A submitted manuscript described an advocacy campaign in a lower-middle income country that used social media to promote an online petition supporting a proposed law. The authors were from an international non-governmental organisation and the country’s government, which together funded and conducted the campaign. The manuscript also reported on data from an online survey of the petition signers that had not been reviewed by a research ethics committee. The authors claimed that the survey was low risk, the participants had consented, and the data were not of a sensitive nature (eg, demographics and how people learnt of the campaign) and were aggregated, thereby precluding identification of any individuals.
Question(s) for the COPE Forum
- Would retrospective research ethics committee review be appropriate to consider in this situation?
Forum Advice and Follow-up
The Forum advised the journal editor to judge whether the study was sound and publishable, but could exercise caution by asking the authors to obtain retrospective ethics approval and to disclose this in the manuscript. However, because retrospective approval is reserved for exceptional circumstances, the authors could also be asked to elaborate on the consent given by participants, in an appendix. The Forum noted that ethics approval can differ by country and mentioned a relevant COPE guidance document. Guidance for editors: research, audit and service evaluations
When the editor advised the authors to obtain retrospective ethics approval, they said that in the study country, approval was not needed for surveys. The editor asked for official confirmation of such exemption, but the authors subsequently did obtain retrospective ethics approval and the manuscript was accepted.
This archived COPE Forum case is categorised under the COPE Core Practice of Ethical Oversight, which states “Ethical oversight should include, but is not limited to, policies on consent to publication, publication on vulnerable populations, ethical conduct of research using animals, ethical conduct of research using human subjects, handling confidential data and ethical business/marketing practices.”
For any studies involving humans, researchers should make sure they are familiar with and adhere to institutional, professional, local/national government, and international regulations. Regulations can be complex, covering issues such as prior ethical approval by an ethics committee (institutional review board); voluntary participation, participant safety, and informed consent; ethical conduct; and data confidentiality, privacy, and storage. Special considerations are needed to use stored human materials/samples and to retrospectively use data/artefacts that were collected for another (especially non-research) purpose.
Human research studies that require ethical approval can be surveys, interviews, and other cross-sectional qualitative studies; prospective and retrospective observational studies; and experimental/interventional studies. They can be in any discipline, not just biomedicine and health care (eg, education, behavioural economics, anthropology, and linguistics). In addition, for the sake of transparency, certain human experimental studies should be registered in an online clinical trial registry before the study starts. Corresponding ethics and transparency declarations should be included in journal articles and other publications and reports, and author instructions should be clear about required statements.
However, as the Forum mentioned in the described case, editors need to be aware that different countries have different approaches to ethical approval of studies. For example, some countries do not require approval for service evaluations, clinical audits, or routine public health surveillance.
According to COPE’s Guidance for editors: research, audit and service evaluations, when reviewing studies that do not require prior ethics approval, journal editors should evaluate quality, usefulness, and ethical soundness, as well as judge whether benefits outweigh harms. Editors should also ask authors to clarify or confirm local regulations. It is the authors’ responsibility to prove that the study was done ethically (eg, with informed consent or consent was not needed), it was automatically exempt from requiring ethical approval, or that exemption was granted by an ethics committee. These details should be included in the manuscript’s Methods section, and some journals also summarise approval and consent information in a dedicated endnote.
In the described case, an online survey was conducted among people who had signed a petition as part of a social-media advocacy campaign. The authors claimed the survey was low risk and did not require ethical review. However, the journal would have needed more information on risks, especially as survey replies might have been traceable to people who signed the petition (which may have been on a third-party platform that linked to social media accounts). The authors should have also confirmed non-involvement of any minors, vulnerable groups, or sensitive topics, and that they were transparent about survey purpose, confidentiality, privacy, anonymity, and any potential author and funder conflicts of interest.
As posed initially by the journal, retrospective ethics-committee review should be sought to grant study approval or formal exemption (or, if applicable, to waive consent or consent documentation). As suggested by Forum, consent procedures should also be fully explained in the manuscript. Although retrospective ethical review is reserved for special emergency cases, the post-study approval that was granted in this case confirms the low risk involved. Nevertheless, editors need to evaluate each case carefully, and some ethics review boards and journals do not consider retrospective applications for ethics review. COPE has previously reported a Forum case (98-31) where a journal rejected and did not allow retrospective ethics review of a questionnaire study. In another COPE Forum case (11-16), a journal rejected a paper on an online questionnaire study for being unethically conducted despite receiving retrospective ethics approval.
Advance ethical approval or official exemption is recommended for survey studies (highlighted in Forum case 20-06), including social-media surveys (highlighted in Forum case 20-05). Journals may request that authors provide a copy of the approval/exemption letter. Journals, and ethics committees, usually also require a copy of the survey questionnaire and informed-consent procedure and/or form. Journal editors and authors should be aware that advance permission may be needed for the use of certain questionnaires, survey tools, or scales (dealt with in COPE Forum cases 03-05, 18-01, and 18-04).
Finally, consent may be needed not only for human participation, but also for publication of data if there is any risk that published data (even if anonymized and aggregated) might be traced to individuals. The risk would be higher for social-media surveys or interviews in which participant fill-in replies or comments refer to personal information, are published as is, and have brief demographic descriptions. The issues involved and precautions needed would be similar to those when publishing case reports (outlined in COPE’s Journals’ Best Practices for Ensuring Consent for Publishing Medical Case Reports).
Trevor Lane on behalf of the COPE Education Subcommittee