A paper reported the clinical outcomes of patients suffering from a neglected disease before and after a change in the national treatment policy which raised the threshold of one laboratory parameter before a more toxic, but more effective drug could be used. The authors did a retrospective cohort study of patients treated under both policies, to see how this change affected outcomes. They did not seek informed consent or ethical committee approval for their study, and provided the following
justification for not having done so:
1) The paper does not report on primary research. All data analysed were collected as part of routine diagnosis and treatment.
2) Patients were diagnosed and treated according to national guidelines and agreements. Testing blood and CSF (as well as recording all other variables included in our analysis) is essential for confirming diagnosis and classifying patients. It is done for each patient without fail and as part of routine care, and is in no way an add-on for purposes of research.
3) The paper does not report on the use of experimental or new protocols. The diagnostic protocol that we evaluated was the national protocol.
4) The program was not set up as a study or research project, but as a treatment program. Hence we did not seek/obtain ethical approval for a study but rather project approval from the Ministry of Health.
5) Our analysis looked retrospectively at outcomes for a large cohort of patients treated. This was done internally as part of an audit/evaluation, so as to improve our quality of care. The decision to change the threshold was based on rather weak evidence, and having confirmed, by this analysis, our clinical suspicions that the cut-off is probably too high, thus exposing patients to a high risk of treatment failure and, consequently, death, we felt that this vital information should be shared with the
wider medical community – hence our decision to submit our findings to a reputable open source journal.
7) The authors had asked their Medical Director for feedback regarding the need for ethical clearance for such a retrospective analysis, and were advised that this was not warranted.
One of the authors had obtained approval from a teaching institution in another country to analyse this dataset, but when he supplied proof of this approval, it stated that it was contingent upon local ethical approval having been received.
This is a rare disease, and the results in this paper provide the largest cohort for analysis since the treatment was established in 1939. How should we proceed?
It was felt that the editorial team needed to weigh up the harm that might result from the publication of this article against the benefits that will be forthcoming. The main harm resulting from publication would be to the autonomy of the study’s subjects. The benefit is that the study will inform other physicians about this treatment, with the expectation of improved therapy for all individuals affected by this rare disease. Only the editorial team could make the final decision as to whether to publish or not. However, the general view was that it was in the public’s interest to have the study published, and that it will probably bring benefits to the very people whose autonomy may be harmed by its publication. An accompanying editorial, drawing readers’ attention to the matter and the reason for deciding to publish the article, would be appropriate.
The authors took a long time to revise their paper. The end result was that COPE's recommendations were followed, the authors requested and received retrospective permission from the government in which their study was done to use the data in this way, and the paper will be published in an upcoming issue with an editorial setting the scene.
Comments
I am often asked to review retrospective papers for peer-reviewed journals and it happens to read similar justifications for the lack of ethcal approval of retrospective, registry based journals. Since this post is quite old; would CORE maintain the same suggestions today? One additioanal challenge relates to the definition of personal data, in that the definition of human subject research entails the use of personal data. In turn, according to current regulations, personal data is any information relating to an identified or identifiable individual; an identifiable person is one who can be identified, directly or indirectly, in particular by reference to an identification number (e.g. social security number) or one or more factors specific to his physical, physiological, mental, economic, cultural or social identity (e.g. name and first name, date of birth, biometrics data, fingerprints, DNA…). According to Horizon 2020 Programme Guidance How to complete your ethics self-assessment, if data are anonymized in a way that requires "excessive efforts to be re-identified" than the data are not considered "personal data" anymore. We should conclude that retrospective, registry-based, research on anonymized data is not human subject research. Is this conclusion correct?
to post comments