Every day we learn of new suspected cases of Ebola infection in the United States, causing worries among the population and renewed efforts by the Centers for Disease Control and Prevention to contain the epidemic.
A student in the postgraduate course in research ethics that I teach at Einstein recently asked whether we could discuss the Ebola epidemic in class. Because I had prepared the syllabus for the course months ago, I hadn't thought of including a discussion of ethics in research on Ebola. But the time was ripe, and I gratefully accepted the student's suggestion.
A letter published online in the Lancet on Oct. 10 is titled "Randomised Controlled Trials for Ebola: Practical and Ethical Issues." The authors are a distinguished group of scientists, ethicists and physicians from all over the world, including several from West Africa, where the epidemic is raging. The letter makes an urgent plea for embarking on research for experimental treatments that would not randomize patients to a control arm that provides only conventional care or conventional care along with a placebo.
Departing From the Randomized Clinical Trial
Instead, the letter argues, "When conventional care means such a high probability of death, it is problematic to insist on randomising patients to it when the intervention arm holds out at least the possibility of benefit." To do so would violate clinical equipoise, the situation in which genuine uncertainty exists about whether an untested treatment has benefits or risks that exceed those of conventional care. There is no such uncertainty in the case of Ebola. Nevertheless, the authors' proposal rejects the gold standard in research methodology: the randomized clinical trial (RCT). How that departure from the widely accepted orthodox trial design will be viewed by research methodologists remains to be seen.
The proposal to use alternative methodologies to explore a beneficial treatment for patients already infected with the Ebola virus also departs from widely accepted ethical guidance for the use of placebos in research. The most prominent international ethical guideline is the Declaration of Helsinki, issued by the World Medical Association. Paragraph 33 says:
The benefits, risks, burdens and effectiveness of a new intervention must be tested against those of the best proven intervention(s), except in the following circumstances: Where no proven intervention exists, the use of placebo, or no intervention, is acceptable.
To say that the use of a placebo, or no treatment, is "acceptable" is not to say that a placebo or conventional care must be used as a comparator. The underlying assumption is that patients will not be made worse off if they receive a placebo when an experimental treatment is not yet proven. The Lancet authors state that "ethical arguments are not the same for all levels of risk," and an experimental treatment holds out at least the possibility of benefit. They make several suggestions for alternative research methods that can obtain provisional answers quickly, eventually leading to RCTs in which patients are randomized to one or another experimental treatment, with no one receiving standard care.
Implications for Future Vaccine-Safety Trials
I agree with the ethical analysis and conclusions in the Lancet article. But they lead me to wonder about the design of preventive-vaccine clinical trials in the future. The article mentions that vaccine-safety trials are currently underway, and these would have to demonstrate safety before moving on to the phase in which efficacy is studied. Because there is currently no efficacious vaccine against Ebola, the same question about a placebo-controlled RCT is likely to arise. Participants in a vaccine trial are healthy volunteers, so it may be argued that they do not have the same urgent need as patients suffering from a disease with a high fatality rate. However, those at high risk of being infected with the Ebola virus (healthcare workers and family members of sick patients) do have an urgent need for prevention, and there exists an urgent public-health need as well.
Clinical prevention trials typically require a large number of participants in order to demonstrate efficacy, but there may well be sufficient numbers in Guinea, Sierra Leone and Liberia, the three most deeply affected countries in West Africa.
It is a question for methodologists, rather than ethicists, to determine whether the results of a nonrandomized trial of an Ebola vaccine would be interpretable. If the same arguments offered by the authors of the Lancet letter are applicable to preventive-vaccine clinical trials, they could result in another example of a departure from the gold standard of research methodology.