Keywords

1 Introduction

…refinement is never enough, and we should always seek further for reduction and if possible replacement. (Russell and Burch 1959, p. 66)

Using animals for research and testing in laboratories has, by its nature, the potential to cause “pain, suffering, distress or lasting harm.” This is precisely how regulated scientific procedures are defined (e.g., Home Office 1986), and as such, scientific research is strictly controlled through legislation in many (but not all) countries (Bayne et al. 2010). Intentionally conducting scientific procedures that have the potential to adversely affect the welfare of animals raises its own ethical issues, and the use of nonhuman primates (hereafter primates), as opposed to other animals, is also a special case. In this chapter, we focus on regulated laboratory studies (and not unlicensed behavioral or cognitive research on primates), describing the ethical framework of the 3Rs, the importance of promoting welfare given the link with quality of scientific output, and the major welfare issues affecting primates in laboratories. Our main emphasis lies with how we can improve the welfare of laboratory-housed primates through coordinated refinements across the lifespan, recognizing the critical role humans play in devising opportunities for reducing harms, and advancing primate welfare.

Animal welfare has been the focus of scientific study for many years yet constructing a single definition and approach to measurement has been difficult (reviewed in Fraser 2009). It is accepted that welfare is broad in concept, multidimensional in nature (Dawkins 2004), and lies on a continuum from poor to good (Broom 1999). In this chapter, we adopt an integrated approach to the concept and assessment of welfare that includes both physical and psychological aspects. Defined by Broom (1986, 2010), the welfare of an animal is its state as regard its attempts to cope with its environment, such that failure to cope leads to profound deviations in biological functioning. Thus, animal welfare, as a biological state within the animal, is relevant to scientists who use primates in biomedical research and testing to benefit humans in some way. When primate welfare is poor, primates are not “fit for purpose” as models of normal functioning.

2 Differences Between Laboratories and Other Captive Settings

Factors affecting the welfare of primates housed in laboratories differ in multiple ways from those in other captive settings (see Table 2). As described in this chapter, laboratory animals have regulated procedures conducted upon them to characterize a pathophysiologic process or intentionally model clinical disease that can result in pain, suffering, or distress similar to the target patient. Factors negatively affecting welfare of primates in zoos may be high visitor numbers (Hosey 2000, 2022), a stressor not present in laboratories. There are some rare cases where zoo-housed primates are released back into the wild, with concomitant stress (such as the golden lion tamarin, Leontopithecus rosalia, Teixeira et al. 2007). Nonetheless, compared to laboratory-housed primates, zoo-housed primates are likely to have better welfare. They are usually housed in more natural social groupings with conspecifics of both sexes and all (or nearly all) age classes. Their enclosures are also comparatively larger, more complex, and include access to outdoor runs providing more choice and control (Coleman et al. 2022). Primates kept as pets, however, have a host of welfare-related issues related to inappropriate rearing, housing, and husbandry (Hevesi 2022), as do those who live in sanctuaries (Brent 2007) given their life experiences.

3 Primate Use in Laboratories

The number of primates used in laboratories worldwide is not known exactly, but estimates over 16 years ago were in the region of one to two hundred thousand (Carlsson et al. 2004). Primates are used as models for humans because of their genetic, physiological, and psychological similarities, primarily in the fields of microbiology, immunology, neuroscience, biochemistry, pharmacology, and toxicology (see Hau et al. 2000; Carlsson et al. 2004; Weatherall et al. 2006; Chapman et al. 2010).

The most commonly used species of primate used for research and testing are rhesus macaques (Macaca mulatta) and cynomolgus macaques (M. fascicularis). A range of other Old World monkeys are used less frequently, such as pigtail macaques (M. nemestrina), baboons (Papio spp.), and vervet monkeys (Chlorocebus pygerythrus). Of the New World monkeys, common marmosets (Callithrix jacchus) are the most frequently used, but also others such as tamarins (Saguinus spp.), and squirrel (Saimiri spp.), capuchin (Sapajus spp.), and night (Aotus spp.) monkeys. Chimpanzees (Pan troglodytes) were the only great ape used in biomedical research, but European legislation (European Parliament and the Council of the European Union 2010 and Turner 2022), and other bans or National Institutes of Health funding limitations mean they are now no longer or rarely used (Kaiser 2013; Graham and Prescott 2015; Grimm 2015). However, Gabon continues to conduct biomedical research on chimpanzees (Kaiser 2013).

Primates are usually purpose-bred for use in research. The use of wild-caught animals is generally no longer accepted (McCann et al. 2007; European Commission 2010). This is because of the stress of capture and transport, and the associated morbidity and mortality (McCann et al. 2007), and presence of disease (Weber et al. 1999). In addition, they may be less suitable as models when their life histories are not known (Howard 2002) as this may introduce unwanted variation and bias experimental outcomes, resulting in studies lacking in experimental rigor and that are consequently less robust in prediction (Howard 2002).

Purpose-bred primates maintain their evolved capacities (i.e., adaptations to survive and reproduce in the wild), and so an understanding of natural history is critical if we are to provide environments that may help to promote welfare. The Jennings and Prescott (2009) report chapters in the Universities Federation for Animal Welfare (UFAW) Handbook (2010; 9th ed. in prep), and Marini et al. (2019) provide important information about the natural history, veterinary, and welfare aspects of the most commonly used primates in research.

Not all animals are protected by legislation, and this varies by country (Bayne et al. 2022). For example, in Europe all vertebrates and some invertebrates are protected, and certain animals, such as primates, cats, dogs, and Equidae, get special protection (European Commission 2010). It appears that this is due to public concern for animals that humans keep as pets or close companions, and our ability to empathize with these animals. There is, however, no robust scientific evidence that the animals with special protection are capable of suffering more than those without special protection (Buchanan-Smith 2010b; Hubrecht 2014).

What evidence is there to suggest that primates require special protection? Their phylogenetic closeness to humans is exactly the reason why they are used, and this similarity is also the basis for our apprehension concerning their use (i.e., they may suffer like humans). The brains of primates are larger in relation to body size than other mammals used in laboratories (Dunbar and Shultz 2007), and brain size is associated with mental capacities and cognitive complexity. However, cognitive complexity does not necessarily mean a greater potential to suffer pain (see Mendl and Paul 2004). The sensation of pain may be the same for an individual which experiences it as to one which is consciously aware and feels pain (Bekoff 2002). An animal therefore does not need to be self-aware to experience pain. Indeed, Broom (2010) has argued that cognitive complexity may reduce suffering as it helps individuals cope with adverse conditions and allows for more possibilities of pleasure. On the other hand, the most cognitively complex primates, the great apes, may also be able to empathize with the suffering of others, and dread future events, increasing their own ability to suffer (Smith and Boyd 2002; Mendl and Paul 2004). These arguments are not fully evaluated in relation to welfare, but suggest that primates, and certainly great apes, are indeed a special case.

In our view, the strongest scientific arguments that primates require special protection are (1) the intricacy of adverse effects resulting from inappropriate rearing (e.g., Parker and Maestripieri 2011) and (2) that their larger brains have evolved for dealing with the complexity of their social and physical worlds. Primates are long-lived compared with other laboratory animals (e.g., rodents), and this poses challenges for care staff who need to provide the opportunities for good welfare throughout their lives, including as their needs change. The provision of appropriate rearing, together with physical and social complexities in the laboratory environment, can be very challenging given the constraints of laboratory life and the requirements of studies. Table 2 outlines the key stages in a cynomolgus macaque life cycle, with potential negative welfare impacts and opportunities for refinement.

There is considerable debate about whether the suffering that primates experience in laboratory research is cumulative (see Honess and Wolfensohn 2010; Pickard et al. 2013; Wolfensohn et al. 2015). What is clear is that some individuals are unable to cope and are euthanized (Wolfensohn 2022). This may be due to additive stacking up when “the residual effects of repeated procedures may add up” or it may be due to additive potentiation when “suffering from earlier events may actually increase the negative impact on welfare of subsequent events” (Pickard et al. 2013, p. 6). In addition to suffering from direct scientific procedures (e.g., surgery, disease modeling, adverse effects from a test item), the intelligence of primates means they may suffer from boredom and fear. Therefore, the consequences of this, inadequate rearing histories, and environments, together with the scientific procedures conducted upon them, can lead to poor welfare (Buchanan-Smith 2010a, b) (Table 1).

Table 1 Examples of issues that may compromise welfare of primates (and many other animals) housed and used in laboratory research and testing, and associated refinements
Full size table

4 Ethical Framework of 3Rs and Welfare

The utilitarian approach is adopted for dealing with the ethical dilemma of using animals in research and testing, is enshrined in legislation, and underpins many local ethical review processes (e.g., European Commission 2010; U.S. Department of Agriculture 2013; including China and India; see Graham and Prescott 2015). This pragmatic approach weighs the ethical importance of the individual and their capacity to suffer against the interests of the other parties concerned (Singer 1975; Sandøe et al. 1997). In practice, this approach is known as a harm-benefit assessment. It is currently applied to primate use prospectively by mandatory ethics boards (e.g., Institutional Animal Care and Use Committees, Animal Welfare and Ethical Review Body) and retrospectively where the actual costs to the animal are reviewed in light of the results of scientific study (European Commission 2010). The perceived harm to the animal in terms of its likely experience of pain, distress, or lasting harm, including intensity, duration, and frequency, is weighed against the anticipated benefits of the research for humans (or other animals or the environment) (Graham and Prescott 2015).

In addition to requiring assessment of the harms and benefits of the proposed research, the legislative framework requires the 3R principle be applied to the project from its experimental design to its execution (Home Office 1986; European Commission 2010). The 3Rs are replacement, reduction, and refinement. Replacement is concerned with the absolute or relative replacement of animals for scientific use. Reduction emphasizes the need to reduce to a minimum the number of animals through good experimental design, the sharing of data and/or resources, or by using modern techniques to obtain more information from the same number of animals (thereby reducing future use of animals). Refinement has been defined as “any approach which avoids or minimises the actual or potential pain, distress and other adverse effects experienced at any time during the life of the animals involved, and which enhances their well-being” (Buchanan-Smith et al. 2005, pp. 379–380). This definition highlights the need to consider all stages of an animal’s life, from birth to death, including the promotion of good welfare of breeding animals not used in research, and involves not just minimization of harms, but takes a proactive stance to enhance welfare through to death with the use of humane end points as required (Wolfensohn 2022). It should also be noted that, although each R is often considered separately, they have a complex interplay (de Boo et al. 2005). For example, reuse may reduce the number of individuals used, but increase the suffering of individual animals. Table 2 highlights some of the main welfare issues that a macaque used in toxicology may experience across their lifespan and describes the opportunities for refinement.

Table 2 Example life cycle of a cynomolgus macaque used in toxicology in the UK, with opportunities for refinement at all stages. PRT: positive reinforcement training
Full size table

The 3Rs provide a platform for uniting welfare together with scientific merit (e.g., refinement: Richter et al. 2010; Tasker 2012; Hall et al. 2015). They can also help increase public support for animal research by highlighting that alternatives are being sought and that animal welfare is prioritized (Leaman et al. 2014). However, despite the widespread scientific support of the 3Rs, there are barriers to uptake of refinements, including staff time, motivation, knowledge, skills, and resources. Laboratories need to have ongoing programs to critically appraise practice in light of new evidence and resources so that the most up-to-date refinements are used (Lloyd et al. 2008). Several publications provide detailed and comprehensive accounts of refinements for primates (e.g., Rennie and Buchanan-Smith 2006a, b; Rensnie and Buchanan-Smith 2006; Jennings and Prescott 2009). To implement refinement successfully requires understanding what welfare is, how it can be assessed, and having a strategy for rapid implementation of changes and their evaluation. Underlying this process should be an acceptance that refinement is a necessary and continuous process—it is a permanent challenge for care staff and scientists (Tasker 2012).

While welfare in the laboratory is formally considered in terms of refinement, one of the greatest influences on the development of animals and their resilience (i.e., their coping ability) as adults is their early rearing environment (Parker and Maestripieri 2011). Hence, the welfare of primates under study may be profoundly affected by the conditions in which they are born, reared, and kept prior to their use in a study.

In macaques, natural weaning from the mother’s milk is usually seen at 10–14 months of age (Harvey et al. 1987); it is a gradual process involving withdrawal of milk and dependence on the mother for caregiving over a period of weeks or months (Lee 1996). Offspring remain with their mother beyond weaning for up to 24 months of age (Ross 1992). In captive breeding, infants are commonly removed from their mothers and natal group at about 6 months of age (Honess et al. 2010). Removal from the natal group and manipulations in the early rearing environment are stressful and result in long-term alterations in the animals’ immune system and its regulation (Coe et al. 1989). More specifically, weaning and removal of the mother are known to have immunosuppressive effects (Coe et al. 1987). Toxicologists testing new pharmaceuticals that are likely to alter immunological parameters in macaques should be aware of the potential confounding effects of differential rearing histories when selecting research subjects (Tasker 2012).

There are also potential problems with the rearing environments of common marmosets, the most frequently used New World primate. This species is characterized by twin births and the cooperation of all members of the family in rearing the young until independence and natural weaning, which occurs at approximately 8 weeks of age (Yamamoto 1993; Buchanan-Smith 2010a). In captivity, dams have higher weights than in the wild, which is associated with larger litters, birth complications, and increased infant mortality (Ash and Buchanan-Smith 2014). Supplemental feeding of litters of three or more, involving removal of the infant or infants from the natal group for hand-feeding, is often practiced during the first 2 months after birth to reduce infant mortality. Depending on how this is done, it has the potential to affect development and confound scientific output (Ash and Buchanan-Smith 2016).

Ash and Buchanan-Smith (2016) used a battery of tests to determine the impact of rearing environment in common marmoset infants in 3 conditions: family-reared marmoset twin pairs, family-reared marmosets from triplet litters where only 2 remain (2 stays), and supplementary-fed triplets. The supplementary-fed triplet infants were never isolated except for very short periods for weighing and had positive experiences with humans from an early age. Furthermore, they are naturally adapted to being passed between carriers (Ingram 1977). The infants were also returned to their family group as soon as possible after feeding, and so spent most of their time with their family group. This supplementary feeding rearing practice had no adverse effects on behavior/cognition, neophobia, nor affective state (Ash and Buchanan-Smith 2016). However, primate infants that are hand-raised entirely by humans have reduced reproductive success and often experience adverse welfare such as increased self-directed behaviors, abnormal behaviors, and inappropriate aggressive behavior (e.g., Porton and Niebruegge 2006). Dettling et al. (2002, 2007) found that early parental deprivation in common marmosets impacts endocrine responses (lower basal cortisol) and several behavioral responses (e.g., they are less mobile and make fewer contact calls than controls in response to social separation/exposure to novelty), as well as blood pressure, which is higher than in controls. These changes make them unfit models of normal healthy humans.

Ideally, the purpose for which primates are bred and subsequently used as research subjects should be known, so practices can be put in place to ensure the animals are “fit for purpose.” This might, for example, involve human socialization and positive reinforcement training for certain husbandry practices and procedures, or exposure to a range of stimuli likely to be encountered, paired with rewards to desensitize the primates. However, many primates are bred in special centers that are often overseas and require transport to the laboratory (Prescott 2001) where they will be used in research (Ha and Sussman 2022). The laboratory of end use may not have direct control over, or the ability to monitor, social groups, weaning age, and conditions, although some countries have legislation to cover designated breeding centers. Ideally, research laboratories should have a coordinated approach with the breeding centers that supply their animals to promote welfare, as well as ensure the primates are “fit for purpose” and to minimize confounding factors that may introduce unwanted variability in scientific output.

5 Importance of Welfare for Quality of Scientific Output

To achieve high-quality and reliable science, several essential conditions must be satisfied. The experiment must yield unambiguous results by minimizing unwanted variation, there must be an absence of confounding factors (Poole 1997), and the study must be undertaken to a required standard (e.g., Organisation for Economic Cooperation and Development (OECD) 1997). In addition to aspects of quality pertaining to the study, quality in its fullest sense should also include the impact of research using animals (see Bateson et al. 2011)—that is, the application of knowledge resulting from research findings.

Poole’s (1997) seminal paper “Happy animals make good science” argues that good laboratory animal science is based upon normal, healthy, and happy animals, unless illness or alleviation of stress is the subject of study. Although the effects of stress and disease are easy to identify, and their confounding effects are well known (e.g., Reinhardt et al. 1995; Festing and Altman 2002; Hall 2007), Poole (1997) argues that unhappiness is also a confounding variable because its effects on biological variables produce increased variation in the data output. He goes further, asserting that most scientists working with animals assume that they have normal physiology and behavior (e.g., heart rates, blood pressure, blood values, metabolism, hormones, and immunological competence). However, these parameters can be dramatically altered by the conditions in which the animals were bred, reared, kept, transported, and the way in which experimental procedures were conducted. Experimenters may assume these parameters to be normal because they commonly encounter them and have no reference for comparison (Tasker 2012). If animals are not well acclimatized, properly characterized, and stable, there is a major risk of confounding and under- or overestimating the treatment effect with no predictive validity to the clinic (Graham and Schuurman 2017). For example, restraint may lead to maximum heart rates, preventing the detection of arrhythmias (Tasker 2012) or significant changes in glycemic control, blunting the response to treatment (Graham and Schuurman 2017). The link between welfare and scientific output is covered in Schapiro and Hau (2022).

There are strong ethical, scientific, and economic arguments to suggest that “no data are better than poor data.” If the data are not of good scientific quality and results are therefore potentially unreliable, inaccurate, or inconclusive, then the primates used in the research will have suffered in vain, violating utilitarian principles underlying animal use. Poor animal welfare and quality of science may also cause delays or lead the research down the wrong path, with more animals being used (going against the reduction principle) and more unnecessary suffering. This not only has ethical implications for the animals, but also wastes time and money. Indeed, Bains (2004) estimated that it takes an average of 12.5 years and $1 billion to take a new drug to market. Recognizing the dependence of reliability of scientific outcomes on animal welfare, it is logical to conclude that these costs are likely to be reduced with improved animal welfare.

6 Welfare Assessment

Given that scientific procedures directly impact negatively on welfare, it is critical for there to be ways of accurately assessing welfare. The list of factors in Table 1 illustrates that primates in laboratories often have reduced welfare, especially in the absence of refinement. However, assessment should include measuring positive welfare states, such as comfort and contentment, as well as negatives ones such as boredom, fear, pain, and/or suffering (Buchanan-Smith et al. 2005). We should focus not only on welfare as a snapshot of the animal in time, but view the animal, as much as is possible, from birth to death, 24/7, across the lifespan (Brando and Buchanan-Smith 2018). This includes day and nighttime assessments, and assessments over weekdays and weekends. For example, Lambeth et al. (1997) found higher wounding rates for chimpanzee during weekdays when care staff are present than at weekends, suggesting something about the weekday routines, such as elevated activities of caregiving, veterinary, research, and other personnel, were causing tension. As well as this welfare may vary across seasons, particularly if there is a mating season when male aggression rises. There are also individual requirements across the different life stages. Younger individuals require special provisioning to allow them to engage more in play, to explore, and to learn contingencies between behaviors and their outcomes. It is known that having control over aspects of the environment improves the welfare of younger individuals more than it does for older individuals (Badihi 2006). Waitt et al. (2010) provide a list of considerations for designing environments for aged primates that include accessibility issues, positioning, size, and type of furnishings, to avoid poor welfare related to age-related arthritis, deteriorating vision, difficulties in thermoregulation, etc. Furthermore, given individual differences in the propensity for welfare states due to personality differences (e.g., King and Landau 2003), we must consider individuals as being unique (Robinson and Weiss 2022).

Hawkins et al. (2011) provide an excellent review of assessment of welfare in laboratory-housed animals. Several practical issues were raised in this review, including how to set up and operate effective protocols for the welfare assessment. The need to tailor welfare assessment protocols to individual animals, as well as individual projects, is emphasized, too, in this review, together with the need to quantify objectively measures relating to the welfare state, and to intervene early to alleviate negative states and minimize them worsening. The problem is that even in our closest living relatives, the primates, it can be challenging to recognize internal states such as pain and suffering (e.g., Flecknell et al. 2011; Sneddon et al. 2014), and although the use of analgesics following potentially painful procedures is improving in primates, it is still not optimal (Coulter et al. 2009). Section 2 of this book provides a comprehensive review of the methods used to assess primate welfare.

7 The Role of Human Behavior Change in Refinement

From birth to death, the lives of primates in designated breeding and supplying establishments are under absolute human control. When seeking welfare improvements, a fully integrated approach is required to ensure refinements are implemented at every stage of the life cycle (e.g., Table 1). The stakeholders include the scientists, study directors, advising statisticians, ethical review panel staff, the veterinarians, the animal technicians, and care staff who are all responsible for the primates’ day-to-day needs—they all have a stake in implementing positive welfare change.

It is often the case that primates spend a rather small amount of time directly engaged in scientific research. An exception is neuroscience, where the primates (usually macaques) may be food and fluid-controlled, restrained in chairs, and tested from 2 to 8 h at a time, 5–7 days a week depending on the requirements of the experiment, and use continues for a number of years (Prescott et al. 2005). But for most primates, the majority of time is spent living in enclosures. Given this, the social and physical complexity of the enclosures, and the control and choices that the primates can make directly impact their welfare. In all cases, the behavior of humans is critical for promoting the welfare of primates housed in laboratories (Rennie and Buchanan-Smith 2006a).

Human behavior change is a growing discipline. It refers to a process that translates knowledge into actions, so that targeted change is implemented. This process is underpinned by multidisciplinary scientific approaches and theoretical frameworks (Michie et al. 2014); it has considerable merit in improving human health (e.g., Ory 2002) and is gaining momentum as a practical way for advancing animal welfare (e.g., Van Dyne and Pierce 2004; van Dijk et al. 2013; Whay and Main 2015). However, traditional approaches to implementing refinements have not focused on stimulating changes in human behavior.

Broadly speaking, two types of intervention are employed to improve animal welfare. These are (1) enforcement of legislation, codes of practice, and supplementary, voluntary accreditation schemes or standards, and/or (2) encouragement, which includes promoting innovation that exceeds minimum standards, and regularly accessing and implementing new knowledge and scientific findings.

In the UK, the appointment of Named Animal Care and Welfare Officers and, in Europe, the appointment of the Institutional Care and Animal Welfare Responsible Person provide oversight and, together with mandatory staff training, ensure minimum standards are met. Pharmaceutical companies and Contract Research Organizations (CROs) are committed to improving animal welfare. They often undergo voluntary Assessment and Accreditation of Laboratory Animal Care accreditation to demonstrate that they meet the minimum standards required by law and are also going the extra step to improve animal care and use. Most large pharmaceutical companies and CROs are signatories of the Concordat on Openness in Animal Research, an agreement across the biomedical sciences sector to improve communication and increase transparency to the public about animal research. There is therefore considerable enthusiasm from the industry to improve animal care and use now and in future.

Good primate welfare is dependent upon creating a strong culture of care. To achieve this, we must overcome barriers, including a lack of knowledge/resources/skills, provide a robust scientific evidence base for recommendations, give ownership of improvements, and the recognition, support, and reward system for those who effectively engage with the refinements we are proposing. To tackle the lack of knowledge and skills, we see communication, training, and dissemination of findings as fundamental to moving the refinement agenda forward.

Given the range of standards worldwide, it is also critical that training is pitched at the right level for dissemination, and the effectiveness of training resources may be enhanced if created with an understanding of models of human behavior change. Considering that primate use is likely to continue until alternative technologies are developed (e.g., see Burm et al. 2014), we need to disseminate evidence-based practice and empower individuals to lead on refinement. The launches of websites, such as marmosetcare.com and the macaque website (nc3rs.org.uk/macaques/), together with other online resources (see Prescott 2016), are a good step toward this goal.

The second approach, encouragement, is where we see considerable opportunities for innovation and improvement, to promote sustainable human behavior changes that result in positive impacts on primate welfare. The cornerstone of encouraging behavior change is to transfer ownership of both the problem and solution to a person responsible for implementing change (Whay and Main 2015). There are two important components necessary for change to happen. The first is appreciating the relevance of the desired behavior change, which must be coupled with taking ownership of the process of change, rather than being told what to do, or even through demonstration. Therefore, creating opportunities for colleagues to explore issues and come up with their own solutions will be more effective than simply being instructed without individual motivation and responsibility (Cunningham et al. 2002). Indeed, there is no one-size-fits-all approach to refinement.

All key stakeholders need to be empowered toward improving welfare, while at the same time fully appreciating experimental aims and impact. The ultimate goal is to synergize better welfare and better science to elevate the quality of the research. Understanding and seeking out people who have powers of influence are important, as change requires targeting several levels. Behavior change may be encouraged using three broad approaches, namely social marketing (an extension of principles used in marketing and advertising to promote change among groups), participatory methods (such as those used in the community development sector), and the creation of action groups (self-help or discussion groups). Useful examples of how these approaches have been implemented and tested in the agricultural sector are discussed by Whay and Main (2015). In practice, both combining and coordinating approaches (e.g., Van Dyne and Pierce 2004; Pritchard et al. 2012; van Dijk et al. 2013) are required to improve welfare. We are keen to advance the evidence base showing that human behavior change techniques improve sustainable uptake of refinements for primates; hence, making primates better models for research.

8 Conclusions

There are specific welfare issues for primates used in laboratory research and testing, including painful procedures and a restricted environment. It is imperative that we apply refinement throughout laboratory primates’ lives. Specifically, efforts must be made to integrate all stages of a primate’s life to improve their welfare by providing opportunities for positive experiences and conditions that enable them cope with challenges. If primates are to be used as models of normal functioning humans, the promotion of welfare will also help ensure that they are “fit for purpose” and will avoid situations where a negative welfare state confounds biological data and leads to research with unreliable or faulty conclusions. By giving ownership of the resources to target audiences, and providing the evidence base underpinning benefits (welfare, scientific output, and financial), knowledge exchange within and across facilities should continue to improve, and with it animal welfare and quality of scientific output.