Quantifying the behavioural and facial correlates of pain in laboratory macaques

In humans, pain reflects biological, psychological and social factors that can make clinical assessments challenging. Facial expressions of pain are consistent within and between individuals and detailed video analysis reveals differences between genuine, faked and masked pain responses. Identifying painful experience in other animals is even more challenging. However, recognising that an animal is in pain is necessary for providing appropriate pain relief. Observational methods for assessing pain are favoured because these do not require additional disturbance of an individual. There are already methods for evaluating the well being of captive animals but these can only detect a broadly positive or negative mood. Facial expressions may be a useful tool for identifying specific subjective states, such as fear and pain, as these occur rapidly in response to specfic stimuli. It is possible that other species do not display pain overtly as it could be harmful to reveal weakness in the presence of competitors. In social species, individuals can also benefit from social support and an expression of pain could be used to recruit support or alternatively to signal others to keep their distance. Research has shown that pain responses in rodents are modulated by the specific social relationship between social partners. It may be wise to conceal pain from competitors but not from friends.

There are currently no standardised methods for assessing pain in nonhuman primates. Recent work in rodents has indicated that facial expressions may be used to recognise the presence and intensity of pain experiences. The proposed study aims to use detailed analyses of facial movement to examine subjective states in laboratory housed macaques. Human observers are disproportionately attentive to faces when looking at images. Focusing on facial behaviour and identifying cues to pain would allow us to exploit this bias. We have developed a standardised tool for the measurement of facial expression in macaques (MaqFACS) and identified some facial correlates of pain within a neuroscience laboratory. However, the potential of this method in assessing pain in other contexts and to different types of pain has not been tested and it is important to determine whether facial measure can be applied more widely. As facial expressions are also social signals, we expect the presence of their allies or enemies, including animal care staff, to influence their ability to cope with challenge. Understanding how these factors influence behaviour will give us insight into how they evaluate their experiences and how we might best improve welfare.

The identification of behavioural indices of affective states in animals is central to assessing the benefits and costs of research and for efforts to refine methods and improve welfare. While affective signals do not directly indicate subjective experience, these potentially differentiate responses to negative stimuli more precisely (e.g. as fear or pain), than physiological or valence measures alone. The proposed study aims to use detailed analyses of facial movement to examine subjective states in laboratory housed macaques. Human observers are disproportionately attentive to facial cues when assessing animal welfare; to exploit this bias requires both the validation of facial cues in the assessment of subjective states and the development evaluation tools that can be applied in the laboratory context. In humans and rodents, micro-analyses of facial movements indicate consistency in the expression of pain across individuals. We have developed a standardised tool for the measurement of facial expression in macaques (MaqFACS) and identified facial correlates of pain within a neuroscience laboratory. However, it is important to validate this method across contexts, with different types of procedures and surrounding protocols. Given that facial expressions are also social signals, we expect the presence of conspecifics or animal care staff to impact upon both the form and frequency of expressions produced. Understanding this relationship is important for the application of behavioural methods in the assessment of pain.

The development of a novel methods for asssessing pain in laboratory macaques has significant value and has very clear applications in the assessment and improvement of welfare. Although macaques represent only a small percentage of the animals used in research under licence in the UK, the challenges of maintaining good welfare in captive primates is widely recognised. Rhesus macaques are widely used as a model species for human health research (e.g. National Institute of Health, USA). The development of MaqFACS (facial measurement tool) was funded by NIH in recognition of its potential to provide a standardised tool to detect subtle behavioural changes in laboratory macaques. It is especially relevant to evaluating the effects of neurolgical disorders, such as Parkison's Disease which causes deficits in facial movement. In laboratory rodents, recent research indicates that a facial coding approach is effective in the recognition of pain. For macaques, a standardised coding tool is already available and the research team has expertise in its application, saving considerable research effort in developing a new coding scheme.

The evaluation of pain in laboratory primates is required for the effective provision of pain relief. This project will also examine the influence of laboratory routines upon the behaviour of individuals expiencing particularly aversive events. As such, the findings have clear pratical applications in relation to the protocols currently used, such as duration of separation from social group, and costs and benefits of sensory access to conspecifics. Any scale has to be sensitive enough to distinguish between different intensities of pain, so that appropriate type and dosage of analgesics can be selected. Improvements to husbandry and procedures to minimise pain and suffering would represent a refinement in research. Standardised measures would provide insight into the relative severity of different procedures, either in isolation or across the lifespan of a research animal, underpinning judgements about the costs and benefits of research with animals. In addition, improvements in veterinary care has implications for health of laboratory animals and the identifcation of humane endpoints, both could potentially reduce number of animals used in research.

Importantly, the research will be carried out within applied contexts and under conditions that reflect the lifelong experiences of laboratory macaques; from breeding centre to research laboratory. Our video microanalysis approach is clearly not feasible as a practical and efficient measurement tool for wider use. However, we aim to translate our findings into a visual rating scale that would be both practical and sensitive enough to guide clinical decisions in an applied context. We will be able to implement the use of the visual scale at the two partner institutions and in collaboration with care and veterinary staff. This research will inform guidelines for the evalualuation and alleviation of post operative pain in macaques (e.g. www.digires.co.uk, www.ahwla.org.uk). Standardising the assessment of pain has the potential to improve pain treatment for macaques across a diverse range of research settings.