In 2013, the world began to witness an unprecedented Ebola epidemic in West Africa that is now smoldering. Weak health systems were primarily to blame for the resulting suffering and loss of life. A few years earlier, one of the discoverers of the Ebola virus said that Ebola is “really a disease of poverty and neglect of health systems.”1 With adequate health systems and basic infrastructure in place, the epidemic would likely have been contained in its early stages.
Absent such systems and infrastructure, the outbreak quickly seemed to spiral out of control. In August 2014, the World Health Organization (WHO) declared the epidemic a public health emergency of international concern, which is a formal declaration of a public health crisis with potentially global reach under international health regulations.2 Clinical trials of experimental Ebola vaccines and treatments soon emerged as a key component of the global response. These experimental interventions were in the earliest phases of testing at the beginning of the outbreak, and whether and how they should be used caused heated controversy, not only among investigators, sponsors, and host communities, but also among bioethicists.
This chapter offers an overview of the ethical debate about the use of experimental Ebola interventions during the ongoing outbreak. It first describes four major challenges for using unproven vaccines and treatments during this epidemic. It then discusses whether it is ethical to conduct trials given these challenges and, if so, under which conditions experimental interventions should be tested. Reflecting the debate up to July 2015, when this chapter was written, issues around trial design receive special attention.
As concern about the Ebola outbreak waxed in 2014, four major challenges emerged around using the available unproven treatments and vaccines. The first challenge was an acute sense of urgency to deploy the experimental interventions and, some argued, get trials off the ground. Previous Ebola outbreaks had fatality rates as high as 90 percent, and, to date, there were no proven vaccines or specific treatments for the disease.3 Many cautioned that experimental interventions were unlikely to make an impact on this epidemic, but the possibility that a “magic bullet” might be among these interventions remained a powerful motive for endorsing their use. When some experts warned that more than a million people in Liberia and Sierra Leone could be infected with Ebola by early 2015,4 and there were concerns that the disease might become endemic in the region or spread globally,5 the use of potentially effective vaccines and specific treatments seemed more pressing still.
The second challenge was profound uncertainty about both Ebola virus disease and the existing experimental vaccines and treatments for the disease. Before this outbreak, Ebola had claimed less than 3,000 lives cumulatively over four decades. Previous epidemics had occurred in rural areas and had been relatively easy to contain, involving at most 500 patients per outbreak.6 Given this, insight into the natural progression of Ebola and its prevention or treatment was extremely limited before this epidemic.7 In addition, several unique features emerged during this epidemic that increased uncertainty about whether the Ebola virus had been, or was, evolving. Specifically, the outbreak occurred for the first time in urban areas, and it spread exponentially for several months, leading to fears that dangerous viral mutations might occur.
Because the burden of disease from Ebola virus was limited before this outbreak, research on Ebola was limited as well. Some governments of high-income countries funded research into vaccines and specific treatments because they feared the Ebola virus might be used as a biological weapon, given the high fatality rate of the disease.8 This research had identified several interventions; however, all of them were in the earliest phases of testing when this epidemic broke. For example, ZMapp—a combination of monoclonal antibodies—was under study in primates when it was first administered to two US citizens who had contracted Ebola in July 2014 in Liberia.9
Other potential specific therapies for Ebola were just as poorly understood. A number of investigational or licensed interventions targeted at diseases other than Ebola were being explored for their anti-Ebola activity, but clinical testing was far into the future.10 In addition, no more than a few Ebola patients had ever received blood or plasma from survivors prior to 2014. The uncertainty around Ebola virus disease was therefore compounded by a profound uncertainty about the existing experimental interventions for Ebola.
Major feasibility constraints were the third challenge in providing clinical care and preventive services or conducting clinical research. The epidemic occurred in some of the least-developed countries11 and brought weak health systems to a near-complete collapse, deepened distrust of health providers within the population, and even resulted in aggression against a small number of them. Moreover, preexisting research capacity and infrastructure were extremely limited in the affected region, and historical cases of abuse and misconduct by researchers from high-income countries loomed over research efforts during this epidemic.12
The fourth challenge was deep ethical controversy about how the experimental interventions should be used in light of the above considerations. This is the topic of the remainder of this chapter.
Two fundamental questions quickly dominated the debate. First, should the existing experimental vaccines and treatments for Ebola be used? Second, if the answer were yes, should they be deployed as part of clinical trials or as part of what might be called “experimental clinical care or prevention,” or—more conventionally—“compassionate use”? These questions were precipitated by the use of ZMapp and other experimental interventions, such as whole blood transfusion, when several health care workers from high-income countries were infected with Ebola.13 The WHO took a leading role in the ensuing debate and convened an international panel that endorsed the use of unproven vaccines and treatments provided a range of ethical criteria were met.14
The WHO recommendation was appropriate for several reasons. Even with effective supportive treatment—fluid replacement, broad-spectrum antibiotics, malaria treatment, and antipyretics—it was considered unlikely that fatality rates could be lowered to below 30 percent.15 Moreover, when the WHO made its recommendation, the fatality from Ebola was estimated to be at 50–70 percent because patients were diagnosed late and treatment centers were overrun.16 Given these grim prospects, it was reasonable to ask patients to assume considerable risks from early experimental interventions—or, more precisely, considerable uncertainty about their risks and potential benefits—in order to potentially save their lives or contribute to curbing a dangerous epidemic.
The argument for the use of the available experimental interventions is weaker for vaccines than for treatments because healthy individuals have alternative means of preventing an infection, such as wearing personal protective equipment and practicing good hygiene. However, investigational vaccines against serious diseases like Ebola tend to pose lower risks than investigational treatments for these diseases, given that vaccines are administered to healthy individuals and therefore expected to be much safer than treatments for patients whose lives are at risk. Moreover, as the high number of infected health workers in this epidemic illustrates,17 some populations have an increased risk of contracting Ebola virus disease due to higher exposure to the virus or inadequate protections (or both). Greater risks from early experimental vaccines thus seem reasonable as well, especially in health workers, burial teams, and other at-risk populations.
Furthermore, testing the existing unproven interventions is critical for an effective response to large-scale epidemics in the future, and robust data can only be collected when sufficiently large numbers of individuals are affected (i.e., during an outbreak). Given that Ebola virus strikes unpredictably, and—before this outbreak—typically affected small numbers of people, there may not be an opportunity to pursue larger trials in the near future. The fact that evaluating the experimental interventions has important social value in helping to potentially control a virulent infectious disease, and that there are few alternatives to gathering data during an epidemic, further supports the reasonableness of accepting increased levels of risk or uncertainty.
Some also argued that using the experimental interventions could be instrumental for curbing the ongoing epidemic, notably by providing infected individuals with an incentive to seek care, potentially identifying a “game-changing” vaccine or treatment for Ebola and promoting public trust.18 Such arguments for the use of experimental interventions should be treated with caution. All of them build on the hope that the investigational vaccines and treatments will make at-risk populations and patients better off, or at least not worse off. However, under ordinary circumstances, only 10 percent of investigational agents beginning phase 1 make it to commercial launch,19 and some cause serious harms that were unpredictable at the start. This percentage may have been even lower for the existing Ebola interventions, given the limited clinical and preclinical data on their effects. More realism about the effects of unproven agents is warranted, even if this can be psychologically difficult when a deadly outbreak is seemingly spiraling out of control.
Instrumental arguments for the use of experimental interventions also assume that such use will not compromise the general response to the epidemic. However, there are several ways in which the use of experimental interventions might compromise containment and care. For example, public trust can be undermined, notably when unproven agents prove harmful or, in the case of clinical trials, communities in need do not receive fair benefits from the research.20 For instance, complex treatments like ZMapp—which is expensive to produce and requires intravenous administration—are unlikely to be implemented in resource-poor settings in the near future, and trust could be undermined if communities do not obtain access to any proven, effective interventions from the research. Experimental vaccines can also complicate containment if they cause Ebola-like symptoms and thereby put further pressure on health systems. These issues are obviously complex and require more research. Yet the above examples suggest that the use of experimental interventions can have a negative impact on the general outbreak response, so that priorities between such use and containment or care may need to be set.
Importantly, the WHO did not adequately address this issue. It stated, “investigational therapeutic or prophylactic options should not divert attention or resources from the public health measures that remain the main priority in outbreak control.”21 Yet this position ignores potentially unavoidable conflicts between using investigational vaccines and treatments and efforts to contain the epidemic or care for patients, especially in the context of weak or collapsing health systems. More generally, the WHO seemed to have no clear strategy for how to integrate the use of experimental interventions into the general response to the outbreak. The organization declared the epidemic a public health emergency of international concern on August 8, 2014, and published a short web summary of its recommendation to use the existing experimental interventions just four days later; this was more than two weeks before its general “Ebola Response Roadmap” was completed.22 Despite the urgency of the situation, a clearer and more realistic vision for how to integrate the use of unproven vaccines or treatments and the broader outbreak response would have been helpful—especially since the WHO was poised to take the international lead for coordinating trials or experimental care or prevention.23 24
While there was broad agreement that it was acceptable to use the available unproven vaccines and treatments, controversy arose as to whether they should be deployed in clinical trials or as part of (monitored) experimental clinical care or prevention. The difference between the two can be more or less pronounced depending on the trial design. Individually randomized, controlled trials (IRCTs) include a control group that does not receive the investigational vaccine or treatment. This is a stark contrast to experimental clinical care or prevention, where all individuals or patients are eligible to receive the investigational agent, barring limitations in supply and countervailing clinical judgment. Uncontrolled trials fall between these two poles because they specify conditions for providing the experimental agents—for example, eligibility criteria and stopping rules—while still providing everyone who qualifies with an investigational vaccine or treatment. Data on patient outcomes are gathered in clinical trials as well as experimental clinical care and prevention, although the data collection is more limited in the latter.
The question of whether trials or experimental care and prevention were appropriate opened deep rifts about the norms that should guide the use of existing experimental interventions during this outbreak. Those who took a clinical perspective argued for experimental clinical care and prevention, trying to do everything possible for individuals who currently suffer from a life-threatening disease or are at risk of contracting it. By contrast, those who took a research perspective argued for clinical trials that enhance potential benefits for individual participants to the extent this is compatible with generating valid data for the benefit of future populations. The disagreement is apparent in the WHO statement that “compassionate use is justified as an exceptional emergency measure”; it remains unclear whether the Ebola outbreak as a whole qualifies as an emergency that justifies compassionate use, or whether the term emergency is more confined.25 Conversely, some commentators have argued that experimental clinical care or prevention is a “serious mistake” and should be avoided.26 While these claims were made early in the debate and might therefore lack nuance, they illustrate the deep disagreement about whether a clinical or a research perspective on the existing experimental interventions should be adopted.
A more balanced position points to considerations that support trials in some cases and emergency clinical care or prevention in others.27 For example, if an experimental intervention is promising based on preclinical or clinical data, and known to be safe, plentiful, and easy to administer, it might be offered to all patients outside of formal trials. Imagine that evidence suggests a daily vitamin C pill could improve clinical outcomes in Ebola patients. Experimental clinical care could be justified in this case because patients are unlikely to experience harm, vitamin C is readily available, and the costs of providing it—with respect to production, supply, and clinical administration—are low.28.By contrast, if an experimental intervention is scarce or costly and its safety uncertain, trials seem more appropriate so that it can be evaluated more systematically and the available stocks are used to this effect. Nonetheless, how these and other considerations are weighed critically depends on whether one believes that clinical or research norms should guide the use of experimental interventions, and how these norms should be specified in a situation of crisis. More analysis of these issues, notably the latter question, is needed.
Overall, the above considerations suggest that clinical trials were generally more appropriate in the ongoing epidemic than experimental care or prevention. Most investigational vaccines and treatments were neither known to be safe nor widely available, and their clinical benefits were highly uncertain. Many agents were also costly to produce or complex to administer. This suggests that trials of the experimental interventions were generally preferable, not only to avoid harm to healthy individuals or patients, but also to determine whether producing and deploying these interventions on a larger scale would be a reasonable investment.
Once it has been determined that an experimental intervention should be evaluated in a trial, it is important to determine how recognized ethical criteria for research—such as a reasonable risk-benefit ratio, fair participant selection, and informed consent—should be applied given the particular challenges of this epidemic.29 The criteria had some reasonably straightforward implications.30 For example, the informed consent criterion requires that information be disclosed, and that voluntary and informed consent be obtained, in culturally and linguistically appropriate formats. The criterion of independent review requires that public accountability regarding the trials be ensured through ethical oversight and review. However, other implications were less straightforward (see table 14.1).
Table 14.1 Ethical issues regarding the conduct of clinical trials in the ongoing Ebola outbreak
• How can meaningful community engagement in the planning, conduct, and oversight of trials be achieved in a time of crisis?
• Which experimental interventions should be prioritized for study? Based on which criteria (e.g., expected safety or efficacy, availability, cost)?
• What level of evidence about the experimental interventions is needed to inform decisions about the need for additional research, marketing approval or withdrawal, or large-scale implementation?
Fair selection of study participants and populations
• Which criteria should determine the selection of trial sites (e.g., ability to benefit participants or meet recruitment targets)?
Acceptable risk-benefit ratio
• When are experimental interventions promising enough for clinical trials? Based on what kind of data?
• How should speed and rigor be balanced in ethical review and oversight (e.g., expedited review vs. review by full committee)?
• How can informed consent or surrogate consent be obtained in a situation of crisis?
Respect for recruited study participants and communities
• How can the confidentiality of recruited or enrolled participants be ensured under the complex conditions of an epidemic?
Note: Because reasonable people will disagree about many of these issues, fair and accountable procedures for making respective decisions are essential. The ethical criteria are adapted from the work by Emanuel and colleagues.31
For instance, how should speed and rigor be balanced in research ethics review? What should determine the selection of trial sites so as to meet the criterion of fair participant selection? Who should be prioritized for trial enrollment? A growing literature has been addressing these issues,32 often building on prior work on the ethics of conducting research during epidemics and other disasters.33
The remainder of this section explores the question that has dominated the ethical debate so far, namely whether trials should be conducted using individually randomized, controlled or uncontrolled designs. This question has polarized commentators, with some advocating IRCTs34 and others rejecting them on both ethical and practical grounds.35 The question also divided a subsequent WHO Ethics Working Group on treatment trials; some members of the group considered IRCTs unethical in certain situations, while others regarded trials without a control group equally unethical if the trial design led to uninterpretable or misleading results.36 The fundamental issue at stake is whether investigators are required to provide the experimental vaccine or treatment under study to all trial participants. If the answer is yes, then an IRCT that randomizes a control group to standard prevention or supportive care without providing the experimental intervention—or providing the intervention at a later point in time—is ethically unacceptable.
The debate was particularly heated in the context of treatment trials, compared to vaccine trials, and will be the focus here. It primarily revolved around three of the recognized ethical criteria for research.37 First, the criterion of a reasonable risk-benefit ratio requires that the risks which research interventions pose to participants be reasonable in relation to the potential clinical benefits for them; the scientific and social value of the research; or both.38 Whether risks and potential benefits are reasonable from the individual participants’ perspective depends, among other things, on their available alternatives for treatment. Some commentators have argued that the outcomes of supportive care—the only available treatment during this epidemic—were so poor that the risk-benefit ratio for participants would be reasonable only when all received the study intervention.39 However, although fatality rates with supportive care varied considerably between locations and over time, they were as low as 30 percent in some treatment centers.40 Moreover, some humanitarian aid organizations argued that more aggressive care could decrease fatality from Ebola to 10 percent.41 If one considers these numbers together with the often profound uncertainty about the unproven treatments, including their potential to harm individuals who battle a life-threatening disease, participants’ interest in receiving an investigational agent looks much less compelling. While these risk-benefit judgments will ultimately depend on the specific treatment under study as well as the outcomes of locally feasible supportive care, it appears that receiving supportive care in the control group was not a clearly unreasonable option for participants in at least some treatment trials.
Second, there are other ethical criteria that interact with risk-benefit considerations for individual participants, notably the criterion of scientific validity. This criterion reflects investigators’ fundamental obligation to conduct socially valuable and scientifically sound research; when a study fails to address an important question by using adequate methods, it is widely considered to be ethically unacceptable.42 However, the obligation to conduct valid research is constrained by investigators’ duty to reduce and limit risks to those that are reasonable in relation to the potential social value of the research, as well as the duty to enhance potential benefits for participants.43
Under ordinary circumstances, these two duties are not equivalent in moral force. Investigators’ negative duty to limit excessive risks acts as a strong constraint on conducting valid research, implying that studies should not be conducted if the risks are unacceptable. By contrast, investigators’ positive duty to benefit participants is much weaker, and it is constrained to the extent that enhancing potential benefits to participants is consistent with designs that produce valid data to socially valuable questions.
The question is whether a public health crisis like the ongoing epidemic changes the relationship between investigators’ duty to enhance benefits to participants and their obligation to conduct valid research. This question is still open and compounded by methodological controversy around the quality of the data that different trial designs are likely to yield under complex conditions like this epidemic. However, once it has been determined that an experimental treatment is not appropriate for emergency clinical care and should instead be evaluated in research, it is difficult to see how the duty to enhance potential benefits for participants could be so strong as to undermine the scientific validity of trials.
Third, in addition to scientific validity, the criterion of fair participant selection is likely to interact with risk-benefit considerations for individual participants.44 Many of the experimental treatments for Ebola were in short supply because manufacturing capacity was limited, and public and private sponsors were reluctant to finance large-scale production based on limited data about the treatments’ effects. This implies that not every patient who might benefit from the existing experimental interventions could receive access. More analysis of how fairness considerations affect the choice of trial design is needed. Yet it is possible that IRCTs—which are effectively a lottery—offer the fairest way of distributing investigational treatments under conditions of scarcity.45
Urgency, uncertainty, feasibility constraints, and ethical controversy likely pose challenges for all research that is being conducted during infectious outbreaks. However, the ongoing Ebola virus epidemic involved a unique constellation of these factors as very early experimental vaccines and treatments were being deployed in a rapidly evolving situation. The debate about whether and how to use these interventions was often polarized and did not pay sufficient attention to the complexity of the situation. The response to future epidemics will hopefully benefit from emerging insights into the numerous factors that influence the ethical use of experimental interventions under similar circumstances.
Many thanks to the editors of this volume—Nicholas Evans, Maimuna Majumder, and Tara Smith—and Franklin Miller and David Wendler for their very helpful comments on an earlier version of this chapter. Thanks also go to other colleagues who have influenced my views on this topic, in particular Ezekiel Emanuel, Steven Joffe, and Verina Wild, and to Rose Mortimer for help with formatting. The work benefited from a Caroline Miles Visiting Scholarship at the ETHOX Centre, University of Oxford, and was funded by the People Programme (Marie Curie Actions) of the European Union’s Seventh Framework Programme (FP7/2007–2013) under REA grant agreement number 301816.