described a problem of two women who wanted their fetuses to be tested for a genetic condition, but where in both cases their partners had refused to give consent. Should the journal publish such a paper without obtaining consent from the partners? The editors think not, but the authors are unconvinced.
_ It would be impossible to completely anonymise the case even if mention was not made of the genetic condition. _ One solution might be for the case to be published in the form of a discussion and alluded to in an abstract sense, although the authors had indicated that they would prefer to leave the case reports in. _ Public interest can override confidentiality.
The journal published the article on the general subject, but with no information on the particular cases. The authors eventually agreed that it would be wrong to publish the case reports without consent from the partners.
A specialist society wishes to post “case of the month” on the society website. The society is not proposing to obtain patient consent from all patients, but will anonymise the case reports instead. It has been suggested a case might be anonymised by changing details including age, occupation, or gender. It has also suggested that there is often much to learn from patients who have died, from whom consent would not be possible. Is this approach reasonable?
To do this would be in breach of the UK General Medical Council guidelines.
Rendering the patients anonymous is not enough.
If a patient has died, permission is not required, but should be obtained from the next of kin as a courtesy.
The facts regarding age, occupation and gender could not be altered.
The editor was advised that patient consent must be obtained, and that patient details must not be modified in an attempt to provide anonymity. The specialist society will inform authors that they must obtain signed patient consent before cases can be published.
An author requested advice about reporting unusual ocular manifestations of a patient who died from a fatal injury. The author sought the permission of the family to report the case, but they also requested that the patient’s name be added to the report in her memory. The author has proposed to add the following in the acknowledgement section: “The authors are grateful to the family of forename/surname for their permission to publish this case report, and at their request, have named the deceased, in memoriam.” What does the committee think?
_ If the deceased patient is identifiable, then there is a breach of confidence, and confidence survives death. _ The family cannot absolve this, but if it is in the public interest, then the name could be provided. _ The editor has to be satisfied that the dead patient would not have objected to any publicity, and that no other person(s) can be hurt or harmed by the publication of her name. _ If these criteria can be satisfied then a redrafted form of words could be published—for example: “The authors are grateful to the family of forename/surname for their cooperation in the publication of this case report, and at their request have identified her as the deceased, as a memorial to her.”
A paper containing three case reports of the same disease was accepted for publication. The disease reported is fairly rare. The parents of one of the cases consented to publication on condition that their daughter was referred to in the paper by her first name rather than as a case number. This particular case has been discussed in the course of a national inquiry, but it is not clear whether the other two cases were discussed. If the editor publishes this paper, the identity of the other two cases might be revealed if he respects the wishes of the parents. What does COPE advise?
Informed consent must be sought from the relatives of the other two cases. The editor should approach the authors to this effect. _ The fact that a national inquiry is under way is not relevant. _ The authors should also inform patients that the journal has a website and that there are regular press releases, so the case might receive wide dissemination.
The other two families had already given their consent. The authors contacted them again to alert them to possible press coverage. No objections were raised so the article was published with the patient’s name in full, as requested by the parents.