Deidentification of facial photographs: a survey of editorial policies and practices.

We analysed all journals from two Journal Citation Reports (JCR) categories: 'Dentistry, Oral Surgery and Medicine' and 'Otorhinolaryngology' published in 2018 for their policies on publishing facial photographs and actual practices of publishing these photographs in articles. We extracted the following data for each journal: JCR category, impact factor, volume, issue, instructions for authors regarding ethical issues, instructions for photograph deidentification, journals' references to standard research and publishing policies, presence and type of published clinical images, separate informed consent for the publication of patient photograph and methods of deidentification. The sample included 103 journals, which published 568 articles with 1404 clinical images. Around a half of the journals (52%) had a policy on clinical images, however, the only predictor of having a journal policy on clinical images was reference in the policy to International Committee of Medical Journal Editors Recommendations (OR=3.00, 95% CI 1.26 to 7.14, p=0.013). Identifiable patient photographs were found in 13% (79/568) of the articles, constituting 9% (128/1404) of the total sample of images. Only 16% (13/79) of articles publishing recognisable patient facial images included a statement about consent for publication of the image. From the total sample of articles, 34% (27/79) contained deidentified but recognisable patient photographs and only 22% (6/27) of them had a statement about patient consent for photograph publication. The patients' consent was more likely stated in the article in cases of recognisable facial images (OR=2.81, 95% CI 1.41 to 5.63, p=0.004). Journals publishing clinical research involving the face and neck region need to establish and enforce policies on publishing clinical images.