A wild Rhesus Macaque in IndiaWIKIMEDIA, J.M. GARGHarvard Medical School will have to pay a $24,036 fine violations of the Animal Welfare Act involving research primates at its New England Primate Research Center in Southborough, MA. The US Department of Agriculture (USDA) fined the school after a lengthy investigation of 11 violations that occurred between February 2011 and July 2012. In some cases, monkeys had to be euthanized after becoming dehydrated when employees of the research center failed to notice faulty watering devices or simply forgot to include water bottles in animals’ cages. One macaque died in April 2012 when it became entangled in a chain within its enclosure.
“The leadership of the school cares deeply about upholding exemplary standards of care,” read a statement from Harvard Medical School, released Wednesday (December 18).
Fining animal care facilities is a relatively rare enforcement tool used by the USDA, but some animal rights activists are decrying the size of the citation. “For an institution that receives $185 million annually in taxpayer funds alone, half of which is spent on animal experiments, a $24,000 fine for years of abusing and neglecting monkeys won’t motivate Harvard to do better by animals,” Justin Goodman, a spokesman for the People for the Ethical Treatment of Animals, wrote in an e-mail to The Boston Globe.
Harvard announced in April that it would be shuttering the New England Primate Research Center—which houses approximately 2,000 monkeys—by 2015, claiming financial reasons led to the decision.
By Stephen R. Latham
Every country’s law permits medical experimentation on animals. While some countries protect particular kinds of animals from being subject to experimentation—notably great apes and endangered species—very few place concrete limitations on what researchers may cause animals to suffer, given sufficient scientific justification. What laws do, instead, is establish standards for the humane treatment and housing of animals in labs, and they encourage researchers to limit or seek alternatives to the use of animals, when doing that is consistent with the scientific goals of their research. The result, of course, is that no existing regulatory scheme is satisfactory to opponents of animal research. The law, in their view, does nothing more than make the animal research scientist into a sort of James Bond villain: superficially polite, offering fine housing and well-prepared cuisine even to those whom he intends, eventually, to kill.
Of course, the goals of animal experimentation law seem much more reasonable if one accepts that research on animals is both important for medical progress and morally permissible. On those assumptions, it makes a great deal of sense for the law to aim primarily at limiting unnecessary animal suffering even as it licenses scientifically justified experimentation. U.S. law accepts those assumptions and adopts that aim.
The system that has evolved in the United States combines elements of sometimes competing regulatory philosophies. The result is a complex, multilayered system that addresses the most important concerns, but, partly because of historical accident, also leaves some gaps. Even proponents of medical research on animals can see obvious ways in which the regulatory structure could be changed to benefit animals. Perhaps more important, though, is the fact that the existing regulatory structure, imperfect though it may be, is elastic enough to accommodate substantial changes that could reduce unnecessary animal suffering.
Multiple Regulatory Approaches
Animal welfare laws must address three main ways in which unnecessary animal suffering can occur in the context of medical experimentation. First, such suffering can occur when a given research protocol is not well justified scientifically. An experiment that was so badly designed that it could never generate any useful scientific knowledge would never warrant animal suffering. Harder cases result when the amount of suffering is ratcheted down, or the experiment’s potential to generate useful knowledge is ratcheted up. A legal regime concerned with avoiding this kind of unnecessary suffering can opt to trust in the judgment of each individual research scientist, or empower someone besides the researcher to make at least some baseline assessment of the scientific value of each new animal research protocol. It can also provide information and guidance to researchers or overseers to improve their decisions.
Second, unnecessary suffering can occur when the amount of animal suffering induced by an experiment is not strictly required to conduct the experiment—perhaps because more animals are used than are necessary; or because less sentient animals could be substituted for more sentient ones, or computer or tissue models substituted for animals entirely; or because crude experimental procedures are producing avoidable stress or pain. A legal framework seeking to avoid these kinds of unnecessary suffering will encourage or require researchers to use the three Rs: reduce (the number of animals used in experiments), replace (animals with nonanimals, higher-order animals with lower), and refine (experimental procedures causing pain or distress).
Third, unnecessary suffering can occur outside the actual research protocol yet still in the research setting because of inappropriate animal handling, housing, and feeding practices. A legal regime seeking to avoid this kind of suffering will dictate humane standards for animal housing and care.
Given these goals, what sort of regulatory scheme would be best at realizing them? One can imagine a variety of available approaches, from strong, centralized state regulation and monitoring of all experimentation to a hands-off reliance on professional self-regulation among laboratory researchers. On the world stage, the United Kingdom is closest to taking the former approach, Japan to the latter. U.S. law falls somewhere in the middle, in part because U.S. law in this area is in fact the result of a gradual, decades-long merging of the government regulatory and professional self-regulatory approaches.
The government regulatory approach is embodied in the sprawling, strange, and often amended Animal Welfare Act of 1966. In its original form, the AWA was designed to control pet breeding and sale practices; it was passed, in part, as a result of public outcry about the mistreatment of dogs sold to laboratories. As amended, it governs the treatment of animals in a wide range of settings, from pet shops to circuses and from zoos to laboratories. Its enforcement is delegated to the U.S. Department of Agriculture’s Animal and Plant Health Inspection Service, whose inspectors make unannounced site visits to research facilities. Violations uncovered on such visits can result in fines and even, in extreme cases, criminal prosecution. The most common complaint about enforcement under the AWA is that it is rigid and mechanistic.
Because of its historical roots in concern for pets, the AWA’s reach is confined to warm-blooded animals, and it contains special regulations addressed to certain animal favorites: dogs, cats, rabbits, and monkeys. Its animal experimentation regulations apply to any school or research facility that purchases or transports live animals in interstate commerce or that receives federal funding. But in fact the law has never reached the bulk of warm-blooded animals actually used in research. Concern about high regulatory costs—and about possible delay in creating guidelines for other, more popular animals—led the USDA to exclude laboratory rats and mice from its oversight from as early as 1970. In spite of lobbying efforts in the 1980s by proanimal groups, a congressional amendment to the AWA in 2002 legally formalized the agency’s longtime practice, excluding rats, mice, and birds from the definition of “animal.”
In general, the law and its implementing regulations have focused on setting demanding, detailed standards for animal housing and basic standards for pain control. It supports only minimal review of the scientific merit of research protocols, but it requires researchers to make efforts to “reduce, replace, and refine.”
The self-regulatory approach to animal research regulation is embodied in the National Institutes of Health’s Guide for the Care and Use of Laboratory Animals.The Guide has existed in some version since 1963, when it was introduced as a voluntary set of professional standards for laboratory animal research. Today, the Guide’s standards are mandatory for all research facilities receiving federal funds. The Guide covers the treatment of all vertebrates, which means that, at least in federally funded research, it closes many of the gaps left open by the AWA. Not only are rats, mice, and birds covered, but also cold-blooded vertebrates like zebra fish—currently the go-to animal for laboratory studies of pain and nerve function.
The change in the Guide’s status to a rulebook has altered its content somewhat. Earlier editions’ expansive aspirational goals have given way in later editions to more readily applicable rules. There has also been considerable pressure to get the AWA’s regulatory requirements and the Guide’s standards to match, since all federally funded researchers are bound by both. Indeed, today, the two sets of standards are, if not identical, at least compatible with one another. But in general, where the AWA regulations are more rigidly prescriptive, the Guide permits lab veterinarians to use their professional judgment in applying general standards to particular species or protocols.
Clearly there is room for reform. If the AWA were amended to include rats, mice, and birds, for example, that would be a major step toward ensuring the humane treatment of all animals in public and private labs.
Federal standards are full of specific requirements for different kinds of studies, but in general, it is fair to say that they offer the most concrete guidance on questions of animal housing and care. The regulations include detailed discussions of square footage, exercise requirements, room temperature, and more. Considerably less guidance is offered on issues of protocol evaluation and implementation of the three Rs.
Of course, this is exactly what might be expected given the incredible volume and variety of animal research in the United States. A central authority can say a lot about how to house and feed monkeys, mice, and zebra fish, and expert advice on those issues will apply to all monkeys, mice, and zebra fish in every lab, no matter what protocols they are being used for. But questions about the other possible sources of unnecessary animal suffering—the scientific justification of a given protocol, or the ways in which animal suffering connected to a given protocol might be avoided or reduced—are too numerous and varied to be answerable in advance by a central authority. With regard to those highly fact-specific questions, U.S. law relies on the expert judgment of local IACUCs.
It is no coincidence that this kind of reliance on decentralized expert committees is also the salient feature of U.S. law governing research on human subjects. The federal Common Rule, faced with a similar diversity of research protocols to evaluate, regulate, and modify, uses the same tactics as the AWA: it mandates creating research oversight committees (institutional review boards), specifies that their membership should include both relevant expertise and community representation, and empowers them to make and enforce a range of judgments about particular experimental protocols.
While the many IACUCs are expected to exercise independent judgment with regard to the scientific issues brought before them, the U.S. government does its best to inform the judgment by providing them with educational resources. The Public Health Service and the Department of Agriculture Web sites are full of guidance documents and educational resources for laboratory researchers and for IACUC members. There are documents, for example, with specific ideas about how and when to substitute lower-order animals for higher-order animals, and other documents providing up-to-date scientific news about newly developed computer models that can substitute, in some cases, for animal experimentation.
Finally, just as in the human subjects research world, federal regulations are quite commonly supplemented by private education and accreditation. Many research facilities seek accreditation by the Association for the Assessment and Accreditation of Laboratory Animal Care, a professional association of veterinarians and laboratory scientists. AAALAC provides education and does prearranged site inspections of labs once every three years. Educational and inspection standards are built largely around the requirements of the Guide, and the NIH accepts AAALAC accreditation as prima facie evidence of a facility’s compliance with the Guide’s requirements.
Toward Reform: Accountability, Uniformity, Balance
The system of decentralized oversight by local IACUCs has several obvious advantages: it permits oversight by people with knowledge of the local researchers and laboratory facilities; it allows IACUCs to develop specialized knowledge, well tailored to the research being done at their facilities; and it is likely more speedy than any alternative program of centralized governmental research oversight would be. On the other hand, the decentralization of oversight has given rise to a number of problems—which, not surprisingly, are similar to those that beset the IRB system in human subjects research.
First, there is a problem of transparency and accountability. IACUCs are for the most part fairly anonymous. Hardly anyone not directly involved in animal research knows that they exist, much less who their members are. And of course, their members are not elected or in any other way publicly accountable for the decisions they make. Most IACUC decisions do not take the form of opinions or any other form of substantive, publishable decision, but of recommendations to researchers for piecemeal alteration of protocols. A central repository of IACUC minutes, and of policies adopted by different IACUCs, might both increase accountability and stimulate new ideas by creating cross talk between IACUCs. But any such repository would have to be created with an eye toward preserving researchers’ intellectual property.
Second, decentralization almost necessarily gives rise to a lack of uniformity in decision-making and in quality of research oversight. One IACUC may conclude that a protocol involves unnecessarily harsh treatment of animals or presents an opportunity for substitution of nonanimal models; another may view the original protocol as unproblematic and requiring no amendment. A number of studies have shown that similar protocols are treated quite differently by different IACUCs. It is unclear what the implications of such findings are. Do they reveal that IACUCs have differing standards relating to animal welfare? That they judge similar protocols differently when they are presented by different researchers? Or some combination of these factors? In any case, enforced uniformity across IACUCs is a dangerous solution to propose for the problem of varying standards, in the absence of clear knowledge about whose standards are appropriate—and whose would be enforced.
A third complaint about the decentralized approach to animal-research regulation involves the perception that the U.S. government is too deferential to local IACUCs and does not take the task of auditing labs sufficiently seriously. In the early 2000s, there were some high-profile allegations made by whistleblowers from the USDA’s Animal and Plant Health Inspection Service (APHIS) that audit findings were deliberately being watered down to be less critical than the field officers originally intended them to be. U.S. audits of APHIS confirmed allegations of lax auditing in some regions of the country. The obvious reform here is to better fund and train both the regulatory overseers and those who audit their performance.
There are other important criticisms of the U.S. regulatory regime not directly connected to its choice of decentralized decision-making. First, there is the question of scientific justification for animal suffering. The AWA does not ask IACUCs to balance animal suffering against the scientific merit or promise of any given experiment. Instead, it asks IACUCs to ensure only that any given protocol has scientific merit and that any animal suffering the protocol induces is strictly necessary to that science. The result is that any study that will advance science, even in a very small way, can be used to justify tremendous amounts of animal suffering, as long as the suffering is necessary to the advance. Though they do seek to modify studies via use of the three Rs, IACUCs almost never reject protocols.
Finally, and most importantly, there is the issue of which animals are protected. As already mentioned, the hundreds of thousands of rats, mice, and birds used in private, nonfederally funded labs are not subject to any federal regulation. (Some individual states’ anticruelty statutes may apply in some cases, but there is very limited case law in the area.) Excluded, also, are cold-blooded animals. This means that there is no federal legal pressure on private firms such as drug companies to reduce or refine animal use, or to replace animals with computer or tissue models—a strategy that may be particularly feasible in studies of toxicology or drug metabolization.
Even in federally funded facilities, the living conditions of rats, mice, and birds are not subject to the USDA’s APHIS inspection; only in AAALAC-accredited facilities is there oversight beyond self-reporting, and AAALAC does scheduled inspections only once every three years. Rats and mice, it should be stressed, are the most commonly used laboratory animals. In addition, U.S. law offers no protection for invertebrate, cold-blooded animals such as cephalopods. By contrast, Europe has recently moved to protect cephalopods in light of their manifest intelligence and sentience. Nor does U.S. law prevent research on great apes, or ban (though it does regulate) the use of wild-caught animals. And the United States is one of only two governments in the world that still permits invasive research on chimpanzees, though the scope of federal funding for chimp research has recently been sharply limited. (See “Raising the Bar: The Implications of the IOM Report on the Use of Chimpanzees in Research,” in this volume.)
Clearly there is room for reform. Some needed reform involves stepping up research oversight. If the AWA were amended to include rats, mice, and birds, for example, that would be a major step toward ensuring the humane treatment of all animals in public and private labs. In addition, the inspection rate for facilities could be more frequent. Publicly funded U.S. labs are inspected by APHIS about once a year, by their own IACUCs twice a year, and by AAALAC (if they choose to be AAALAC-certified) once every three years. Compare this to the U.K. system of inspecting about once a month. Other reforms could involve improving rigid and not-terribly-useful existing regulations, like cage-size requirements currently based on animals’ body size rather than on their behavioral needs. Most significantly, the law could be reformed to permit a more explicit balancing of harms to animals (including both suffering and death) against the scientific gains at which the research aims. Empowering IACUCs to engage in such balancing is hardly radical; IRBs, for example, are already empowered to engage in such balancing in the human subjects research area, and this has not caused research to grind to a halt. Such a reform would require us to confront directly the question of how much suffering humans can impose on other species in return for small but real gains in knowledge.
Finally, a great deal can be accomplished even within an unchanged legal regime. The most urgent need is for more to be done to implement the three Rs. The familiar calls for better education about replacement techniques and more aggressive IACUC intervention on behalf of reduction and refinement are, of course, well justified. But even more dramatic reduction might be achieved if the goal of reduction were pursued not only within but also across protocols. There might be significant gains from putting animal-sharing procedures in place at the institutional level. At the moment, animals are commonly euthanized whenever the particular research project they’re involved in comes to an end, without regard to the animal’s age or health status. If a protocol involves attempts to breed, for example, mice with particular genetic traits, the pups born without those traits are routinely euthanized. If research facilities could work with researchers to use healthy animals from one study in another, rather than default to their euthanization, then fewer animals would need to be bred for suffering.
Stephen R. Latham is director of the Interdisciplinary Center for Bioethics at Yale University. He has published on a broad range of issues at the intersection of bioethics and law. He is a former board member of the American Society for Bioethics and Humanities, a former graduate fellow of Harvard’s Safra Center on Ethics, and a former research fellow of the University of Edinburgh’s Institute for Advanced Studies in Humanities. His current research includes a project funded by the Robert Wood Johnson Foundation to create a database of state statutes and cases criminalizing HIV exposure and a project on a legal framework for newborn whole-exome screening.