A paper on a vulnerable population was published in a journal. The journal followed their usual procedures for processing papers on vulnerable populations, by requesting and reviewing further information on the ethics approval and consent procedures of the study (e.g.: recruitment procedures; blank version of the consent document participants read and signed; the study protocol that was approved; ethics approval certificate).
However, the paper published individual profiles (Short Tandem Repeat (STR)) which contained genetic data from the participants. A reader contacted the journal, highlighting that these data are identifiable and could pose a risk to the vulnerable participants in the study. The publisher was not previously aware of this, and after discussions with the editors the journal asked the authors to remove those data.
This prompted a wider discussion about data availability for manuscripts on vulnerable populations, and whether it is ethical to encourage data availability for such papers even when written informed consent for its publication was obtained from participants. There is growing concern in the media regarding the welfare of some vulnerable populations and an increasing pressure to protect those participants in research. The publisher is keen to do everything it can to protect vulnerable populations, and would therefore benefit greatly from COPE’s advice on the following three questions.
Questions for the Forum
- Do you recommend that data deposition be discouraged for studies on vulnerable populations?
- Is there anything else we can do as publishers to protect vulnerable populations?
- Should we discourage the sharing of all individual genetic data from all studies on human participants unless written informed consent was obtained for the publication of that data, or just STR data?
The Forum was informed that this journal has a range of additional measures in place to protect vulnerable populations who are the subject of their submissions. These include requesting extra ethics documentation and study justifications from authors, seeking advice from specialist chief editors, and only proceeding if the topic of the research is not deemed to be sensitive. The Forum applauded the efforts made by this journal to act responsibly in the management of sensitive data but noted the limitations to what a journal can realistically be expected to oversee. Editors should be able to expect that the person or team who collected the data has taken responsibility for data security and the protection of the populations they are studying, and that arrangements should be in place for its management via institutions and data repositories.
Comments reported that journals vary in their attitude to data availability, some encouraging it, others making it mandatory and others still leaving it to the discretion of the author. The Center for Open Science has taken a lead in recommending data availability and opinion could usefully be sought from there.
Journals and publishers could also take an active stance in encouraging authors to deposit data in subject-specific repositories where checks are already in place to manage and curate them. This would remove the need for each journal to request further supporting information and would increase the chances that any difficulties in study design or data security would be identified before the publication stage is reached. A blanket policy against discouraging the use of data on vulnerable populations is not advisable as reporting such research could be of positive benefit to those groups.
We have implemented COPE's feedback, and that of the Center for Open Science, in how we handle cases on vulnerable populations, and we are very grateful for the advice given.