A journal received a manuscript using secondary analysis of existing medical records in which there was no indication that ethics approval was obtained from a recognized ethics review board or that participants gave their informed consent to be included in the study. Instead, the authors explained that the study had been based on a secondary analysis of existing medical records and that no patient information had been passed on to a third party. Patients were identified by codes rather than names so that links could be made with different outcomes. Patient names were neither available nor utilised in the analysis and all records were held securely in a password-protected database which was only available to programme team members. They also stated that 'The conduct of this study did not inflict direct harm on patients, nor expose them to unnecessary risk as a result of accessing and reviewing their medical records'.
The journal’s understanding is that if the secondary data are anonymised and contain no identifying information, then ethical review is generally not required, but no ethics review appears to have been done in this study, at any stage.
Questions for COPE Council
- Should the journal reject the manuscript?
- Should the journal ask the authors to apply for institutional ethical approval to confirm the anonymous or identifying nature of the data?
- Could this manuscript be published without patient consent?
Advice on this case is from a small number of COPE Council Members. Most cases on the COPE website are presented to the COPE Forum where advice is offered by a wider group of COPE Members and COPE Council Members. Advice on individual cases is not formal COPE guidance.
We would suggest the editors consider the following before deciding on how to proceed:
- How large is the data set—if it is small, is there a greater chance of identifying individuals?
- Where were the data collected and for what purpose—if the location is very identifiable or the data were collected for a unique drug trial for example, would that increase the chance of patient identification?
- Why was the study not approved by an institutional review board—was there no appropriate oversight committee in the location?
- Was the original study institutional review board approved and what were the conditions of that approval?
- Was the original or any other secondary analysis of data from the original study published and, if so, what approvals were or were not noted by the authors in the publications?
At the very least, we suggest the editors need to go back to the authors and get answers to the questions above and find out if retroactive approval could be obtained. If there was institutional review board approval of the original study, that would be a useful piece of information in the decision-making process. Council would caution against rejecting the article because it could be published elsewhere without anyone questioning the ethics.
It may turn out to be reasonable to publish the study, but institutional review board approval is always a good idea in research involving human subjects, even if at the end of the process it is decided that there is no risk. Researchers deciding for themselves that there is no risk is a conflict of interest, no matter how reasonable the case.
The final decision will require a lot of input related to the issues identified and a thoughtful discussion of the editorial committee based on any additional information received.