Using pictures in the BMJ

We encourage authors to include pictures in submissions to the BMJ to add useful and relevant information. We also use pictures to help the journal to look good, to be well read, to entertain and stimulate readers, and sometimes to cast different perspectives on familiar subjects. Surveys show that readers may lose interest when faced with slabs of unbroken print, so we include some pictures to make articles more engaging and to draw readers in.

But there are pitfalls in publishing pictures in a medical journal, and we receive a handful of complaints each year about ours. Readers have two main concerns—that publishing pictures of patients may compromise privacy and confidentiality and that pictures may be altered and might thereby mislead.

The BMJ has strict policies on preserving privacy and confidentiality. We insist that authors obtain patients' written consent before we agree to publish pictures of real patients taken in a clinical setting. This applies even when an image only shows something that seems unlikely to lead to identification of the patient—for example, a small skin lesion or a single toe. Patients can and do recognise themselves, especially those with unusual or rare conditions.¹ And we know that masking someone's eyes does not prevent them from being recognised, a practice we abandoned years ago.²,³

Some authors think we are too pedantic in seeking consent for every clinical image. But our policy on images is just a subset of our general policy, that we need consent from patients for any information that comes from the doctor-patient relationship,⁴ and it complies with the General Medical Council's rules on publishing images of patients.⁵ Patients may also have rights akin to ownership over an image of themselves and do have the right, we believe, to give consent for photographs to be filed in their case notes while refusing permission for those pictures to be published.⁶

Given this policy on consent for images of patients mentioned in BMJ articles, is it inconsistent of us to publish pictures provided by agencies in news items and other articles? We believe that the BMJ would be at a disadvantage among other media if we didn't use such images, and pictures can often tell a story more powerfully than words. But we cannot take responsibility for the consent of people who are shown in pictures that we have obtained from agencies, libraries, other publications, and other commercial sources. We state clearly where pictures have come from, and we assume that they and their photographers have obtained relevant permission from models in any images showing people. Reputable picture agencies and other sources are unlikely to take the legal and financial risk of selling sensitive images without appropriate consent. If we doubt that someone photographed could have given consent—owing to severe mental illness, dementia, or learning disability, for example—we use our discretion and try to avoid images that might allow that person to be identified.

So much for protecting patients. What does the BMJ do to protect readers from misleading images? And has the advent of digital photography tempted us to use visual trickery? Journals, including the BMJ, have been cropping and masking photos for decades, and digital imaging has simply extended the possibilities. We alter clinical images only occasionally, when using them primarily as art rather than information, for example on the BMJ's cover (see box on bmj.com). When we have altered an image substantially, we state this in the legend or cover note.⁷ Similarly, when we buy from agencies scans, electron micrographs, scintillograms, thermal images, and other clinical images whose colour has been enhanced or changed, we publish these with explanatory legends.

The other questions we are often asked about pictures are more technical, and are mostly about finding and preparing images for submission, and getting permission and copyright clearance. To find answers to these questions please read our advice to contributors.⁸