A roadmap for a public discussion of the ethics of animal research

July 19, 2019

Allyson J. Bennett

Last week we wrote about US legislative and other efforts that address research with nonhuman animals, particularly that conducted by federal agencies. Together those efforts have generated several federally funded workshops, panels, and committees over the past several years. Scientists, policymakers, advocates, and bioethicists have talked, written, and otherwise participated in discussions, white-papers, and reports about the use of nonhuman animals in research aimed at public health and safety. The National Institutes of Health (NIH) now appears to be in preparation to launch yet another workshop to address the broad topic of ethics and when/where/how nonhuman animals are used in research (workshop Day 1 description here). [Update 7/22/19 below]

Here at Speaking of Research we support public dialogue about research with nonhuman animals. As we’ve written before, starting with the foundational moral and ethical considerations as the basis for decision making can accurately situate serious reflection and recommendations. In fact, among the chief critiques and complaints about previous panels and workshops aimed at evaluating nonhuman animal research is that they did not begin with the big questions that provide a solid basis for discussing ethics. In 2011, for example, we wrote this about the conclusion of the Institutes of Medicine (IoM) committee that addressed NIH’s chimpanzee research:

“First, the charge of the IoM committee to assess the “scientific necessity” of the work, while specifically avoiding ethical issues, was clearly ill-posed, and – as the committee quickly realized – nearly impossible to carry out. We acknowledge the committee held serious discussions about the science of chimpanzee research and the availability of alternative methods, but it is notable that these were guided by principles that are ethical in nature.  Namely:

1. The knowledge gained must be necessary to advance the public’s health.
2. There must be no other research model by which the knowledge could be obtained, and the research cannot be ethically performed on human subjects.
3. The animals used in the proposed research must be maintained in either ethologically appropriate physical and social environments or in natural habitats.

Moreover, the IoM committee explicitly recognized that “ethics was at the core of any discussion […] on the continued used of chimpanzees in research”.

It is evident that the tension about the use of chimpanzees in research is not merely about science.  In fact, it is not even primarily about science, as arguably chimps can stand as valid scientific models in many areas of research.  It isn’t even about the cost of research.

It is largely about ethics.”

Several years later, a new NIH panel on research with other nonhuman primates encountered the same issue and was critiqued similarly.

In view of the outcomes and criticisms of previous workshops and panels, what can be done differently to achieve the goal of engaging in a genuine public dialogue about scientific research with nonhuman animals? As we’ve written in response to previous panels, we recommend thinking about the following:

1) Where to start?  Begin at the beginning and include the broad range of stakeholders

Among the first questions are whether nonhuman animals should be used by humans at all. If so, for which purposes? Under what conditions? Which animals? Use of animals for research is a small proportion of human use of animals, which includes for food, fiber, entertainment, labor, sport, show, and companionship (graphic here; more estimates here).

In contrast to the use of animals for these other purposes, however, only animal research and testing explicitly require advance ethical consideration in the formal US regulation and oversight system (as in many other countries). Among the requirements for approval are:

• Justification of the potential benefit in terms of advancing knowledge and serving humans, other animals, global society, the environment; also, justification and rationale for the species, or animal model, proposed for use
• Consideration of non-animal alternatives if they exist and are feasible to answer the scientific question
• Provision of animal care and treatment to minimize unnecessary pain or distress, balanced with scientific goals

Any workshop, panel, or consideration can (and should) begin with a serious review of the current system in place, making sure that all participants are beginning from the same basic knowledge. Further, the information should be conveyed clearly and specifically to the public, media, and interested policymakers. A central criticism of previous panels/workshops has revolved around who is, and who is not, included. In some cases, the range of appropriate participants (or interest holders) is narrowly defined, or inaccurately conveyed as two discrete camps. For instance: scientists on the one hand and ethicists on the other. Or, pro-research vs anti-nonhuman animal research. Both undermine meaningful dialogue because they encourage the perception of a simple dichotomy. Further, this kind of framing introduces an implicit and wrong assumption that the existing system of proposal, review, and conduct of scientific research proceeds without scientists engaged in ethical consideration and decision-making. At worst, it can provide the impression that bioethicists alone are qualified and charged with decision-making to ensure ethical science. Expertise across a diverse range of topics and content areas is important to meaningful dialogue that can best inform decisions. What is also important for public decision-making is representation of the full range of public stakeholders. Those would include patients and patient groups, for example.

Why start here? Doing so would help avoid proposals that duplicate existing systems. More importantly, it would allow for a factual and rational critique of proposals for change, including a thoughtful assessment of potential benefits and harms of proposed changes. It might also provide a solid foundation for identifying evidence and questioning claims about whether the current system is sufficient, or not.  In this regard, it would avoid the problem of beginning with an untested or undiscussed presumption that the current system is flawed and requires overhaul.

At the same time, beginning with foundational assumptions provides a public clarity of the positions held by various stakeholders. We’ve written about this previously in terms of “fair partners” in dialogue. For instance, for those opposed to all use of nonhuman animals by humans, the only acceptable outcome of an ethical consideration or policy about animal research is to decide it is not permissible (see here for further discussion). For those individuals and groups, all the evidence in the world about discoveries, scientific advances, medical progress that depend on the use of animals in research is irrelevant. Yet, that is unlikely the view of the American public and policymakers given that the vast majority of them consume other animals and a wide range of other uses are legal and acceptable in the US and elsewhere. For this majority, the questions about ethics surround the purpose of the use of animals, the potential benefits, whether there are alternatives, and consideration of how the animals are cared for, whether pain or distress are minimized—all questions that are embedded in the current system for review, approval, and oversight of animal research (see here, here, here, here for additional information).

Finally, even a basic consideration of the foundational questions will show that a species-by-species, or agency-by-agency, discussion of research is beginning a conversation at mid-point and on shaky ground. Thus, instead of hosting panels about primate research by NIH, or dog research by the VA, or monkey research by the FDA, we’d propose that federal agencies and others consider a much broader public discussion—all animals, all uses?

2) Define terms.

One notable thing in several of the panels and workshops on animal research and testing has been the use of terms that are poorly defined and only narrowly interrogated, if at all. Among them is the concept of “necessity.” Its use in this context appeared during the IOM panel, with the apparently simple idea that the panel should investigate whether chimpanzee research was “necessary.” It appeared again in a private workshop held to discuss nonhuman primate research.  We wrote about this one in 2017 and, in 2014, wrote about the concept of “necessity” advanced by a bioethicist who has continued to be part of these panels, Prof. Jeff Kahn (ex. here, here).

A similar concept, indispensable, appears in other guides for animal research. For instance, a recently released European report reads: “The use of animals for scientific purposes can be ethically justified only if an animal experiment is “indispensable” and the expected scientific outcome justifies the harm imposed on the animals in the course of the experiment.”

So what is the problem with “necessity” and “indispensable”? On face value it could appear these are quite reasonable criteria.  After all, why would anyone do research that isn’t necessary?  But the question is, necessary for what? To whom?

We could make a list of many publicly supported functions, or even many inventions of humans and query whether they are necessary– schools, roads, libraries, hospitals, military forces, birth control, books, music, the internet, cars, and so on.  What is quickly obvious is that the definition of necessary is tightly linked to evaluation of immediate and long-term outcomes, a balance of harms and benefits, and consideration of a broad swath of interest holders. That is—who is affected, how, and when.

At the most basic level, we can say that food and water are necessary, or indispensable for survival and health. Protection from adversaries may also be necessary for survival of those under attack, but the same military force that protects one set of interest holders may well eliminate another, thus complicating any simple conclusion about necessity.  Schools, roads, libraries, hospitals, military forces, birth control, books, music, internet, and cars can all be similarly critiqued.  Judgements about their necessity all depend on the needs and desires of interest holders.

The same is true for scientific research broadly, including the use of animals in research and testing.

3) Accurately represent the nature of science.

Inherent to the question of “necessity” is the question of how discoveries and advances are made. Sometimes discussions about whether animal research in general, or a specific study, should be conducted center on whether it will lead, or has led—either indisputably, or in high probability—to a cure, treatment, or medical advance. Proposals that center on evaluating a direct link between a particular study and an advance in drug development, treatment, or cure for a disease, however, most often proceed from a set of flawed assumptions. Among them is the assumption that science progresses on a linear track and within a very short timeframe. As we’ve written before, that is not the nature of basic (or fundamental) science (see Isaac Asimov’s essay here). Prof. Dario Ringach and others have written a number of Speaking of Research posts on this theme, see here, here, here for examples and further discussion.

A basic review of what is known about biology, brain function, immunology, genetics, behavior, and other areas of science will show that advances in knowledge occur over long timescales. Further, the outcome of a study is not known in advance of doing the research and the breadth of impact of a finding—including its applications—are most often neither known, nor necessarily predictable.

Of many examples, we might take the fact that the timeline between the discovery of neurotransmission and brain chemicals in the 1920s-50s was the foundation for the development of drug therapies to treat schizophrenia, depression, and other mental health disorders. The discoveries depended upon use of a range of nonhuman animals—frogs, rabbits, clams, cows, monkeys, and others. Yet those early studies of yet-unknown substances in blood were not aimed at developing a drug for schizophrenia. Nor would it be possible to identify the widespread contributions and benefits of the research for many decades.

Another example can be found in the case of the first drug approved for the treatment of HIV in the 1980s. That drug, AZT, indisputably lengthened the lifespan of those with HIV. It also provided key knowledge relevant to the development of subsequent drugs that were more effective and that have allowed individuals with HIV to live much longer, with a medically manageable condition.

Why does AZT provide a good example of the nature of science, animal research, and medical benefit?  The story is told NIH (here):

“NCI scientists developed AZT (azidothymidine) in 1964 as a potential cancer therapy.  It proved ineffective against cancer and was shelved…In the 1980s, it was included in an NCI screening program to identify drugs to treat HIV/AIDS. AZT decreased deaths and opportunistic infections, albeit with serious adverse effects. In March 1987, AZT became the first drug to gain approval from the U.S. Food and Drug Administration for treating AIDS.”

In other words, AZT was a drug developed and declared a failure for treatment of cancer. As such, it became part of a kind of “database,” but did not become a cancer treatment.  What that meant is when new compounds were being screened to combat a new disease, AZT was identified, tested, and employed. That previous “failure” became a success, something that could not have been known in 1964, became reality in 1987—a reality that benefited many patients.

As summarized by NIH’s National Institute of Allergy and Infectious Diseases:  “In the 1980s, the average life expectancy following an AIDS diagnosis was approximately one year. Today, with combination antiretroviral drug treatments started early in the course of HIV infection, people living with HIV can expect a near-normal lifespan.” (emphasis added)

What all of this simply means is that consideration of animal research (or any scientific research) should take into account the nature of the scientific process.  Despite the unpredictability of results, non-linearity, and long time-scales between discovery and application, what we know about medical and scientific progress is that it works in the long-run and is among the best tools we have to solve problems, address health challenges, and make advances that benefit humans, other animals, global society, and the environment.

That doesn’t mean “anything goes” with respect to research generally, or to animal research specifically. What it does mean is that a panel, workshop, or discussion that seriously examines ethics and decision-making should be grounded in explicit acknowledgement of the nature of science, the role of basic research, and the best fact-informed assessment of the range of likely consequences of any set of decisions.

4)  Consider the harm of inaction

In the 2016 NIH workshop on the ethics of nonhuman primate research, neuroscientist Prof. William Newsome made the case that not doing basic research would be a tragedy for the world. As reported by Science: “Even research that doesn’t have an immediate translation to people—like figuring out how the monkey brain works—is necessary, argued Newsome, because it could eventually lead to significant new knowledge that might improve human health. ‘It will be a tragedy for the world if we don’t leave room for basic science.’ Most attendees seemed to agree, with some stating that not doing research on monkeys was ethically indefensible because humans would suffer down the line.”

Evaluating the consequences of not doing something, or the harm of inaction, is difficult, but key to ethical consideration. Writing at Speaking of Research in 2012, Prof. Dario Ringach addressed the point in a post titled: “The morality of inaction: Reframing the debate,” writing:

Opponents of the use of animals in research challenge scientists and society as a whole to answer a simple question — How can we possible justify harming other living beings in the course of scientific studies?

In framing the moral debate with this question there is an implicit assumption that needs to be clarified. That is, opponents of the research assume that they do not need to provide a moral justification for their own position since they are not the ones harming animals.

However, there are many circumstances where choosing not to act turns out to be untenable.

If we were to find a toddler drowning in a bathtub, we would feel morally obliged to act and to save her life, particularly because doing so would not require us to assume any significant risk to ourselves.  Inaction in this case would be morally wrong and unjustifiable.  This illustrates the fact that inaction is not morally neutral; it requires justification.

In this regard opponents of animal research are avoiding an important question. What would be the consequence of inaction?  That is, what would be the cost to mankind of not doing medical research with animals?  And if the answer is that there is substantial cost in terms of lives and suffering, how can one justify not doing the work?

Some research opponents prefer to avoid the question altogether by denying that the research is relevant to human or animal health, implying that harm is done to animals without producing any benefits. However, no matter how many times animal activists attempt to rewrite medical history the facts are clear. Animal research contributed to the development of vaccines for polio, smallpox, diphtheria, tetanus, whooping cough, measles, rubella, mumps, hepatitis A/B, influenza, rotavirus, chickenpox, meningitis, human papillomavirus, all of which combined saved billions of human lives. Animal research also played an important role in the development of antibiotics, blood transfusion, lung surfactants for neonatal care, insulin, antidepressants, anti-retroviral therapy, and so on. The scientific consensus indicates that animal research is critical at the present time to advance medical research and human health.  Arguments along these lines fail.”

Patients, patient groups, and others affected by biomedical and behavioral research are important stakeholders in all of the questions we’ve outlined here. Their voices are particularly important to include in discussions that include weighing the harm of inaction. In fact, the point was drawn well in a debate between neuroscientist Tipu Aziz and animal rights philosopher Peter Singer. As we wrote previously: “While Peter Singer, author of Animal Liberation, has certainly supported principles of animal rights on a Utilitarian basis, his views on the matter are more nuanced than many activists give him credit for. In 2006, while debating the issue with Prof Tipu Aziz – a brain surgeon who has conducted pioneering work into deep brain stimulation – Singer agreed that Aziz’s use of monkeys could probably be justified. The Independent writes:

“In the film Monkeys, Rats and Me: Animal Testing, Singer is seen in discussion with the Oxford academic Professor Tipu Aziz, who has been conducting experiments on macaque monkeys as part of his work to find a treatment for Parkinson’s disease and other illnesses. Told by Aziz that tests on some 100 monkeys has led to positive treatment for 40,000 patients, Singer responds that he ‘would certainly not say that no animal research could be justified.’” (see and read more here and here).

5) Discussion of resources, infrastructure, animal care practices, reproducibility are all premised on the answers to the question of whether animals should be used by humans at all

The NIH review of chimpanzee research included consideration of animal housing and care, captured under the language “ethologically-relevant” conditions.  We, and others, have written extensively about that elsewhere. The 2016 NIH workshop on monkey research also included consideration of animals’ housing and care. The planning stage of the September 2019 NIH workshop appears to include some of the same.

Animal care, housing, psychological well-being and treatment are all relevant and core ethical considerations in weighing the use of animals in research. They are relevant to assessing potential harm in balance with scientific objectives and potential benefit.  But they are also a set of questions that arise if, and only if, the decision is made to house animals in captive settings. We have previously proposed a framework for broad consideration of housing animals in captivity (see below). The framework identifies four major choice points for decision making, starting with basic questions about whether captive populations of animals should be maintained at all, or increased at all. The subsequent questions—what purposes are permissible, and standards of housing and care, are premised on ethical consideration of human use of other animals within captive settings.

Similarly, previous NIH reviews have centered on analysis of the scientific need for creation of additional animals for research, or allocation of resources for that work. For instance, in a recent two-part analysis report, NIH surveyed research institutions and commercial facilities in order to better understand current and anticipated use of nonhuman primates in research (see here, here, and our analysis here). As mentioned above though, the question of how many, what kind, when and where, and for what type of research animals may be used are all questions that are secondary to the larger discussion about whether animal research should be conducted at all. Further, these are not questions that are specific to nonhuman primates. They could just as well be asked about fish, mice, or fruit flies. Nor are they questions specific to publicly funded research. They could just as well be asked about pharmaceutical companies and the use of nonhuman animals in both drug development and testing. Finally, they are not specific to animals housed and bred in captivity for research. They could be asked about animals in agriculture, entertainment, exhibition, sport, and companionship.

The upcoming NIH workshop is titled: “NIH Workshop on Optimizing Reproducibility in Non-human Primate Research Studies by Enhancing Rigor and Transparency.“ The NIH Office of Science Policy website contains the description in the image at the beginning of the post, with the workshop focused on: “build(ing) upon current reproducibility efforts by exploring best practices for enhancing rigor and transparency in studies specifically using non-human primate models. … exploring the role of ethical factors that may influence study design given that these models are of such critical importance to numerous research fields such as neurobiology, infectious diseases, regenerative medicine, reproductive biology, and more.”

Research methods, validity, rigor, and reproducibility are all topics that are of core interest to science and scientists. They are not, however, unique to animal research or to nonhuman primate research.  Nor are they a substitute for a serious, thoughtful and comprehensive discussion of ethical considerations. Each of these merits its own time and full development prior to putting the intersecting pieces together. Covering the full range in 2 days with a limited set of participants seems likely to leave everyone dissatisfied and again raise more questions and heat than answers or light.

Update 7/22/19: The posting of the videocast for NIH workshop previously planned for September 16, 2019 now reads “Cancelled.” https://videocast.nih.gov/summary.asp?live=33343&bhcp=1

The NIH Office of Science Policy page contains the following notice: “Please note that the “NIH Workshop on Optimizing Reproducibility in Non-Human Primate Research Studies by Enhancing Rigor and Transparency” originally scheduled for Sept 16-17, has been postponed and will be rescheduled.“