∗University of Pennsylvania
‡California State Polytechnic University
Throughout history, people have used animals—whether for food, fiber, sport, adornment, labor or companionship—as a means of satisfying human ends and interests. But animals also have interests—in avoiding pain, fear, distress or physical harm, and in pursuing their own needs, desires and goals (DeGrazia, 1996). Relations between human and non-human animals become morally problematical where there is a conflict of interests between the two: where the human use of the animal either causes the latter pain, fear or harm, or it in some way thwarts or prevents the animal from satisfying its own needs and goals.
During the last 20 years, purveyors and proponents of animal-assisted interventions (AAI)1 such as the Delta Society have made concerted efforts to professionalize the “industry,” and establish selection and training standards that aim to minimize the risks of harm to all concerned, including the animals (Hines and Fredrickson, 1998). However, the field has experienced explosive growth in recent years and, in many cases, these standards have been established in the absence of any systematic or empirical evaluation of the potential risks to animals imposed by current practices. Indeed, the general but unsubstantiated feeling across the industry is that these are “good” activities for animals to be engaged in. The fact that a large number of animals fail to respond to the nurturing and training they receive has not generally been taken as evidence that they do not want to, or are unable to, participate. Instead, practitioners tend to respond to failure by changing the selection or the training procedures, as if the animals are theoretically capable of responding positively to any demands made of them. Indeed, with the laudable exception of some contributions to this volume, it is unusual to find more than a passing reference to animal welfare concerns in any recent review of risk factors in the AAI field (see, e.g., Brodie et al., 2002).
Much of the rest of this Handbook has been devoted to demonstrating how the use of therapy and assistance animals significantly enhances human health and well-being. The primary question we address in this chapter is whether this end morally justifies the means of achieving it. Specifically, our goal is to re-examine the animal/human partnership from the animal’s viewpoint to see what the benefits might be for the animal, or to see if the raising, training, and use of therapy and assistance animals is causing significant degradation in their welfare. In doing so, however, we recognize that there is a serious shortage of reliable scientific evidence to reinforce some of our claims and suggestions.
Behnke (2005) points out that the ability to use ethical judgment and discretion are defining features of the helping professions, and that arriving at an ethical course of action implies weighing and balancing the values of professional knowledge and realizing that complex situations do not necessarily result in simplistic solutions. With these observations in mind, we urge practitioners and clinicians to examine their ethical responsibilities for the welfare of their therapeutic adjuncts, and insure that the safety and well-being of these animals are safeguarded at all times.
23.2 General welfare considerations
The term “animal welfare” generally refers to the state of an animal, and the extent to which it can be said to be faring well or ill in a particular situation or at a particular point in its life. Different animal welfare experts tend to give priority to different aspects of an animal’s state when assessing its welfare: some emphasize unpleasant or pleasant subjective feelings (Boissy et al., 2007; Dawkins, 1980; Duncan, 1993), while others focus on the animal’s ability to express “natural” or species-typical behavior (Rollin, 1995), or its capacity to adapt to, or cope with, the demands of its environment (Broom and Fraser, 2007). An animal’s welfare can also be said to be broadly synonymous with its quality of life (Duncan and Fraser, 1997). For the purposes of this review, we will consider welfare as comprising all of these different ideas and concepts.
23.2.1 The Five Freedoms
Although different kinds of welfare problems attend the specific roles and activities performed by different classes of therapy or assistance animal, there are some basic welfare considerations that tend to apply to all of them regardless of how they are used. Most authorities accept that these basic considerations are reasonably well summarized by the “five freedoms,” originally formulated by the so-called “Brambell Report” (Command Paper 2836, 1965). This original report described the need to provide animals with the opportunity or freedom “to stand up, lie down, turn around, groom themselves and stretch their limbs.” The Five Freedoms were further refined and expanded by the Farm Animal Welfare Council in the early 1990s to take their most current form (FAWC, 2009):
Freedom from thirst, hunger and malnutrition—by ready access to fresh water and a diet to maintain full health and vigor.
Freedom from discomfort—by providing a suitable environment including shelter and a comfortable resting area.
Freedom from pain, injury and disease—by prevention and/or rapid diagnosis and treatment.
Freedom from fear and distress—by ensuring conditions that avoid mental suffering.
Freedom to express most normal behavior—by providing sufficient space, proper facilities and company of the animal’s own kind.
Of these “freedoms,” the fifth is the probably the hardest to define and the easiest to overlook or ignore. While it is obvious to most people that animals have physical requirements for adequate food, water, protection from the elements, and so on, it is much less widely acknowledged that animals also have social and behavioral needs (Dawkins, 1983; Hughes and Duncan, 1988), and that these needs differ markedly between species (Mason and Mendl, 1993). Understanding animals’ social and behavioral needs by primary caregivers is part of the ethical obligation attending animal ownership and use. It is of some concern, therefore, that few practitioners in the AAI field receive adequate ethological training on such matters. Judging the value of a particular behavior or social interaction to an animal may sometimes be difficult. However, in general, if an animal is strongly internally motivated to perform a particular behavior or social interaction, and if its motivation to perform appears to increase following a period of deprivation, it is an indication that the activity or interaction is probably important to the maintenance of that animal’s welfare. Common indications of deprivation include animals performing abnormally high frequencies of displacement or vacuum2 activities, repetitive or stereotypic behavior, apathy and prolonged inactivity, and/or self-mutilation (Broom and Johnson, 1993; Dawkins, 1988).
Conversely, and in contrast to free-living animals, most therapy and assistance animals are trapped in systems where they have little control over their social lives, and where they cannot avoid or escape unwelcome or unpleasant social intrusions. Denying animals control over their physical and social environment is also known to have adverse effects on their physical and mental well-being (Hubrecht, 1995). Animals need to have an opportunity to habituate to the environment and to the activities in which they are involved. This allows them to adapt to the situation and cope with potential sources of distress or discomfort. If a stressful situation overwhelms the animal, the animal’s welfare will be dramatically compromised. Fine and Eisen (2008) reveal a case study where they discussed a few occasions where therapy dogs were integral in the treatment of an extremely active child diagnosed with ADHD. Close attention needed to be given to assure that the animals were not overwhelmed or became anxious at times where the stimulation was high (loud noises or over-petting). They recommended that clinicians must be cognizant of what occurs within therapy and its impact on their therapy animal. If any signs are noticed, there needs to be an opportunity for respite or refuge for the animal.
For service and therapy animals, problems of basic welfare are most likely to arise in circumstances where animals are either residential within health care settings, or spend large amounts of time in holding facilities such as kennels or stables. In the former context, inadequate advance planning, selection, and staff commitment and oversight can lead to animals being improperly cared for (Hines and Fredrickson, 1998). Small mammals, birds and reptiles that are caged or confined are probably at greater risk of neglect or improper care, and non-domestic species that tend to have more specialized requirements than domestic ones are also likely to be at greater risk. “Improperly cared for” in these contexts should have the broadest definition. Most often it is defined as animals that are inadequately fed, watered or cleaned. However, any failure to attend to individual needs should be regarded as improper care. Overfeeding animals to the point of obesity is just as negligent as underfeeding. Giving an animal the opportunity to exercise or interact with conspecifics is not enough without insuring that the individual takes advantage of the opportunity.
23.2.2 Aging and retirement issues
Further welfare challenges arise when therapy and assistance animals begin to age. Many dogs, for instance, display clear evidence of progressive cognitive as well as physical impairment associated with aging, including disorientation, failure to recognize familiar individuals, restlessness, and house-soiling (Neilson et al., 2001). Naturally, under these circumstances, an animal’s schedule for therapeutic involvement will need to be curtailed. This may cause some disruption and adjustment for both the clinician and the animal, although careful planning may help to mitigate this. Each case should be handled individually to ensure that the needs of the animal are best met. In some cases, it may be ideal to stop using the elderly animal altogether. However, if the animal has grown accustomed to participating in therapy sessions for many years and finds a sudden and radical change in schedule or activities this too may be stressful. It is imperative for the therapist to appreciate the animal’s age and physical health. Adjustments need to be made, and the first priority should be the welfare of the animal. If in doubt, it may be a good idea to consult with an impartial third party, such as the animal’s veterinarian or an animal behaviorist.
Based on observations when his first therapy dog (Puppy) began aging and slowing down, one of the present authors noted that, despite her decline in energy, she remained eager to go to work. Every day, Puppy would be found next to the door when it was time to leave for work. Curtailing her work schedule had been a consideration for a few months, since she appeared tired and deserving of a break. However, when left at home, she seemed unhappy and would just lie next to the door. This was a dog that was used to being on the go. Altering her activities without any proper planning appeared to be devastating, especially because it modified her daily routine (Fine, personal observation). A transition plan was therefore formulated that allowed Puppy to have an adjusted schedule, with plenty of opportunities for respite and sleep. A day bed was set up for Puppy in one of the conference rooms. When she was tired, Puppy quickly learned that she could retreat to the room and nap for as long she needed. Whenever she rested, the clients were told that she was not to be disturbed. Some of the children just wanted to stick in their heads and see her. We eventually shortened her day at the office, which did not seem to bother her. She would just be picked up earlier in the day and went home and relaxed. Puppy lived until the ripe age of 14. She passed away on a weekend, actually working a few days before her death. By planning for retirement and making provisions for her, Puppy was able to preserve her sense of integrity. Attention was given to her health as well as her emotional well-being. Throughout her final years of therapy, Puppy received frequent health checks by a veterinarian to assess her health and to monitor her capabilities of continuing her therapy role. This process has now been followed with several of the writer’s older animals. The approach seems to be the most humane for all parties involved and is strongly recommended for older therapy animals.
23.2.3 The problem of stress
As originally defined by Hans Selye (1957), stress is the body’s natural physiological response to environmental stressors. The processes underlying this stress response are now reasonably well understood: when humans and other animals are subjected to unpleasant or painful stimuli, their bodies respond by secreting a group of hormones from the hypothalamus, and the pituitary and adrenal (HPA) glands. These hormones ordinarily serve to prepare the body for so-called “fight or flight” reactions, and, once the emergency is over, hormone secretion generally declines to a normal baseline level. However, under certain circumstances, particularly when the source of pain, anxiety or distress cannot be readily avoided or controlled, the stress response of the HPA system may become prolonged or “chronic,” thereby producing a number of deleterious consequences for the health and welfare of the individual. It is essentially impossible to know precisely what levels of stress an individual is experiencing at any given moment. Instead, we are reliant on various indirect measures or “indicators” including levels of stress hormones (cortisol, catecholamines, etc.) in body fluids—blood, saliva, urine, etc.—and outward manifestations of stress involving overt changes in behavior. Behavioral indicators of stress vary greatly between species, and in many they have never been studied or described in any detail. In the dog, studies and anecdotal observations suggest that sweating paws, salivating, panting, muscle tension, restlessness, body shaking, paw lifting, yawning, aggression, hypervigilance, and intensified startle reflex may all be behavioral manifestations of stress (Beerda et al., 1998, 1999; Butler, 2004), while in cats, alert inactivity, tense muscle tone, crouching posture, and pupil dilation may be indicative (McCobb et al., 2005). In horses, signs of stress can be characterized by vocalization, pawing, increased incidence of head movements, increased aggression, sweating, and increased respiratory and heart rates (Kay and Hall, 2009; Stull, 1997). Unfortunately, in other species of animals that can be used in therapy, such as reptiles or fish, the signs of stress may be harder to identify. In general terms, attempts by the animals to escape or increased propensity towards aggressive behaviors can be perceived as signs of discomfort, distress or stress. Therapists must pay close attention to the animals and be aware of stressful signs. Beyond that, detailed attention to the animals and their actions, supported by a precise knowledge of the species characteristics and behaviors, both normal and abnormal, is required to make an adequate assessment.
There are many potential sources of chronic stress in the lives of therapy and assistance animals. Trainers, practitioners and end-users of these animals should be educated to recognize the warning signs and act accordingly. It is ideal if stressful situations are anticipated and corrected before the animal shows signs of being affected. Planning for regular breaks in between therapy sessions in which animals are given an opportunity to rest and are given positive human contact, if appropriate, can alleviate the effects that potentially stressful situations may have on them. If a stressful situation presents, prompt action to ameliorate its impact on the animal’s welfare is required. This may result in the removal of the animal from the stressful environment, giving time for the animal to fully recover.
23.2.4 Use of non-domestic species
Most domestic animals have been selected to show a higher degree of tolerance of stressful situations and stimuli compared with non-domestic species, even those reared entirely in captivity (Hemmer, 1990). The implications of this are that, if a particular animal-assisted intervention or activity is stressful or potentially stressful for a domestic species, such as a dog, cat or horse, it is likely to be even more so for a non-domestic one. The majority of wild animals kept in captivity also have more specialized nutritional and husbandry needs than their domestic counterparts and, in many cases, their specific needs and requirements are not as well known, greatly increasing the likelihood that they will receive inadequate care unless practitioners, staff, and consulting veterinarians are properly versed in their particular species-specific requirements. Clinicians should become educated on proper handling, nutrition and care of these animals so that they are not inappropriately interacted with. These unsuitable actions could lead to behavioral and physical challenges. Non-domestic species are also harder to train, and their entrained responses extinguish more quickly in the absence of appropriate reinforcement. Some species, such as parrots and many non-human primates, are also highly intelligent and socially manipulative (Cheney and Seyfarth, 1990), and this tends to make them potentially unreliable or disruptive as social companions. All of these factors render non-domestic species less suitable for use in AAA/T and assistance animal programs, and more likely to experience welfare problems if used.
Among therapy animals, this point is well illustrated by the increasing use of parrots in residential settings, such as hospices, long-term care facilities, and correctional institutions. Recent advances in avian medicine, nutrition, and behavior reveal that most of these birds have highly specialized needs relating to air quality, nutrition, lighting, housing, sleep, and both environmental and social enrichment. Since avian wellness and welfare is difficult to maintain in institutional contexts due to a lack of centralized, informed, and consistent care, it should be questioned whether it is ethically appropriate to place birds in these settings at all (Anderson et al., 2005).
Similar concerns are raised by recent efforts to train and use capuchin monkeys (Cebus sp.) to assist profoundly disabled people. In most cases, these programs have found it necessary to neuter and surgically extract the canine teeth from the monkeys before they can be used safely with such vulnerable human partners. Monkeys may also be required to wear remotely controlled, electric shock-collars or harnesses in order to provide the user with a means of controlling the animal’s potentially aggressive and unreliable behavior. Clearly, the necessity of using of such extreme and invasive measures raises doubts about the practical value of such programs, as well as serious ethical questions concerning the welfare of the animals involved.
23.3 Animals used in therapy
Although the boundary may at times seem blurred, particularly with respect to so-called “emotional support animals” (ESAs),3 a reasonably clear distinction exists between therapy animals and service (or assistance) animals, at least in the USA. However, whereas service animals are relatively strictly defined under federal law (Americans with Disabilities Act, 1990), therapy animals form a heterogeneous category that can encompass everything from pet visitation to swim-with-dolphins programs (Iannuzzi and Rowan, 1991). The particular welfare issues confronting therapy animals are therefore correspondingly diverse.
23.3.1 Animal visitation programs
The majority of therapy animals are personal pets (usually dogs) that, together with their owners, provide supervised visitation programs to hospitals, nursing homes, special-population schools, and other treatment centers (Duncan, 2000). Such animals (and their human partners) are typically certified as being suitable for the task of visiting based on their responses to simplified temperament tests. However, while most such tests evaluate the animal’s reaction to acute stressors, and their willingness to tolerate intimate or invasive handling by strangers, rarely if ever do they attempt to ask the animal if he or she is actually motivated to interact socially with unfamiliar humans given the choice. As one author on the subject points out, the distinction is a crucial one:
Nothing else dogs do compares to the kinds of intrinsically stressful social interactions that take place when they visit clinical, educational, or post-trauma situations. No other canine-related event, no sport nor competition requires a dog to enter the intimate zones of unfamiliar humans and remain there for several minutes of petting and hugging…Most dogs have been bred for generations to distinguish between outsiders and the family, and to act accordingly. There has never been a breed of dog designed to enjoy encroachment from strangers. Dogs who actually enjoy interactions in clinical and educational settings are very rare, and the uniqueness of their talent should be appreciated.
(Butler, 2004, p. 31)
Even if this last sentence might be considered exaggerated by some, the fact remains that visiting animals rarely enter therapeutic settings of their own volition, and many of them are likely to find the experience of being constantly approached and handled by strangers—often strangers with abnormal behavior or demeanor—stressful and/or anxiety-provoking. If this is the case with dogs, it is reasonable to conclude that the other species occasionally used in visitation programs are even more likely to experience welfare problems.
Signs of stress-related fatigue are commonly reported in visiting therapy animals, leading some practitioners to conclude that such visits should be limited to one hour or less (Iannuzzi and Rowan, 1991). As one author notes, “[E]thical handlers develop time frames and environmental policies that allow their dogs to visit only within environments that are comfortable for them, and they leave before, not after, their dogs develop major symptoms of stress” (Butler, 2004, p. 37). Unfortunately, many handlers appear oblivious to the stress signals emitted by their animals, perhaps because they enjoy the social aspects of visitation more than their dogs do. In these cases, it is important that the handler or the therapist be responsible for the animal participating in a visitation therapy program and actively assess the animal’s status for changes in behavior. This will allow for the anticipation of situations that may cause stress to the animal. This problem-solving process should lead to the establishment of the necessary modifications to the visit or to the prompt removal of the animal from the stressful environment before clear signs of discomfort appear. A good knowledge of the individual animal’s behavior is required to pick on the subtle cues that the animal gives to show its discontent or tiredness. For example, bringing a therapy dog on a unit for youth with psychiatric disorders, the therapist must be aware of unusual behaviors that the dog may emit which may be demonstrative of various stressors. When this occurs, the animal should be removed from the situation and plans must be made to avoid further discomfort. Even if the handler has no precise knowledge of the dog, the presence of some general signs of stress, such as body stiffness, lowered tail, whining or increased panting, without any other obvious reason, should lead to an assessment of the animal and, even if in doubt, to its removal from the specific situation.
23.3.2 Residential programs
Although reliable evidence is lacking, the potential for stress-related fatigue or “burnout” in therapy animals is probably greatest among residential programs—prisons, nursing homes, inpatient psychiatric hospitals, long-term care facilities, etc.—where the animals are potentially “on duty” all day, every day of the year (Iannuzzi and Rowan, 1991). Animals housed in residential settings must therefore be provided with adequate “downtime” as well as access to comfortable and safe havens where they can escape entirely from the attentions of residents should they wish to. A minimum number of opportunities for resting need to be included in the daily schedule and planned to be of appropriate length and as often as required depending on the potential for stress associated with the specific animal use. Unfortunately, none of the associations representing therapy animals have prescribed a protocol to identify a realistic length of time that an animal should be in therapy sessions without a break. This must be considered, and prescribed breaks and respite should be implemented. However, due to individual differences in animals and various species, adjustments must be made on an individual basis. The authors urge the readers to take this into strong consideration to assure that ample time is given to animals to relax. It is better to do it this way rather than wait until the animal shows signs of being fatigued.
While the goals of therapy or rehabilitation may be enhanced by encouraging inmates and residents to participate in the care and training of such animals, it is essential that these activities are appropriately supervised, and that one of more staff persons is fully committed to accepting primary responsibility for insuring the welfare and well-being of the animals. The lack of a clear “chain of command” with respect to animal care responsibilities is the most frequently cited reason why animals’ needs are sometimes neglected in residential facilities. Regular, routine veterinary examinations should also be required both to monitor the animals’ health and to evaluate their stress status (Iannuzzi and Rowan, 1991).
The potential for serious animal abuse and cruelty is probably also greatest in residential programs, particularly those based in correctional or psychiatric facilities. Although thankfully rare, occasional cases of outright cruelty to therapy animals have been reported or alluded to in the literature, usually in situations in which the human/animal interactions have not been closely supervised (Doyle, 1975; Levinson, 1971; Mallon, 1994a,b). Once again, this speaks to the necessity of establishing clear and careful guidelines regarding staff oversight and control of all animal-based interventions.