We received a paper which describes genotyping results from a large number of individuals (>50) from five unrelated families, in which family members had various blood and liver conditions. On submission we noted that the paper included specific details regarding the clinical histories of individuals in each family. Some individuals were described in substantial detail, others only briefly. For example, information about probands included age at presentation, sex, ethnicity, clinical history, occupation, clinical complications (some quite specific), clinic attended (for some individuals), history of alcohol consumption, ages of relatives, clinical details for relatives, age at death etc.
The authors responded that when initially obtaining consent to the research project, patients consented that “[t]he results from studies on the research samples may be published, but individual patients will not be identified in the publications”. They claim that details included are not identifying, and that it would be impractical now to trace all living relatives. They asked whether our consent form has been approved by an IRB, and say that before using it they will need to have it reviewed by their IRB.
We are unsure how to proceed but feel that when consenting to the research project the individuals may not have realised they would have been described to this level of detail, and that we should respect their privacy. One option might be to ensure the authors seek consent to publication from all probands, but then remove extraneous detail regarding the relatives. However, it is possible that the IRB may have useful input regarding the publication of individuals’ details from this study.
The editor was asked if the personal information was essential to the study. If it were deleted, would the paper still have value? The editor confirmed that the information was crucial to the study and could not be deleted. All agreed that the editor needs to weigh up the need to publish against any harm that might be caused by publication. Most believed that the harm involved was probably quite severe and so outweighs the need to publish. One test that can be applied is to consider whetheran investigative journalist would be able to identify the patients from the study (in this case, most Forum members felt that they could, therefore the risk of identification is high). The Forum did not believe it was that impractical for the authors to obtain consent and so they thought it was reasonable to ask the authors again to obtain consent. Did the authors make it clear to the patients how much detail and history would be published? Also, some noted that with pedigree studies, it can be very easy to identify patients. As with all such cases, the editor needs to make a judgement concerning the harm that might be caused to the patients in publishing and the benefits of publishing. Most believed that the risk of identification was high and that the editor should not publish without consent. The advice was to go back to the authors and ask them to obtain consent. COPE was asked about whether it could produce a generic consent form for all journals (to avoid problems of having to use multiple forms if a manuscript is reject by one journal and submitted to another). COPE Council agreed to consider this suggestion.
The editor informed the authors that the case had been taken to COPE, and the agreement of those who attended was that patients and family members would need to give consent to publication of the case(s) having read the paper. The authors responded after several weeks to say they had traced the patients and obtained consent to publication. They updated their paper to state this in the revised text, and also updated the text to minimise some details of family members, and to state that details were minimised to maintain anonymity but that further clinical details could be made available to other researchers on request.