Article Text
Statistics from Altmetric.com
In the Journal of Medical Ethics, Mirjam de Vos and colleagues1 publish a provocative clinical ethics case study. They describe the case of a young child with severe brain damage whose parents wished to discontinue life-sustaining treatment against the advice of medical professionals. As noted in the accompanying editorial by David Isaacs, the case raises important questions about parents’ role in decision making for critically ill children.2 It also raises interesting questions about patient (or surrogate) consent for case studies.
Case studies are important in medical ethics. They provide a valuable link between ethical reflection and practice. Case reports can be useful in identifying new issues for ethical analysis that have not previously been considered.3 They can also provide an insight into the complexity of clinical medicine, and some of the challenges of applying ethical theory in the real world.
It is widely accepted that case reports should not include identifying details such as patient names, addresses or photographs unless those details are essential and patient consent has been obtained. However, deidentification may not be sufficient. Even if identifying details have been removed, the Committee on Publication Ethics recommends obtaining patient consent.4
Why is consent required even when no identifying details are published? Patient consent might avoid two possible harms from publication of case reports. The first of these arises if the patient were to identify himself or herself in a case report and feel aggrieved that their personal medical information had been shared without their permission. They might thereby lose trust in health professionals. Although this harm might be thought unlikely, the ready availability of medical papers on the internet renders this concern more plausible. Even before the internet and open access journals, medical journal editors have reported being contacted by patients who had recognised themselves.5 Prior consent would avoid this sort of patient distress. The second, more serious, harm is the possibility that other individuals might recognise the patient and thereby learn confidential medical details about them. Patient consent would potentially reduce this risk by ensuring that patients agree to the details that are published about them.i
The BMJ policy has long been that consent should be obtained in all case reports unless the patient and family are untraceable and the case has been anonymised to such a degree that neither the patient nor anyone else could identify them with certainty.8 Journal policies are not all as stringent as this. A recent survey of 491 academic journals found that only 40% of journal websites indicated a requirement for patient consent with case reports.9
However, the nature of case reports in medical ethics is such that consent may be particularly challenging.3 The types of cases that yield ethical dilemmas often involve patients who are not competent to provide consent, differences of opinion between patients or surrogates and professionals, communication breakdown and difficult family dynamics. They may involve unusual or distinctive features that make true anonymisation difficult.10 End-of-life dilemmas will often have been followed by the death of the patient, and professionals may not have ongoing contact with families.3
One reason why professionals may not seek consent for case reports in medical ethics is a feeling that patients may refuse permission for publication and/or may be upset by the request. The case report in this issue and one recently published by David Isaacs in the Journal of Paediatrics and Child Health (JPCH) provide some (admittedly anecdotal) evidence relevant to that concern.
De Vos and colleagues had not initially sought consent for publication of their case report. The editors of the JME asked them to do so, at which point they approached the family. Interestingly, the family they approached declined to provide consent—because they felt that this would bring back too many unpleasant memories and emotions. However, de Vos and colleagues then approached another family that they had interviewed (with a similar experience). That second family was very keen for their child’s case to be published. Indeed, the authors eventually included them as coauthors of their paper.ii
In Isaacs’ case, paediatricians initially sought publication of a case involving parents who requested continued medical treatment against medical advice (the opposite type of conflict from the de Vos case). The case report was rejected by the BMJ on the basis that patient consent was required.11 The authors were not willing to approach the family, and eventually published their report (without consent) in JPCH.12 Eight years later, the authors sought to publish a follow-up to the case report. At that point, the editors of JPCH felt that it was inappropriate to publish without consent, and Isaacs and colleagues13 then approached the family. The family consented, and the case report was published.
As these two examples indicate, families may well provide consent for publication of case reports, even where those cases involve conflict between the family and professionals. Nevertheless, consent will not always be forthcoming if sought and may not always be possible to obtain. Are there circumstances where it would be appropriate to publish ethics case reports in the absence of patient consent? Some authors have argued that where authors cannot or should not obtain consent, that they should be prepared to forfeit the benefit of publication.6 At the Journal of Medical Ethics, we take the view that consent should be sought wherever possible (box 1). Where consent is not possible, harms to the patient can be alleviated by strict attention to anonymising patient details. Significant harm is unlikely where only the patient or physician directly involved would be able to identify themselves.iii Alternatively, fictionalised or composite cases that have been inspired by real experience may provide sufficient clinical flavour to enhance or support ethical discussion without risking patient harm. Finally, we accept that there are situations where there is a strong public interest in publication of case reports in the absence of consent. We will consider publication of such case reports, however, will require authors to make a strong case that any risks to individuals involved in the case are minimal or non-existent and are outweighed by the benefits of publication and debate (box 1).
The JME policy on patient consent for case reports
Consent from the patient or from surrogates (for incompetent patients) should be sought for case reports published in the journal.
Case reports may be published in the absence of patient consent in the following circumstances:
-
Cases in the public domain, where all details have previously been published
-
Cases where it is not possible to seek consent (patient and family deceased or uncontactable)
-
Cases where it is not appropriate to seek consent (eg, would harm the patient)
In situation 2 or 3, authors will be expected to make a case to the editors that they have anonymised or modified the case sufficiently to minimise or remove risk of harm to the patient. Second, they need to demonstrate that there is a sufficient public interest to outweigh any possible harms.
Footnotes
-
Competing interests None.
-
Provenance and peer review Commissioned; internally peer reviewed.
-
↵i It would not remove the risk altogether. In the only published instance of harm from medical case reports of which we are aware, a patient complained to the General Medical Council after a friend recognised her in a psychiatric case report to which she had consented.6 In that case, the General Medical Council found the psychiatrists not guilty of professional misconduct because the patient had provided consent.7
-
↵ii We should note that we do not have the consent of the first family to mention their discussions about consent with de Vos and colleagues. Nevertheless, we judged that here our criteria for publication in the absence of consent apply: only the individuals concerned could identify themselves, harm is likely non-existent and there is a significant public interest in discussion of the issues.
-
↵iii We have focused in this editorial on patient consent. We do not regard it as usually necessary to obtain the consent of professionals who may be mentioned in a case report even if they may recognise themselves. Consent would be required if there is a possibility that the physician (or other professional) would be able to be identified with confidence by a third party.
Linked Articles
Other content recommended for you
- When lecturers need patient consent
- Case reports on dangerous infectious diseases: a review of patient consent
- Managing self-harm: the legal issues
- Public interest in health data research: laying out the conceptual groundwork
- Qualitative documentary analysis of guidance on information provision and consent for the introduction of innovative invasive procedures including surgeries within NHS organisations’ policies in England and Wales
- The patient who refuses nursing care
- Opt-out paradigms for deceased organ donation are ethically incoherent
- The impossibility of informed consent?
- Discussing surgical innovation with patients: a qualitative study of surgeons’ and governance representatives’ views
- Ethics of organ procurement from the unrepresented patient population