Videos, photographs, and patient consentBMJ 1998; 316 doi: https://doi.org/10.1136/bmj.316.7136.1009 (Published 28 March 1998) Cite this as: BMJ 1998;316:1009
- a Ethox, Division of Public Health and Primary Care, Institute of Health Care Sciences, Headington, Oxford OX3 7LF
- b Oxford Medical Illustration, John Radcliffe Hospital, Oxford
- Correspondence to: Dr Hood
- Accepted 16 February 1998
A video showing real surgical operations was about to be sold through high street shops during the closing months of 1996. The BMA, GMC (General Medical Council), and the Institute of Medical Illustrators were quick to condemn this commercial exploitation of sensitive and confidential material. Media coverage was generally critical.1 A temporary injunction stopped any sales of the video, and a subsequent court order permanently prevented its distribution. Although the film's producer claimed that the surgeons concerned had given their permission for the video footage to be used, it emerged that many of the patients had not.
It is common practice to illustrate medical books with photographs of patients. In how many such cases have the patients given valid consent for publication? The advent of digital imaging has allowed photographs and video recordings to be stored, accessed, and distributed around the world with ease. Consequently, there is an increasing demand for medical images. Does valid consent for the use of a photograph in book publication cover the use of the same picture in electronic publishing? In most cases we doubt whether there is sufficient documentation to find out for what uses the patients originally gave consent.
• The internet and electronic publishing are powerful tools for the dissemination of medical information and have created a demand for medical images
• Consent should be requested from patients for all medical photography and for the subsequent use of their images whether or not they can be identified by the picture
• Specific consent should be obtained if an image will be used in electronic publishing and we describe a new consent procedure that covers such use of pictures
• Review of this procedure after 4 months shows that 85% of patients continue to give consent for publication of their image despite explicit discussion of the possibility of the image becoming available on the internet
Respect for patient autonomy is generally regarded as one of the central ethical principles in medical practice.2–4 Such respect has two related implications for the use of medical images: the first is to do with consent, the second with confidentiality. In general, information about patients that doctors obtain in the course of their clinical work is confidential.5 A doctor should normally have the patient's consent before sharing that information with others beyond the healthcare team, particularly if the individual might be identifiable from that information.
This issue has been discussed with regard to published case histories in the BMJ. 6 7 The editor wrote: “Authors and editors must thus ensure that patients have given their consent to publication whenever there is a possibility that the patient may be identified.” Although some exceptions to this position have been suggested, such as when interests of public health outweigh the importance of confidentiality,8–10 it is broadly the position agreed by the International Committee of Medical Journal editors.11 Images taken in the medical context, just like the information that a patient gives a doctor, form part of a patient's confidential records and should therefore be treated in exactly the same way.
Fully informed consent
Existing BMA and GMC guidelines state that patients have the right to be given as much information as possible on where an image might be used. 12 13 It seems clear that, if a patient has given permission for a picture to be shown only to appropriate professional staff, such an image should not be used in publication. However, does explicit consent for publication of a photograph in a medical textbook cover its use in electronic publishing?
It has always been possible for members of the general public to have access to medical images by browsing through textbooks, although we imagine this rarely happens in practice. The situation with electronically published images could be very different. Even if these are published in CD ROM “textbooks” it is easy for them to be copied and, for example, put on to the internet, where they would become readily available to a large number of people. Indeed, for educational purposes, it may well be desirable to put such images directly on to the internet. With current levels of security on the internet, there is little to ensure that such images are not widely seen, distributed, or misused.
It is therefore important that when an image is taken for medical publication the patient is made aware of the possible forms of publication now in existence and of the lack of control that it is possible to exercise over who will see the images. If patients have given consent for book publication only, it is doubtful whether such consent validly covers publication in electronic formats and on the internet.
Patients' rights—to ownership or just to confidentiality?
Medical images can be categorised into those from which the patient can be identified and those from which identification is unlikely. Just as much care needs to be taken in using this categorisation with images as with case descriptions.6 A traditional way of preserving anonymity when a photograph includes a patient's face is by blacking out the eyes. It is questionable whether this successfully disguises identity.11 Digital imaging can distort features a little more effectively (see figure), but what seems unidentifiable to a doctor may not be so to patients and their family or friends. Distinguishing marks, tattoos, posture, and gait may all reveal identity.
In many cases identification is most unlikely, such as from a photograph of a small area of skin or from a chest radiograph. In such cases is consent required? It could be argued that if patients cannot possibly be recognised from a picture they have no right to restrict its use. This has been the position taken in a number of discussions about written information 6 11 and is presumably the principle behind Smith's conclusion: “If we have an epidemiological paper with data on 5000 individuals will we require consent from all of these people? The answer will always be no when, as is usual, the data are presented in a combined form: no individual is identifiable.”7
However, patients may have rights akin to ownership, rather than confidentiality, over an image of themselves. For example, it would certainly seem wrong for a doctor to take a photograph of a skin lesion without the patient's consent, and, if it is wrong to take the photograph without consent in the first place, should the patient not be able to restrict the use of the photograph? Patients do have the right, we believe, to give consent for photographs to be filed in their notes but refuse permission for them to be placed in the public domain. If this is correct then it would seem that, even when an image does not identify the patient, the patient's consent would be needed for publication. This is the position now taken by this journal, and by the GMC in its recent guidelines on visual and audio recordings of patients.13 According to these guidelines, even when a patient could not possibly be identified, a recording cannot be used beyond the medical setting without specific consent. Continuing this line of reasoning, even when a patient has given consent for book publication it is not sufficient to assume that this allows electronic publication.
New procedure for obtaining patients' consent for publication of medical images
1 The clinical photographers have been trained in asking consent from patients. A new consent form has been produced for this purpose 2 The patient is asked to read through the consent form (available on our website www.bmj.com) 3 The photographer explicitly discusses each of the sections of the consent form with the patient and invites questions 4 The patient is informed precisely about the nature of the images to be taken and whether the patient is likely to be recognisable 5 Patients who give consent and who may be identified from the images are given two weeks from the date of photography, during which they can withdraw consent, before the pictures are available for publication 6 Three categories of consent are presented to patients: (a) use of the images in their confidential notes, for medical teaching, and for publication; (b) use restricted to patient notes and medical teaching; (c) use limited to confidential notes alone 7 In discussion with the patient attention is drawn to the possibility that the image will be used in electronic publication 8 The patient may view the images at any time and can withdraw consent, in which case the image is deleted permanently from the database. However, it is emphasised at the time of consent that full recovery of the image may not be possible once it has been made available for publication 9 It is made clear that refusal to give consent for the image to be made, or to be used in any specific way, does not affect the patient's medical care 10 If the patient is aged under 16 then consent is requested from a parent or guardian. However, the views of competent minors are taken into account, and if they refuse to give consent no images are taken 11 After a video image has been taken the patient is asked to confirm the initial consent
Revised consent form
Oxford Medical Illustration are currently involved in collecting an electronic database of photographs and video recordings for teaching and publication, including electronic publication. The controversy prompted by the surgical video led us to carefully consider the issue of patient consent. Review of consent forms used by several medical illustration departments showed them to vary greatly in design and in the degree of information presented to patients. None of these forms mentioned the possibility of electronic publication or of distribution through the internet. To address this, we constructed a new consent procedure and consent form (see box). In developing this procedure, we rejected the possibility of relying on “impliedconsent,” which is raised in the GMC guidelines.13 While such consent may sometimes be valid with regard to recording the image, it cannot provide valid consent for the image's publication.
Particular care needs to be taken with regard to minors. The guidelines of the International Committee of Medical Journal Editors allow parents or guardians to give consent for the publication of written information.11 In our procedure the photographer requires consent from both parent (or guardian) and a competent minor (even if below 16 years old). The fact that the image cannot be removed once it is in the public domain is emphasised.
This procedure and consent form have been in use for several months. Some patients have expressed concern over the potential use of their images on the internet, and for this reason have refused to give consent for publication. Others have not wanted their image to be used for commercial gain, in which case the image is not entered onto the main database. An audit of the first four months of using the new procedure showed that only 15% of the 518 patients referred for clinical photography refused to give consent.
As consent procedures in medicine, medical research, and medical reporting continue to develop, the procedure for the publication of medical images needs to be scrutinised. This issue is relevant not only to departments of medical illustration but also to individual physicians who wish to continue using personal slide collections for lectures and publication.
Contributors: CAH and PD developed the new consent procedure. CAH, PD, and TH jointly developed the wording of the consent form. TH and CAH explored the ethical issues associated with medical photography. The paper was written jointly by CAH and TH and edited by PD. Revisions were by TH and CAH.
Funding: This work was funded by Oxford Medical Illustration and Oxford University Press.
Conflict of interest: None.