Masking the Eyes Is Not Anonymisation, and the ICMJE Says So

This is the policy that most often catches a clinician writing their first case report, because the standard is higher than the one used for consent to treatment and higher than most people expect.

The patient reads the manuscript. Not a summary, not a consent form describing the article in general terms. “Informed consent for this purpose requires that an identifiable patient be shown the manuscript to be published.” The reason is that identifiability is contextual: a detail that looks innocuous to the author may be the one that makes the patient recognisable to their own family, and only the patient can see that.

Internet availability is part of what is consented to. Authors “should disclose to these patients whether any potential identifiable material might be available via the Internet”. Consent given on the understanding that this will appear in a specialist journal is not consent to indefinite public searchability.

Eye masking does not anonymise. “For example, masking the eye region in photographs of patients is inadequate protection of anonymity.” The black bar is a convention that has outlived its usefulness, and the ICMJE names it specifically.

And it has to be visible in the paper. “When informed consent has been obtained, it should be indicated in the published article.” Identifying information — names, initials, hospital numbers — should not be published, and non-essential identifying details should be omitted entirely rather than obscured.

A patient’s right to privacy does not end when their case becomes interesting. The ICMJE puts it in one sentence: “Patients have a right to privacy that should not be violated without informed consent.”

The requirement to show the manuscript exists because authors are poor judges of what identifies their own patients. A rare presentation, a named occupation, a date of admission and a town will between them identify someone precisely, and none of them looks like identifying information to the person writing the case up.

The internet clause exists because the consequences changed. A case report in 1995 was findable by a librarian. The same report now is findable by a colleague, an employer or a stranger typing a few clinical details into a search box, in perpetuity. A patient consenting today is consenting to that, and ought to know it.

How a journal applies it. The ICMJE tells journals to put the consent requirement in their own instructions for authors, and most do — usually with their own consent form, which has to be used rather than a local one. BMJ, for example, requires its own form, requires the patient to have read the article, and does not accept electronic signatures.

A practical consequence: consent obtained on your hospital’s form at the time of treatment will usually not be accepted by the journal. Check the target journal’s form before you approach the patient, so you only ask once.