Respect for Persons (Autonomy) and Informed Consent

Within the Belmont Report, “respect for persons” incorporates two separate but related moral requirements: to acknowledge autonomy and to protect those with diminished autonomy. Though it is possible that the Commission had other intentions in selecting the principle of “respect for persons,”^{Footnote 32} its documented intent “requires that subjects, to the degree that they are capable, be given the opportunity to choose what shall or shall not happen to them.”^{Footnote 33}

Autonomy commonly refers to the capacity for self-determination. It is both a capacity for self-rule, and an ability to protect or further one’s own interests. As a result, it is intimately related to vulnerability. Subjects who lack autonomy lack the ability to adequately represent their own interests, which makes them vulnerable to exploitation and harm.

There are many open questions about the degree to which humans and other animals display autonomy or autonomous action.^{Footnote 34} It is now generally accepted that animals—including chimpanzees, crows, elephants, macaques, orcas, and wolves, among others—form daily goals and order their behavior accordingly.^{Footnote 35} Alliances or coalitions, displayed by animals such as chimpanzees, dolphins, and others, reflect goals and planned behaviors that extend beyond momentary concerns. Older members of these animal communities behave in ways that satisfy their preferences, which effectively track what they value and their basic interests.^{Footnote 36} Many animals also provide for their young, indicating that they also track the preferences or interests of animals in their care.^{Footnote 37} These behaviors and attendant capacities evidence a capacity for self-determination. In humans and animals, the capacity for self-determination can change acutely or chronically over the course of a lifetime as a result of neurological, psychological, and emotional maturation or degeneration, or other changes in aptitude. Nonetheless, self-determination, of varying degrees, is essential to wellbeing in humans and many animals, and interference with self-determination can lead to physical and mental distress, as well as disorders such as learned helplessness, posttraumatic stress disorder, and depression.^{Footnote 38}

As the Belmont Report indicates, autonomy is also characterized by an ability to make informed decisions free of undue influence. External “circumstances that severely restrict liberty,” as in the case of prisoners, can adversely impact one’s ability to exercise autonomy.^{Footnote 39} Sources of fear and other forms of intimidation, such as captivity, can limit the expression of self-determination.

Though it is important to acknowledge capacities like autonomy and their implications, as the Belmont Report points out, it is as important to protect those with diminished autonomy, whether such diminishment is due to factors that are intrinsic or extrinsic to the individual. In such cases, two general rules apply. To the degree they are able, even individuals with compromised autonomy should be provided opportunities to choose what will or will not happen to them—a standard of welfare that is also increasingly recognized as applicable to animals.^{Footnote 40} However, to the degree that individuals are incapable of making informed decisions, they should be protected from harm.

The application of respect for persons in human research typically takes the form of seeking and securing informed consent before inclusion in a research study. Informed consent requirements generally include three elements: the delivery of sufficient information, comprehension, and voluntariness.^{Footnote 41} Guidelines provide opportunities for those with diminished capacities to represent their own interests in the decisionmaking process by respecting “the objections of…subjects,”^{Footnote 42} allowing individuals to dissent even if they are not deemed to have the capacity to provide informed consent. In some cases, surrogate decisionmakers charged with choosing what is in the potential research subject’s best interests may offer permission on the individual’s behalf. Regardless of the circumstance, coercion, in which an individual is threatened with harm, and undue influence are prohibited: “Inducements that would ordinarily be acceptable may become undue influences if the subject is especially vulnerable.”^{Footnote 43}

There remain unanswered empirical and conceptual questions about autonomous capacities in animals. Human-animal interactions, and the crudeness of our understanding of animals, are unlikely to permit meaningful informed consent. Therefore, it is important to consider whether concepts used in human research such as dissent, assent, and surrogate decisionmakers are sufficient for formulating decisions about the use of animals in research.

Within pediatric research ethics, researchers and regulators increasingly recognize that the sustained dissent of even very young children should be respected, unless perhaps the study promises direct benefits that are otherwise unavailable. This requirement is motivated by obligations to avoid harm associated with disrespecting the corresponding level of self-determination of the child.^{Footnote 44} Elsewhere, requisite capacities needed to express sustained dissent have been described.^{Footnote 45} Briefly, young children who express sustained dissent (1) have the capacities to experience pain or distress; (2) can anticipate its occurrence; and (3) can express a preference that the pain or distress stop (or not occur). Many animals also possess these capacities, as evidenced by relevantly similar neurological structures, escape or avoidance behaviors, and the expression of behavioral preferences. However, an absence of escape or avoidance behaviors does not necessarily indicate an absence of dissent.^{Footnote 46} Like children, animals are more likely to adapt or suppress certain behaviors in an attempt to protect themselves from individuals in a position of power or influence—further underscoring their vulnerability. Captive conditions and other practices used during the course of laboratory research interfere with animals’ capacities to express dissent, deepening their vulnerability to manipulation and coercion. Therefore, in the current context, absence of sustained and recognizable indicators of dissent alone is generally not a sufficient criterion for permitting research with animals to proceed.

There are unresolved questions about whether animals can express assent in ways humans reliably understand.^{Footnote 47} Assent, compared with dissent, may be considered a more morally rigorous requirement for participation in research or other activities. As an affirmative expression of agreement, assent implies active participation, whereas an absence of dissent can suggest passive or confused participation and therefore less certainty about a willingness to participate. However, in most cases involving animals, it is unlikely that, even in the best of circumstances, they could make an informed decision about the potential risks or benefits of research generated by perceived human needs.

The potential use of surrogate decisionmakers in animal research also presents special difficulties. First and foremost, consent from human surrogate decisionmakers is always suboptimal compared with consent from a capable research subject. However, current regulatory guidelines involving individuals who cannot provide consent (e.g., young children) allow for the input of a surrogate decisionmaker (e.g., at least one parent) who can provide or refuse permission on the child’s behalf. In these cases, respect for sustained dissent and assent of the child (if they are capable of providing it) are also required. However, as outlined above, the case of animals raises particular challenges related to dissent and assent. Since convincing arguments for why nonbeneficial research is acceptable in those who cannot provide consent are lacking, a conservative standard is required.

In practice, respect for the autonomy and vulnerability of animals would require respecting their freedom and choices to the extent that they are capable of asserting their decisions, and to the extent we are capable of understanding their choices. At the very least, the following criteria would be required: an absence of punishment, coercion, restraint, or confinement; real options and alternatives appropriate to the individual and their normal circumstances; and surrogates who are not unduly compromised in their capacity to represent the animal’s interests by their own personal or professional conflicts of interest. Surrogate decisionmakers should also be capable of making an independent, informed decision free of coercion or undue inducement. If a surrogate cannot reliably understand the cognitive, ethological, and ecological factors required for the health and wellbeing of the individual animal, and impartially represent their individual interests, a cautionary approach would require avoiding enrollment in research.

Beneficence and Assessment of Risks and Benefits

According to the Belmont Report, the principle of beneficence is interpreted as an obligation. Beneficence is often taken to mean “doing good,” but in the Belmont report, the principle is based on two general rules: (1) do no harm (nonmaleficence; a principle believed by many scholars as also embodied in the Hippocratic Oath); and (2) maximize possible benefits while minimizing possible harms.^{Footnote 48} These rules, interpreted literally, would clearly be in tension, as there is a difference between not harming and merely minimizing harm (or risk of harm). However, in human research, the principle of beneficence is operationalized through a requirement that possible risks to participants are minimized and that there is a favorable benefit/risk ratio for participants.

The dual requirement that harms are minimized and that there is a favorable benefit/risk ratio is noteworthy, since either rule by itself would be insufficient. For example, if the only requirement for the principle of beneficence were that the risk of harms be minimized relative to proposed research, some human research could be deemed permissible even when potential harms significantly outweigh potential benefits. Conversely, having only a requirement for a favorable benefit/risk ratio without an obligation to minimize harms would allow research that included preventable or unjustifiable harms to proceed. Thus, it is only in tandem that these two rules (when properly executed) ensure that undue and preventable harms are avoided.

The existing regulatory situation is quite different for animals. For example, the 3Rs framework utilized by most institutional review committees often leads to claims that animal research does require benefit/risk analysis, since funding sources are expected to assess benefits in their funding decisions and institutional review committees attempt to reduce harms associated with specific protocols.^{Footnote 49} However, in the absence of a direct comparison of potential harms and benefits, problematic research resulting in significant harms to animals can readily proceed. This problem is particularly exacerbated in exploratory research involving animals, in which purported benefits may only amount to faint promises of discoveries. In practice, harms to animals are afforded comparably little weight.^{Footnote 50} The situation reflects a problem that the Belmont Report (and subsequent regulatory policies) avoided by including a dual criteria notion of beneficence.

The IOM Committee’s finding regarding chimpanzees provides a clear example of the failings of the current approach used in animal research. The Institutional Animal Care and Use Committees (IACUCs) at institutions performing research with chimpanzees completed their official duties in reviewing research protocols, minimizing harms relative to proposed protocols. Nevertheless, most or all of federally funded chimpanzee research was ultimately deemed unnecessary, and presumably scientifically and ethically unjustified, by the IOM Committee. In this case and others, the absence of a direct comparison of potential harms and benefits, or a pro forma exercise to meet review committee expectations, would continue to permit ethically problematic research.

Though duties of nonmaleficence and beneficence are alluded to in some animal research guidelines, current provisions do not afford sufficient attention to these principles. For example, the most recent edition of the Guide for the Care and Use of Laboratory Animals requires that, for certain research, “the IACUC is obliged to weigh the objectives of the study against potential animal welfare concerns.”^{Footnote 51} Recently, American and European organizations overseeing the use of animals in research released a two-part set of recommendations for benefit/risk analysis in animal research.^{Footnote 52} However, as noted by Anna Olsson and colleagues, the report “has no legal status and is less than clear on a number of points.”^{Footnote 53} Individuals responsible for reviewing protocols involving animals still have no clear framework by which to prohibit research with an unfavorable benefit/risk profile. As a consequence, full rejection of a protocol is rare; instead, protocol review commonly focuses on modifications along the lines of the 3Rs that may still fail to create a favorable benefit/risk profile. Sincere attention to the principle of beneficence in animal research would, in contrast, necessarily include: (1) a full assessment to the extent possible of actual and potential harms; (2) a full assessment of potential benefits; and (3) an approach that allows for a transparent and rigorous direct comparison of the risk of harms and potential benefits, along with conditions on what benefit/risk profiles are acceptable.

Within human research, the principle of beneficence requires that the nature and scope of physical, psychological, social, and other harms be considered. However, consideration of a full scope of harms is far less common or perhaps even possible in the general practice of animal research. If considered, risks associated with animal research would include, but not be limited to: harms associated with breeding and transport; separation from parents, siblings, and other conspecifics; an inability to control access to that which meets one’s own basic needs; a lack of safety and security; thwarting of an animal’s preferences; an inability to fulfill one’s full developmental potential; deprivation of a normal, species-appropriate, natural environment; and being killed.^{Footnote 54}

In particular, confinement and captivity in themselves introduce significant risk.^{Footnote 55} Certain forms of captivity are inherently harmful to animals because these conditions deprive them of ethologically appropriate environments and activities, of liberty and movement, and of basic psychological and social needs, including stimulation, play, and family. Some forms of captivity also impose experiential harms, including stress, boredom, pain, fear, and loss. Currently, within the practice of animal research, the risk—that is, the magnitude and probability—of all of the aforementioned harms is often severe and certain.^{Footnote 56}

Within human research, measured risks and benefits may also include those that affect “the families of the individual subjects, and society at large (or special groups of subjects in society).”^{Footnote 57} Similarly, for animals, in addition to the harms to prospective subjects, other potential harms should be considered. For instance, in the case of free-ranging animals or descendants of ‘wild-caught’ animals, the direct environmental costs as well as the indirect environmental costs of supporting these practices should be included in any risk assessment.

In contrast to much research on humans, it is rare for justifications offered for animal research to be focused on benefits to individual animals included in experiments.^{Footnote 58} No clear moral justification for this position is offered in existing laws, policies, or guidelines, though nonspecific, perceived future benefits to humans are often cited as a rationale for the use of animals in research. As in the case of human research, it is important for institutional review committees to ensure research proposals do not exaggerate potential benefits of research. Though it is impossible to predict the potential value of various scientific discoveries, the possibility of unexpected future benefits also exists for research involving humans. Nonetheless, such uncertainty has not prevented ethics committees from requiring a realistic assessment of potential benefits as part of the approval process for human research proposals. Similarly, uncertainty should not be an impediment to conducting a realistic assessment of potential benefits (or risks) for research involving animals.

The Belmont Report also goes farther, suggesting “the assessment of risks and benefits requires a careful arrayal of relevant data, including, in some cases, alternative ways of obtaining the benefits sought in the research.”^{Footnote 59} Even if research has in the past resulted in a particular benefit such as an advancement of knowledge, it is in many cases possible that the benefit could have been achieved through some other less harmful means. It is similarly important for an assessment of benefits of research on animals to accurately reflect whether such benefits could only be procured through the proposed research, or whether anticipated benefits could be achieved via alternative methods now or in the future. ^{Footnote 60}

Identifying an appropriate approach for a rigorous comparison of potential harms and benefits is arguably the most conceptually challenging aspect of benefit/risk analyses. Even in the case of inter-human comparisons, there is rarely agreement about how to weigh particular risks against particular potential benefits. For individuals who can provide consent, individual differences influence a willingness, or lack thereof, to undergo specific risks in anticipation of potential benefits. The case is made more difficult when an individual is unable to provide consent. Nonetheless, even when conflicts occur between commitments to the production of societal benefits and risks of harms to individuals, human subjects are commonly protected from serious harms including severe pain, psychological distress, permanent disability, suffering, death, and prolonged captivity—factors which are commonly experienced by many animals used in research. A commitment to nonmaleficence would require that animals be similarly protected from serious harm. However, in current practice, potential benefits to humans, and even the scientific question itself, are given much more weight than known or potential harms to animals. As other authors have argued, a principle of equal consideration of interests would require that harms to animals should be weighed equivalently to relevantly similar harms in humans.^{Footnote 61} Likewise, potential harms to humans from ungeneralizable or misleading findings should also be included in a full benefit/risk assessment.^{Footnote 62}

Justice and the Selection of Subjects

“Who ought to receive the benefits of research and bear its burdens?”

This is the question of justice posed by the Belmont Report.^{Footnote 63} Attention to justice emerged as a result of perceived injustices associated with the systematic selection of individuals for research because of their easy availability, socioeconomic status, compromised position, or manipulability. For example, in the nineteenth and twentieth century, economically disadvantaged patients seeking healthcare within public institutions bore the burdens of research, whereas wealthier private patients reaped its benefits. Later in the twentieth century, researchers systematically targeted prisoners in Nazi concentration camps, African American men living in Tuskegee, Alabama, the Havasupai of the Grand Canyon, hospitalized children with cognitive disabilities, and other vulnerable or exploited populations.^{Footnote 64} Though the harms imposed on these disparate populations ranged from deceit to torture, each abuse has contributed to an historical background for how “conceptions of justice are relevant to research involving human subjects.”^{Footnote 65}

Consideration of the principle of justice within the Belmont Report has given rise to moral requirements for fair procedures and outcomes in the selection of subjects, including attention to social justice:

Generally, individuals, especially those with impaired autonomy, should not be subjected to risks disproportionate to the benefits of the research, and if individuals are members of groups unlikely to benefit from the research, such individuals may need to be excluded from research altogether.^{Footnote 67}

Currently, animals are deliberately and commonly chosen for research as a result of their easy availability and manipulability, as well as institutional and cultural biases in society. Like human populations who were historically and systematically targeted, animals also have a “frequently compromised capacity for free consent.”^{Footnote 68} Animals overwhelmingly bear the burdens of research without accessing the benefits. Though vulnerable human populations, including children and prisoners, are protected as a result of their compromised abilities to provide informed consent, animals are currently excluded from similar protections despite possessing similar vulnerabilities.

The concept of vulnerability is widely used but not well-defined in ethical guidelines. The Belmont Report, like many other statements, guidelines, and codes for research ethics, does not define vulnerability. It lists some of the groups or individuals who can be considered vulnerable subjects, including “racial minorities, the economically disadvantaged, the very sick, and the institutionalized,”^{Footnote 69} all of whom, historically, were frequently exploited because they were oppressed, easy to access, powerless, or had no one to advocate for their rights and protection. The Common Rule also lists prisoners, children, and pregnant women and their fetuses as groups requiring special protections.

The Council for International Organizations of Medical Sciences (CIOMS) provides perhaps the most comprehensive definition of vulnerability to be found in research codes.^{Footnote 70} Vulnerability is defined as either (1) lacking the traits of an autonomous person (the incapacity to protect one’s own interests, i.e., vulnerability is defined as intrinsic to an individual), or (2) being unable to act autonomously due to constraints on one’s freedom to act (the inability to protect one’s own interests, i.e., vulnerability is defined as the result of extrinsic conditions). A previous iteration of the CIOMS guidelines stated that “vulnerable persons are those who are relatively (or absolutely) incapable of protecting their own interests. More formally, they may have insufficient power, intelligence, education, resources, strength, or other needed attributes to protect their own interests.”^{Footnote 71}

Under this substantive understanding of vulnerability, it is clear how both children and prisoners are vulnerable. The former group is primarily vulnerable because they fundamentally lack the ability to comprehend and consent to research, or to advocate for their own interests, and the latter is primarily vulnerable because the very conditions of captivity and institutionalization produce external constraints on their autonomy and ability to make free, uncoerced choices and protect their own interests.

Recent work attempting to better understand the concept of vulnerability, as it applies to humans in research and beyond, can be used to further demonstrate how animals are vulnerable subjects in the context of research.^{Footnote 72} Catriona Mackenzie, Wendy Rogers, and Susan Dodds have developed a taxonomy that captures the variety of ways in which the concept of vulnerability is deployed.^{Footnote 73} They identify three types of vulnerability: inherent, situational, and pathogenic. Inherent vulnerabilities are intrinsic features of existence to which humans and animals are susceptible as biological creatures (e.g., illness, disease, accidental injury, pain, suffering, and death), while situational vulnerabilities are extrinsic and arise from the context or situation in which individuals or groups find themselves. Pathogenic vulnerabilities form a subset of situational vulnerability and can be brought about either through morally dysfunctional relationships, or when attempts to remedy existing vulnerabilities instead worsen them.

Animals in research are vulnerable in all of these ways. As biological creatures they are inherently or intrinsically vulnerable; their embodiment, basic survival needs, and more complex needs endow a susceptibility to harm that can be realized during the course of research. As in the case of humans, animals’ experiences matter to their wellbeing. Since many animals have complex needs and are susceptible to suffering if their needs aren’t met, a just approach necessitates that their fundamental needs are treated as similar to those of humans—with full and equal consideration. Treating these animals otherwise, simply because of their easy availability and convenience, undermines obligations to justice.^{Footnote 74}

The treatment of animals in society, including in the current research environment, also exposes animals to situational vulnerability. Even animals with the potential to develop the cognitive capacities necessary to make autonomous decisions and protect their own interests are typically deprived of these capacities in captivity, for the same reasons that involuntarily captive and institutionalized humans lack them: the conditions of captivity act as an external constraint on their abilities to act to protect their own interests. Such constraints include their complete dependence on their captors, as well as the use of coercion and inducements to attain compliance or cooperation.

Animals are also pathogenically vulnerable. Motivations for including them in research reveal at least one source of their pathogenic vulnerability. Animals are often selected as research subjects because of asymmetrical power relations, dependency, and species-based discrimination. Human institutions systematically embed a failure to recognize many animals as entities with intrinsic value, instead treating them as expendable instruments for human convenience. The increasing use of genetically modified animals in research is a particularly profound example of pathogenic vulnerability. Humans develop and breed these genetically modified individuals in an attempt to replicate or simulate a human medical condition, with the intention of subjecting them to the harms of research.

In light of the dependent position of many animals in society, their intrinsic and extrinsic vulnerabilities, and an inability to provide informed consent, such animals should be treated as vulnerable subjects. Within human research, vulnerable subjects typically cannot be exposed to risks exceeding those encountered in normal daily life in safe environments (i.e., minimal risk) or a minor increase over minimal risk, particularly when there is no direct benefit to the individual. The Belmont Report does not provide direct ethical guidance for nontherapeutic research with vulnerable subjects, such as children, that exceeds a minimal risk standard.^{Footnote 75} The Common Rule restricts such research, but does not prohibit it so long as it is likely to result in generalizable knowledge about a “serious problem affecting the health or welfare” of the individual’s “group” (in the case of a child, the group would be children with the condition targeted by the research).^{Footnote 76} Studies that involve more than a minor increase over minimal risk, with no prospect of direct benefit for the individual who cannot provide consent, must be referred to the federal government, which can convene an expert panel to review the research proposal. If approved, such studies typically require multiple permissions, including the permission of each parent and the assent of the child in the case of research involving children. Some have proposed that a similar system could be invoked for animals.^{Footnote 77} However, such approaches may not be morally sufficient to guide decisions about the use of animals in research.

Animals are in some ways more vulnerable, as individuals and as a class, than human populations who cannot provide informed consent. For example, whereas a parent or guardian who represents a child’s best interests can provide permission on their behalf (sometimes with a child’s assent), it is much more difficult for a human to represent the interests of an animal, since many humans seem not to understand the needs, desires, or interests of other animals as well as they can (presumably) understand other humans.^{Footnote 78} There are possible exceptions to this, such as companion animals, whom conscientious guardians may know well.^{Footnote 79} However, generally, these differences suggest that the risk threshold for research involving animals should be as low as, or lower than, the risk threshold mandated for vulnerable human populations, including children.^{Footnote 80} It would also suggest a need for greater certainty that the risk threshold is not exceeded. A cautious view would require that if animals cannot consent or assent, almost no form of known harm—particularly serious harm—is acceptable. Permissible research might include that which might provide a direct benefit to the individual (with a favorable benefit/risk profile), in the absence of dissent, as well as permission from an appropriate surrogate decisionmaker. As with surrogate decisionmakers for humans, knowing and understanding an individual’s interests, needs, and preferences is necessary but not sufficient to represent that individual’s best interests.