Abstract
As part of my professional role I sit on a number of ethics committees. In front of one of these there recently came what I would have to describe as a troubling application. I will describe it as best I can without giving any non-blinded details as I am genuinely conflicted about what our response to this should have been (or rather I am convinced about what our response should have been - I’m just unsure about how we ensure that that response is shared…)
This piece of research noted that evidence in regards to an existing treatment for a condition had what appeared to the the unfortunate side effect of increased mortality rates in children who were prescribed higher doses of this particular drug for a particular condition. The FDA was quite rightly concerned about this, and asked the company producing said drug to provide evidence to assure them that higher doses didn’t cause higher mortality rates with the subgroup of adult patients with this condition.
My committee had two main concerns with the study which I will briefly outline, since I’m more interested in the subsidiary questions raised by the research than discussing the specific piece of research.
After some initial discussion the committee hit on two major issues with the study.
The first was that the study design was a prospective randomised blinded trial where new patients would be recruited to different dosages of the drug not based on their individual needs (though they would all need some amount of the drug and would all get at least that amount), but instead the needs of the trial - this meant that many participants would be given much greater doses than they needed - which given that the main aim of the study was to see if higher deaths occurred at higher dosages this seemed difficult to ethically justify - why not look at patients currently prescribed high doses rather than risk new patients’ lives with no corresponding benefit for them?
The second more technical issue was that the study had been powered only to show that the higher dose had no more than double the mortality rate than the lower dose - which both didn’t seem to meet the FDA’s request and made us somewhat suspicious that there might be in house data already showing a less than double but still significant increase in mortality.
Unsurprisingly we rejected this application after some discussion and extensive feedback for the researcher. However this is a multicentre international trial which is obviously going through the approval process simultaneously at many institutions in many countries and given what we know about the variability of research ethics committee decision making will no doubt be approved in some places meaning that some patients will be unnecessarily exposed to potential harm without benefit for them(of course it may be that no actual harm eventuates but this is beside the point).
Were we obligated to try and share our decision more widely with the REC community? And if so how?
In the absence of an international research ethics system with single worldwide approval (a model already being contemplated rather bravely for the EU) it is likely that multiple committees internationally will always be involved in approving such research and this will lead to differing time frames, regulatory systems, information sheets, application forms etc.
We wouldn’t want it to be the case that research had to get approval from all RECs approving it in all countries and couldn’t be commenced until they had shared their opinion since this would drastically delay international research.
Within those constraints how might the international REC community share information and opinions about research which seriously bothers then?
One option is to expand the trial registration process so that ethics committees could enter caveats if they were seriously concerned about a trial. These could even be treated in a similar way to how serious adverse events are dealt with requiring local research ethics committees to be informed and able to withdraw the approval for the trial.
This will still no doubt delay some good trials and have costs attached to it, but delaying good research is merely failing to benefit, whereas allowing ethically bad research to proceed when we can prevent it is harming.
In this issue:
This issue primarily focuses on several issues broadly relating to the ethics and law of social science research.
The first paper of the issue reflects back to the lengthy discussion in this journal in 2012 of the justification of the ethical review of social science research.
The second paper provides a short report of the Ottawa statement on the ethical design and conduct of the emerging methodology of cluster randomised trials.
The third paper asks us what lessons can we take from the classic cases of Milgram and Zimbardo’s research in terms of minimising harm.
The final paper raises the challenging question of research confidentiality in sociology in the UK.