Report of an RSPCA/AHVLA meeting on the welfare of agricultural animals in research: Cattle, goats, pigs and sheep

Full text

43
Report of an RSPCA/AHVLA meeting on the
welfare of agricultural animals in research:
cattle, goats, pigs and sheep
*PENNY HAWKINS1, R. EDDIE CLUTTON2, NGAIRE DENNISON3,
MIRJAM GUESGEN4, MATT LEACH5, FINULA SHARPE6, HUGH SIMMONS7,
ADRIAN J. SMITH8, JOHN WEBSTER9and UTE WEYER7
1Research Animals Department, RSPCA, Wilberforce Way, Southwater, West Sussex
RH13 9RS
2Royal (Dick) School of Veterinary Studies, University of Edinburgh, Easter Bush Campus,
Midlothian EH25 9RG
3Home Office Animals in Science Regulation Unit, PO Box 6779, Dundee DD1 9WW
4Institute of Veterinary, Animal and Biomedical Sciences, Massey University, Private Bag 11
222, Palmerston North 4442, New Zealand
5Pain & Animal Welfare Science (PAWS) Group, School of Agriculture, Food & Rural
Development, Newcastle University, Newcastle Upon Tyne NE1 7RU
6HLS, Woolley Road, Alconbury, Huntingdon, Cambridgeshire PE28 4HS
7AHVLA, Animal Sciences Unit, AHVLA Weybridge, Addlestone, Surrey KT15 3NB
8Norecopa, c/o Norwegian Veterinary Institute, P.O. Box 750 Sentrum, 0106 Oslo, Norway
9Old Sock Cottage, Mudford Sock, Yeovil, Somerset BA22 8EA
*Corresponding author: penny.hawkins@rspca.org.uk
Introduction
The Royal Society for the Prevention of Cruelty to
Animals (RSPCA) and the Animal Health and Veterinary
Laboratories Agency (AHVLA) jointly convened a
meeting to bring together animal technologists,
researchers, veterinarians and students with an
interest in the welfare of cattle, goats, pig and sheep
used in research and testing, for a programme of talks
and discussion sessions. The meeting, held in June
2013, addressed a range of topics including cognition
and emotion in agricultural animals, welfare in
containment systems, positive reinforcement training,
pain assessment and alleviation and assessing actual
severity. It was initiated as a follow up to an
international meeting on harmonisation of the care and
use of agricultural animals in research held by
Norecopa, the Norwegian consensus-platform for
alternatives, in September 2012.
Norecopa: a national consensus-
platform for the 3Rs
Adrian J Smith, Norecopa
In 1999, following an initiative at the 3rd World
Congress on Alternatives and Animal Use in the Life
Sciences in Bologna the same year, a European
organisation was established with the purpose of
bringing together all four stakeholders interested in
animal research: regulators, academia, industry and
animal welfare organisations. This organisation, known
as ecopa (European Consensus-Platform for
Alternatives, www.ecopa.eu), offers membership and
support to national platforms which follow the same
principles. Currently there are 14 national platforms in
Europe that are approved by ecopa.
Norway’s national consensus-platform, Norecopa
(www.norecopa.no), was established in 2007.
Institutions and individuals can become members of
Norecopa and most Norwegian research institutions
have done so. Norecopa signals its commitment to all
3Rs of Russell & Burch by describing itself as a
‘national platform for replacement, reduction and
refinement of animal experiments’.
Inspired by the 2004 FELASA meeting in Nantes
entitled ‘Internationalisation and Harmonisation of
Laboratory Animal Care and Use Issues’, Norecopa has
arranged a series of international consensus meetings
on the care and use of animals in research. These
meetings have so far covered fish (in 2005 & 2009),
wildlife (in 2008) and agricultural animals (in 2012).
April 2014 Animal Technology and Welfare

Tech-2-Tech
Norecopa wishes these meetings to be a lasting
resource for the international research community, so
each meeting has its own website
(http://www.norecopa.no/sider/tekst.asp?side=21)
where all the presentations are available, together with
links to guidelines on the care and use of the species
in question. In addition, participants write a consensus
document describing the challenges to implementing
the 3Rs which were identified during the meeting.
These consensus documents have been used actively
by Norecopa and others to initiate tasks designed to
address these challenges. Norecopa has also
established email-based discussion forums following
these meetings but interest in these has so far been
low.
The consensus document resulting from the 2012
meeting on agricultural animals included
comprehensive recommendations and a list of tasks
relating to the implementation of the 3Rs when using
agricultural animals in research and testing. These
included further meetings and discussion fora on 3Rs
topics for a range of stakeholders. There was
considerable support for this among UK researchers,
animal technologists and veterinarians, so delegates
from the RSPCA and AHVLA went on to convene a
meeting based in the UK in June 2013.
Norecopa welcomed the RSPCA/AHVLA meeting as an
initiative which should help to advance effor ts to
increase implementation of the 3Rs in the care and use
of agricultural animals.
Agricultural animals in research:
cognition, emotions and ethics
John Webster, University of Bristol
Agricultural animals used in research are considered to
be sentient. Definitions of sentience vary, as does the
nature of sentience itself within the animal kingdom.
However within the homeothermic classes of mammals
and birds, I stand by my definition of sentience as
‘feelings that matter’.1Sensations resulting from
environmental stimuli motivate the sentient animal to
actions that are designed to avoid suffering and
promote a satisfactory emotional state. If the actions
are successful, the animal learns to cope. If they fail
then the animal is likely to suffer. Failure to cope may
occur either when the stresses are too severe, complex
or prolonged, or when the animal is constrained in such
a way that they are unable to take constructive action
necessary to relieve the stress.
The primary motivation for sentient animals, including
humans, is emotion; the need to feel good, avoid
suffering and ignore the things that do not matter (see
Table 1 for some examples of animal emotions and
their positive and negative aspects). Cognitive ability, or
Positive aspect
Emotion
Negative aspect
Security
Fear
Anxiety
Pleasure
Comfort
Pain
Satiety
Appetite
Hunger
Joy
Hope
Despair
Table 1. Aspects of different animal emotions
the reasoned interpretation of sensation and
experience, can help to improve welfare by teaching the
animal to cope. However, it may in some circumstances
increase the degree of suffering when the coping
mechanisms fail. Nevertheless, suffering is an
emotional state; it is a fallacy to assume that the
greater the (apparent) cognitive ability of a species (still
less its phenotypic similarity to humans), the greater its
capacity to suffer. There is no good argument to
suggest that the capacity of a chimpanzee to suffer is
any greater than that of a dog or pig.2
I also contend that sentient animals do not simply live
in the present. Primitive sensations like hunger or pain
lead a domestic animal to hope that a meal will be
provided on time or fear that they may receive a
beating. An animal with the capacity for hope will also
have the capacity for despair.3
How then should we apply these concepts of sentience
to our approach to the care of agricultural animals,
especially those used in research, where we are
obliged to weigh any harms to the animals against
benefits to the society of humans (and other animals)?
I suggest the following rules of engagement.
1. Harms should be assessed not only in terms of their
short-term effects on the physical state of the animal
but also in terms of their longer-term effects on
emotional state (e.g. anxiety, learned helplessness).
2. The day-to-day husbandry of the animal must not
only be sufficient to meet physical needs for
nutrition, comfort, etc. but should also, wherever
possible, provide the resources necessar y to allow
the animal to achieve a satisfactory emotional state
through their own actions.
3. Species selection must be made on a case-by-case
basis where the benefits are assessed by weighing
the scientific evidence relating to the predictivity of
the animal ‘model’, against the harm that may
accrue to the animals both from the scientific
procedures and from their lifetime experience within
the research environment.
4. The capacity of an animal to experience suffering
must be defined in terms of their own sentience, not
their status in human society. It is unethical to
promote agricultural animals (e.g. minipigs) as
‘more acceptable’ than dogs or monkeys simply on
44

Tech-2-Tech
the basis of public identification of the pig as a food
animal, the dog as a companion animal and
primates as those species closest to man.
Welfare challenges in high disease
containment research
Hugh Simmons, AHVLA and NADIR (Network
of Animal Disease Infectiology Research
Facilities)
Disease-causing infectious agents must be contained if
they are (i) exotic and contagious to animals, (ii)
infectious to humans, (iii) genetically modified – or a
combination of two or all three of these factors. These
three different containment requirements are covered
by three different pieces of legislation in the UK, which
specify four different levels of containment in each
piece of legislation. There is some variation in the
detail of the containment required at each level. The
highest are Categories 3 and 4; Category 3 applies to
pathogens such as Classical Swine Fever, Blue Tongue
Virus and Tuberculosis (TB), while Category 4 applies
to Foot and Mouth Disease, Rabies and High
Pathogenicity Avian Influenza.
These disease containment legislations define key
elements of building design and management
necessary to ensure containment, such as air handling
units with high efficiency par ticulate air (HEPA) filters,
effluent treatment plant (ETP) on the drainage system,
sterilisation or incineration of waste and changing
facilities where staff change into the required personal
protective clothing and equipment (PPE). These factors
can impose restrictions on the kind of housing and
environment that can be provided for the animals, for
example long stem fibre roughage and bedding may
block ETPs, or PPE may affect interactions between
animals and staff. Once such highly specialised
facilities have been built, it can be difficult or
impossible to update them, to take account of some
new developments in housing and care, due to their
construction and complexity.
Considering all of this, undertaking research using farm
animals into diseases that require high containment
can present challenges that must be overcome to
ensure animal welfare.
Currently all the European organisations with these
types of facilities for farmed animals, including AHVLA,
are members of NADIR. This is a FP7 project funded by
the European Union, with the aim of sharing good
practice in a number of areas including animal welfare
(http://www.nadir-project.eu). Some of the approaches
used by NADIR and its members to improve animal
welfare are set out below, using the Five Freedoms as
an analysis tool (Box 1).
Freedom from hunger and thirst
Fitted troughs, drinkers and water bowls are often
found in Category 3 buildings just as in low security
facilities. However, Category 4 buildings are more
challenging due to the greater stringency on how waste
can be removed from the room or building. In a lot of
designs virtually all the waste has to be washed down
the drain to the ETP. This means a complete diet has to
be offered without long stem fibre for ruminants, which
reduces normal rumination function. As there is no
suitable alternative source of long stem fibre, studies
are planned so as to minimise the duration of category
4 housing and animals are fed a diet with high content
of fibre (but not long stem).
Freedom from discomfor t
The risk of environmental discomfort should be low, as
building management systems within containment
facilities keep temperature and relative humidity within
appropriate limits. With respect to physical comfort,
animals need somewhere comfortable to rest, which
can present problems if it is not possible to provide
conventional bedding materials. If animals are held on
non-slip flooring without bedding, they may be
uncomfortable and can develop sores or bursae at
pressure points during longer experiments. One way in
which animal comfort is assessed is to evaluate how
long they spend lying on different materials and then
provide the preferred material unless there are
significant veterinary or scientific constraints.
If some discomfort is unavoidable, the aim is to
address this within the experimental design so that
non-infectious phases are conducted outside, keeping
the time in containment to a minimum.
Freedom from pain, injur y and disease
The nature of the infectious agent used and design of
the experiment, will also both be critical to the animal’s
freedom from pain, injury or disease. In common with
animals kept in low security accommodation at AHVLA,
a Health Plan is implemented for each animal to
eliminate any health problems that are not related to
the study. It is particularly important that animals are
in good health before going into high containment; not
only to minimise any effects on welfare of the
experimental procedures and any husbandry
Box 1. The Five Freedoms
1Freedom from hunger and thirst.
2Freedom from discomfort.
3Freedom from pain, injury and disease.
4Freedom to behave normally.
5Freedom from fear and distress.
45

Tech-2-Tech
constraints but also due to the logistical problems of
treating any intercurrent disease in high containment
facilities. This would have a potentially catastrophic
impact on the validity of the study, especially as high
containment experiments are generally done on small
numbers of animals.
It is obviously impossible for animals to be ‘free’ from
pain or disease when they are used in disease
research, so refinement is critical. The impact of
disease can be reduced by implementing an effective
health and welfare monitoring protocol that includes
humane endpoints. Staff contact is unfortunately
limited to defined periods when animals are in higher
categories of containment but CCTV can be used for
continuous observations. Although this cannot
substitute for interactions with an empathetic and
experienced observer, or extra monitoring for animals
on critical points of severe protocols, CCTV can help to
maintain surveillance and can also provide footage that
can be sampled and reviewed.
The ‘welfare score charts’ used at AHVLA are always
tailored to individual projects and approved by the
institutional ethics committee. Humane endpoints are
refined using predictive indicators wherever possible.
For example, clinical signs and disease progression
vary between strains of classical swine fever but real
time analysis of white blood cell (WBC) and platelet
counts have enabled predictive endpoints to be
established. Animals are now humanely killed when the
WBC and platelet counts fall, so that they do not have
to experience the full progression of the disease.
Other measures to reduce suffering and refine
procedures at AHVLA include:
G‘pre-start’ meetings for researchers, veterinarians,
animal technologists and care staff to ensure that
all are aware of critical indicators and decision
points
Gtelephone contact list for out of hours, if euthanasia
is required
Gprocedures timed so that the phase with the
greatest potential to reach end point will occur early
or mid week
Gsamples taken early in the morning so that results
will be available that day, enabling better decision
making with respect to endpoints
Gexperienced staff are responsible for animal
monitoring and care for groups at greatest risk, e.g.
unvaccinated controls
Guse of microchips to monitor body temperature,
which reduces handling
Freedom to behave normally
The legal minimum space allowances for animals in
high containment are no different from those for
animals in conventional housing. However, containment
systems can have an impact on the kind of
environment that can be offered to the animals and
consequently on their freedom to behave normally.
Environmental enrichment is possible in containment
building designs and should be tailored to the species
being housed, e.g. long fibre foodstuffs and social
groups are very important for ruminants; straw and
other materials or objects to manipulate are good for
groups of pigs, reducing bullying and injuries within the
group.
Freedom from fear and distress
A key concept here is facilitating emotional resilience,
i.e. providing a set of conditions that allow emotional
adaptation to different forms of adversity at different
times in the animal’s life. Ways of fostering emotional
resilience include allowing acclimatisation periods to
the housing system before animals are used and – very
importantly – training animals before the experiment,
including giving food rewards. Care is also taken to
select individuals and breeds with suitable
temperaments, as not all breeds are suited to high
containment. This is also a human health and safety
issue when working with large animals in contained
spaces.
In addition to the Five Freedoms, at AHVLA we also sum
up our approach as ‘Four Ps’:
–Positive cultural attitude towards identifying any
issues or problems. The facility user group holds
both ‘pre-start’ and ‘wash-up’ meetings before and
after studies, for open discussion and learning for
future studies.
–Proactive ongoing development of welfare score
sheets, continually refining humane endpoints with
the goals of moving to predictive endpoints, as well,
as providing comfort and environmental enrichment.
–Preventative measures to directly or indirectly
reduce suffering, such as building design,
experimental design, developing emotional
resilience in animals.
–Productive approach, acknowledging that better
welfare and better science go hand in hand.
Clicker training in minipigs at
Huntingdon Life Sciences
Finula Sharpe, HLS
Note: much of this presentation has previously been
published in Animal Technology and Welfare, see
Arblaster (2010).4
Introduction
Handling minipigs throughout long term scientific
procedures can pose both welfare and practical
problems. Mature minipigs can weigh 30 kg and
46

Tech-2-Tech
animals of this size and strength are difficult for the
majority of animal technologists to lift or carry safely,
which can result in stress to the animal and physical
strain for the handler.
Clicker training has therefore been introduced as a
method of avoiding stressful carrying and sling
restraint for minipigs undergoing dosing during dermal
studies.5The aim of the clicker training programme at
HLS was to train 52 minipigs to walk voluntarily from
their pen into a dosing trolley, stand for the dosing
procedure and walk back to their pen once it had been
completed.
Method
The training programme used operant conditioning, in
which the minipigs were trained to form positive
associations between certain behaviours and fruit
rewards such as pieces of apple and grape. The
minipigs began training at the age of 4 weeks and each
animal was trained individually in an empty pen for a
session lasting approximately 5 minutes with the
handler seated on the pen floor.
Animals were taken through 6 stages of training (listed
below) which gradually progressed towards walking into
a dosing trolley. If at any point a minipig was showing
signs of confusion then the session would be stopped
and training taken back to the previous stage to re-
establish the desired behaviour.
Stage 1: ‘Click’ of clicker each time the minipig takes a
treat from the trainer’s hand.
Stage 2: Treat and simultaneous ‘click’ of the clicker
each time the minipig moves towards a target.
Stage 3: Treat/click each time they touch the target.
Stage 4: Treat/click each time they follow the target
around the pen.
Stage 5: Treat/click each time they follow the target
out of the pen and back in again.
Stage 6: Treat/click each time they follow the target
into the dosing trolley and out again.
Progress was recorded on a daily ‘tick’ sheet, which
highlighted, for example, if the animal ‘touched the
target’,‘followed the target around the pen’ or was
‘nervous’ of the handler. This enabled us to ascer tain
whether or not animals were ready to move onto the
next stage of training. Training sessions were carried
out daily for 2 weeks before the start of the study.
Outcome
The initial aim of training all minipigs to walk voluntarily
into the dosing trolley was achieved by the third week
of the study. This was aided by animals having access
to their daily food ration while standing in the dosing
trolley, as it provided a distraction from the dosing
procedure. Some minipigs were sufficiently confident
to walk into the dosing trolley by the end of the 2 week
acclimatisation period, while others were more
cautious about taking treats from the trainer’s hand
and took a little longer.
To ensure consistency in the training method, each
technologist involved was given training and a set of
instructions to follow, in addition to the six stages of
training:
Clicker training minipigs
Git is important that the minipig is alone in the pen
during each training session
Gonly reward the minipig when they per form the
desired behaviour
Gthe moment a minipig per forms the desired
behaviour they should be rewarded immediately with
a simultaneous treat and ‘click’
Git is important to have a positive attitude at all times
and not get frustrated if the animal doesn’t per form
what is required, they should be ignored until they
can focus on the trainer again
Gif the minipig becomes confused or unfocused,
training should be stopped and taken back to a safe
phase that the animal is more familiar with
Gif the animal becomes scared and runs away while
training outside the pen, calmly walk (do not run)
after them and bring them back with the target and
a reward
Gbefore training animals to walk into the dosing
trolley, raise and lower the trolley outside the pen so
the minipig can get used to the sound
Gonce trained the minipigs will not need the target
any more, as they will be attracted by their daily food
ration in the trolley
Conclusion
The aim of the clicker training programme was achieved
and this positive behaviour continued throughout the
remainder of the study. While the age of the minipigs
was a significant issue, it was an unavoidable factor
due to the constraints of the study protocol, which
required animals to be 6 weeks of age at the start of
the study. Clicker training did also serve a purpose in
allowing the minipigs to become more familiar with
handlers, outside of normal husbandry procedures.
Clicker training minipigs provides a significant
improvement in the dermal dosing procedures used at
Huntingdon Life Sciences, as it removes the need to
handle the minipigs and therefore reduces stress to
the animals. Using positive reinforcement as a method
of refining dermal dosing procedures therefore
improves animal welfare. It also reduces physical strain
placed on the animal technologists involved. Clicker
training could also be used for other studies using
minipigs, where animals need to be moved from the
animal unit to a dosing room.
47

Tech-2-Tech
Pain assessment and alleviation in
agricultural species used in
research
R. Eddie Clutton, University of Edinburgh
Introduction
Non-avian agricultural species (cattle, sheep, pigs and
goats) are used for a number of different purposes in
research6. Two examples of research fields where pain
may be caused are (i) the investigation of legal, though
potentially painful husbandry operations, e.g.
castration; and (ii) experimental surgery with recovery,
e.g. organ transplantation. Pain assessment and
alleviation are clearly very important in both of these
contexts. When applying a harm-benefit assessment it
may be possible to justify inadequate pain
management in (i), for example to demonstrate that a
husbandry procedure is a welfare issue or to evaluate
potential techniques for pain relief. However,
inadequate pain management can never be justified in
the case of (ii).
Sheep, goats and pigs have also been used to try to
study human pain. In accepting the validity of these
species as human pain ‘models’, the scientific
community must accept that these species feel pain7
and act accordingly: the provision of effective analgesia
when these species undergoing painful procedures is a
matter of justice as well as an act of refinement. The
extensive use of pigs in biomedical research is
constantly justified on the basis of their physiological
and anatomical similarity to human beings so it is
surprising that with few exceptions, the literature
devoted to laboratory pig welfare is notable by its
absence of references to pain recognition and
treatment.
Pain recognition
Acute pain behaviours resulting from husbandry
procedures in young agricultural animals are well
recognised. Little is known about similar behaviours
after experimental surgery and it may be imprudent to
extrapolate information, because age and breed are
confounding factors. Pain evaluation is also
complicated because, in general, agricultural species
show less obvious (to the human eye) pain behaviours
than horses or carnivores. Behaviours may also be
counterintuitive to humans, for example sheep may
display ‘obsessive’ eating behaviour when in pain but
this may be interpreted as animals feeling well. The
fact that some agricultural animals are less likely to
mount ‘fight-or-flight’ reactions may account for their
popularity as ‘large animal’ models (along with legally-
condoned ‘speciesism’).
Unfortunately, all of this means that attempts to
categorise pain behaviours in order to assess pain
severity may be misleading. Indeed, no validated, nor
widely accepted, pain scoring system currently exists
for farm animal species undergoing procedures that
are unrelated to husbandry8. On a more positive note,
if animal technologists are familiar with individual
animals undergoing potentially noxious procedures,
they can be better able to detect subtle behavioural
changes that indicate the presence of pain and initiate
suitable treatment. It can be time-consuming to build
up this level of trust but empathetic staff play an
important role in ensuring that animals are given the
benefit of the doubt with respect to experiencing pain –
and resources should be made available for this.
Pain alleviation
Information on analgesic drugs in agricultural species
(including in the scientific literature) is extremely
limited beyond those used in painful production
disorders, e.g. mastitis. This may be because there
has been no perceived need for analgesia (if pain
behaviours are ‘subdued’) or because legislation
drafted to protect human food supplies limits
commercial incentives for the pharmaceutical industr y
to fund pain studies in ‘food animals’. Determining the
efficacy of ad hoc analgesic therapy is complicated by
the limited pain behavioural repertoire in the species
involved.
However, the literature is gradually increasing and
attitudes are changing. Strategies being increasingly
employed as part of anaesthetic techniques to
minimise an animal’s post-operative discomfort are:
pre-emptive analgesia (the administration of analgesics
before surgical injur y is created); polymodal pain
therapy (the administration of analgesics from several
therapeutic classes at relatively low doses, such that
the combined effect is adequate analgesia, but a
reduced risk of side-effects); partial intravenous
anaesthesia (in which analgesic drugs are used at
higher doses in order to contribute to the state of
anaesthesia and thus reduce the requirement for
general anaesthetic) and prolonged post-operative
analgesia (which involves anti-inflammatory medication
being given until the surgical wound has healed). All
four strategies may be combined within a given
anaesthetic technique.
Conclusion
The dearth of information on pain recognition and
treatment in agricultural animals, relative to companion
and other laboratory species, prompts consideration of
a major limitation on their use in potentially noxious
recovery experiments. There is no justification for
withholding analgesia, or some kind of pain relief
technique, when conducting potentially painful
procedures on agricultural animals.
48

Tech-2-Tech
Assessing pain using the Lamb
Grimace Scale (LGS)
Mirjam Guesgen1and Matt Leach2, (1)
Institute of Veterinary, Animal and Biomedical
Sciences, Massey University, New Zealand
and (2) School of Agriculture, Food & Rural
Development, Newcastle University
Pain in animals is of considerable public concern.
Within biomedical research, pain is of additional
concern because it not only compromises animal
welfare but can also have implications for the validity of
scientific results. In order to address this and alleviate
pain, it is essential to assess its severity and duration
effectively. New, behaviour-based methods for
assessing pain are providing evidence for a
relationship between pain and behaviour and providing
additional tools for pain management.
For example, facial expressions are routinely used to
assess pain in humans, particularly those who are non-
verbal and it has also been shown that human
observers tend to focus on animals’ faces when
assessing pain. This led to studies of ‘pain faces’ in
non-human animals and the use of facial expressions
as indicators of pain is now being developed for several
species, such as rodents9,10,11 and rabbits.12
In these studies, analysis of video stills has enabled
the development of species-specific ‘Grimace Scales’,
which have been shown to be accurate, reliable and
valid measures of pain. A further benefit is that
evaluation studies can be done using video footage of
animals already undergoing scientific or husbandr y
procedures for other purposes, thereby gaining
information that can help to reduce suffering but
without causing additional harms.
The aim of this study was to identify whether lambs
produce noticeable changes in facial expressions when
experiencing pain, and if so, to develop and validate a
Grimace Scale that incorporates these expressions.
First, the LGS was defined using five so-called ‘action
units’: orbital tightening, mouth changes, nose
changes, cheek flattening and ear changes. Each one
is scored on a 3-point scale; 0 = not present, 1 =
moderately present and 3 = obviously present. Figure 1
depicts the Rabbit Grimace Scale12 to give an
impression of the appearance of animal facial action
units; the Lamb Grimace Scale will be published at a
later date.
The LGS was then trialled using still images of the
faces of restrained Romney X lambs from video footage
taken before and after sham docking or routine tail-
docking with a rubber ring in an on-farm situation.
Images were randomly selected and scored for four of
the action units (not ear position, as lambs were
restrained) by four treatment-blind observers. The
results showed that in this instance, the LGS had high
reliability within and between observers, but low
accuracy, i.e. people were scoring consistently but
there was a high proportion of false positives. This was
likely due to restraint stress, which can also influence
facial expressions.
In a follow up study, video sequences of the faces of
unrestrained lambs were taken before and after tail-
docking or sham docking and scored by treatment blind
observers as above, with ear position also included.
This time, reliability was also high within and between
observers, and accuracy was significantly improved,
with fewer false positives.
In summary, we found that the LGS changes in
response to tail docking and that this can be reliably
coded by observers. However, as LGS is also affected
by restraint, further validation is required to see
whether it is possible to identify those action units that
are associated with pain and with distress.
This study demonstrates that lambs change a number
of their facial features when experiencing pain and that
these features can be accurately and reliably coded
and quantified. It is hoped that the LGS can go on to
become a routine tool for pain and welfare
assessment.
Lifetime experiences and actual
severity
Ngaire Dennison, Animals in Scientific
Procedures Inspectorate (ASPI)
The UK Animals (Scientific Procedures) Act, 1986 (ASPA)
requires that a project licence to carry out a programme
of scientific work using living animals may only be
granted where there has been a favourable evaluation of
the proposed project. This evaluation must include a
harm-benefit analysis of the programme, to assess
1
2
3
Figure 1. Facial action unit in a rabbit: orbital tightening
Legend: The eyelid is partially or completely closed,
and the globes may also be drawn in toward the head
so that they protrude less. Score 0 = not present; 1 =
moderately present; 2 = obviously present (the eye
closure reduces the visibility of the eye by more than
half)12.
49

Tech-2-Tech
whether the harms that would be caused to animals in
terms of pain, suffering and distress could be justified
by the expected outcome, taking into account ethical
considerations and the expected benefits to human
beings, animals and the environment.
Part of the prospective assessment of harm requires
that the series of procedures that an animal undergoes
during an experiment be classified as ‘non-recovery’,
‘mild’, ‘moderate’ or ‘severe’, with respect to suffering
that may occur. In addition, due to the changes in the
ASPA required to implement the recently enacted
European Directive 2010/63 EU, there is now a
requirement for the assessment and reporting of the
actual severity of the regulated procedures. There are
many benefits associated with assessing and reporting
actual severity (Table 1) but with the proviso that
judgements have to be suitably informed, objective and
consistent.
Actual severity assessment should:
Gensure that what actually happened to the animal is
recorded and understood
Gprovide evidence-based information for future
prospective harm-benefit assessments
Gfacilitate effective retrospective review
Gidentify areas where the application of the Three Rs
should be focused
Gprovide feedback to help refine ongoing studies
Gprovide input for thematic reviews
Glead to statistics that will better reflect the actual
harms to animals
Gimprove transparency and understanding of the
impact of science on animals
Table 1. Benefits of actual severity assessment.
The law requires that the assessment of actual severity
be made by a competent person and, as with the
prospective classification, classified as non-recovery,
mild, moderate or severe – but in the case of actual
severity it must be based on the actual impact on the
individual animal. A number of sources of information
should be used to determine the experience of the
individual, including day-to-day observations of
behaviours and clinical signs observed throughout the
procedure. This requires good planning, team work and
training, as well as appropriate choices of indicator and
recording systems. Training, including ensuring
appropriate attitudes, is critically important for staff
working with farm animals as these animals can be
perceived by those without the appropriate education
as ‘stoic’ (due to lack of understanding of their
behaviour, as discussed above) or, simply as
‘production’ animals, somehow less deserving of
consideration than companion animal species.
Further guidance on all of the aspects of assessing
actual severity outlined in the paragraph above has
been produced by the European Commission,13,14 which
applies across all species but the examples provided
are of limited help to those working with farm animals
for two main reasons. First, the worked examples are in
rodent and rabbit models and it is important to
consider cases of relevance to large animals to identify
any species- or model related difficulties in assessing
severity for work involving larger animals. Second, the
examples tend to relate to the classification of a single
procedure, whereas farmed animals are often on study
for a long time and are more likely to undergo a series
of experimental techniques. Re-use of farm animals in
further scientific procedures for a second and separate
purpose may be requested for a number of reasons,
including reduction of overall animal numbers,
avoidance of the requirement for naïve animals for
surgery, or for bio-security reasons. There may be an
additional financial element to considerations, if
significant resource has been expended on training or
otherwise preparing the animals.*
There are understandable concerns in the scientific
and welfare communities about achieving compliance
with the legal requirement for assessment of actual
severity, especially in terms of ensuring accurately and
consistency and also with respect to how other
potential causes of discomfort, pain or stress within
the life time experience of the animal (e.g. transport)
should be factored in. People are often anxious that
they will ‘get it wrong’ and may simply opt for reporting
the prospective severity limit, which is not helpful
because it (i) provides little advantage over the
previous system of reporting and (ii) provides no
information about studies that are more or less severe
than expected. Other concerns relate to the time taken
to do the assessment for each animal – but if the
assessment is done properly, the benefits in Table 2
should outweigh this – and confusion about what is
meant by ‘consistency’ in assessments.
It is important to recognise that the level of actual
suffering may differ from the prospective judgement
and also may var y between different animals
undergoing exactly the same procedures, depending
on the individual’s responses to these. That is,
consistency is about everyone scoring each animal in
the same way; it is not about expecting the same
model to give the same results between or even
necessarily within establishments.
––––––––––––––––––––––––––––––––––––––––––––––––
*In the UK, any such re-use must be authorised in advance
by the Secretary of State. The fact that the animal has been
used before must be included within the harm-benefit
assessment, the actual severity of the previous procedures
must be assessed and checked against defined criteria and
a veterinary surgeon with knowledge of the lifetime
experience of the animal has to advise that the animal’s
general state of health and wellbeing have been fully restored
following the previous use.
50

Tech-2-Tech
So, how to achieve this? Besides making full use of the
available guidance from the European Commission
(and the forthcoming guidance from the Home Office in
the UK), it is a good idea to communicate with and
consult others with relevant expertise. Within the
establishment, this could include animal technologists,
the Named Animal Care and Welfare Officer (NACWO)*,
veterinarians the Named Information Officer (NIO),
researchers and internal animal care and use
committees including the Animal Welfare and Ethical
Review Body (AWERB). Externally, useful contacts could
be members of the Home Office Inspectorate (HOI), or
scientists and technicians from other groups doing
similar work.
Communication with the local Home Office Inspector
(HOI) is absolutely essential if the severity limit of the
protocol is found to have been exceeded, e.g. if an
animal on a study with a moderate severity limit has
experienced severe suffering. Some concerns have
been expressed that this would (automatically and in
itself) be non-compliant with the requirements of ASPA.
However, provided each case is promptly reported to
the HOI, as required by standard condition 18 of the
project licence, the user should normally still be
compliant with the Act. Condition 18 would only be
breached if the HOI was not notified.
Discussion session
The rest of the meeting was devoted to a discussion
session in which delegates exchanged views on the
way in which they assessed severity and the kind of
resources they felt might be useful to help achieve
good standards of objectivity and consistency. An
overview of the outcomes is presented below; note that
‘TurningPoint’ was used to survey the audience so the
numbers of responses will vary.
Audience demographics
The audience of around 70 delegates comprised mostly
animal technologists (28, of whom 12 were NACWOs);
also 14 researchers, 9 veterinarians, several members
of AWERBs, NIOs, Named Training and Competency
Officers (NTCOs) and two HOIs. Most people (36)
worked in government research or diagnostic facilities,
with fewer employed within academia (19) and industry
(10). Delegates worked with pigs (44 people), sheep
(34), cattle (31), poultry (23) and goats (10). The
majority of attendees (53) had attended because they
felt there were welfare challenges associated with farm
animal use, whereas 27 people felt reasonably happy
with the situation but wanted to ensure they were
updated.
What people were doing to assess suffering
Day-to-day welfare was mainly assessed by animal
technologists, whom delegates also believed to be best
at the task (see Figure 2). Most delegates (31)
recorded observations using clinical record sheets that
were mostly ‘free text’, whereas fewer (10) used
numerical or binar y ‘score sheets, three used
electronic data logging systems and 11 reported that
the recording system depended on the species and
experimental protocol. In a session with species-
specific breakout groups, delegates listed the indictors
that they commonly used to assess both good and
suboptimal welfare (Appendix 1).
Figure 2. Delegate poll: who assesses animal welfare
day-to-day and who does it best?
Legend: Delegates could vote for up to three options
for ‘who assesses welfare’ and one for ‘who does it
best’. With respect to the latter question, 7 answered
‘it depends’.
Figure 3. Delegate poll: are the Directive and/or the
documents produced by the EC and its Expert Working
Groups of any help with respect to assessing the
welfare of agricultural animals?
Legend: The figures refer to the number of delegates
voting for each option.
––––––––––––––––––––––––––––––––––––––––––––––––
*For non-UK readers, the NACWO is a senior animal
technologist responsible for (inter alia) ensuring day-to-day
care of the animals, and the NIO is responsible for ensuring
that staff have access to relevant information specific to the
species at the establishment.
51

Tech-2-Tech
The requirement to record and repor t actual severity
Delegates were asked whether their facility was ready
to take on the task of assessing actual severity; 19
said ‘yes’, 28 said they were ‘working on it’ and 17
said ‘no’. Most people were entirely, or at least partly,
satisfied with the training they had received relating to
welfare assessment (24 and 22 respectively, as
opposed to 15 who were not satisfied) but only 12 of
67 delegates who answered (18%, see Figure 3) had
read the guidance from the EC on severity assessment
and found it useful.
Discussion on a potential forum for large animal users
A straight question was put to the audience; would you
use a large animal network, such as a discussion
forum or virtual/actual meetings, if one were available?
Fifty-four delegates answered ‘yes’, as opposed to
twelve who replied that they would not. The potential
benefits of a UK forum are set out in Figure 4. There
was a fairly equal preference amongst delegates for an
online discussion forum and meetings and 22 people
felt that the IAT could do more for those working with
large animals. Clearly, any forum would have to be
hosted by an appropriate body, and the audience were
given a number of options to vote for: a scientific
welfare organisation (18 votes), a leading agricultural
research facility (15), the NC3Rs (11), a body like IAT
or LASA (10), ‘other’ (8) or the authorities (1).
While there was support for a UK forum of some kind
at the meeting, Norecopa’s attempt to set up a similar,
multinational forum was met with limited interest.
Further thought and consultation will be necessary as
to the kind of format that could be supported by
relevant bodies. The outcome of this discussion will be
passed on to IAT, LASA, the NC3Rs, the UK Animals in
Science Committee and the Home Office. The RSPCA
and AHVLA are also exploring the potential for another
joint meeting at the time of writing.
Overall action points
The following action points are suggested on the basis
of the talks and discussions on the day:
–Be prepared to accept and recognise emotional
states in agricultural animals; challenge
assumptions to the contrary.
– Consider how a positive emotional state might be
encouraged for the animals in your care, taking into
account (as appropriate) factors such as selection
of species, strain and individual; the potential for
refining housing, care, procedures and the way in
which they are conducted; recognising and
assessing desirable behaviours that indicate
positive wellbeing.
– Review large animal housing and care with
reference to the Five Freedoms, and AHVLA’s ‘Four
Ps’ (above) – including animals housed under all
biosecurity levels.
– If animals in containment do not have suitable
housing and husbandry refinements, identify any
obstacles and work to overcome these.
– Consider whether there is the potential to use
positive reinforcement training within any
procedures, either to obviate restraint or provide
stimulation for the animals.
– Make no assumptions about the nature and level of
pain an animal may be experiencing on the basis of
their behaviour alone. Keep up with the literature
and current thinking on the assessment of pain and
suf fering, including new approaches such as
Grimace Scales.
– Challenge any requirements to withhold analgesia,
or some other appropriate pain management
technique.
– Review the guidance on actual severity assessment
produced by the European Commission and relevant
national regulator (e.g. the UK Home Office), seek
further advice and clarification if needed and
feedback to the authors if the guidance is not useful
at your facility.
If you have any thoughts relating to a discussion forum
for those working with agricultural animals, or ideas for
further meetings on agricultural animal use, contact
research.animals@rspca.org.uk
Acknowledgements
Thank you to all of the delegates and speakers for their
support and enthusiasm for the meeting and for the
useful and interesting discussions. Many thanks also
to AHVLA staff for their ver y welcome assistance, which
helped to make the day such a success.
References
21Webster, J. (2005). Animal Welfare: Limping towards
Eden. Blackwell, Oxford.
22Webster, J., Bollen, P., Grimm, H. and Jennings, M.
(2010). Ethical implications of using the minipig in
regulatory toxicology studies. Journal of Pharmacological
and Toxicological Methods, 62, 160-166.
Figure 4. Delegate poll: what would you want from a UK
forum?
Legend: The figures refer to the number of delegates
voting for each option.
52

Tech-2-Tech
23Webster, J. (2011). Anthropomorphism and zoomorphism;
useful fallacies? Animal Welfare, 20, 29-36.
24Arblaster, F. (2010). Clicker training in minipigs. Animal
Technology and Welfare, 9, 115-118.
25Bertelsen, T., Nielsen, T.C., Lund, G and Gade, L.P.
(2008). The use of clicker training in minipigs. Ellegaard
Newsletter, 30, 6-7, http://minipigs.dk/fileadmin/filer/
Newsletters/Newsletter_30.pdf (last viewed 30 October
2013).
26Norecopa (2012). Consensus Statement: Harmonisation
of the Care and Use of Agricultural Animals in Research,
http://www.norecopa.no/norecopa/vedlegg/Consensus-
statement-agricultural-animals.pdf (last viewed 30
October 2013).
27Morton, D.B. and Griffiths, P.H.M. (1985). Guidelines on
the recognition of pain, distress and discomfort in
experimental animals and an hypothesis for assessment.
Veterinary Record, 116, 431– 436.
28Molony, V and Kent, J.E. (1997). Assessment of acute
pain in farm animals using behavioural and physiological
measurements. Journal of Animal Science, 75, 266-272.
29Langford, D., Bailey, A., Chanda, M., Clarke, S.,
Drummond, T., Echols, S., Glick, S., Ingrao, J., Klassen-
Ross, T., LaCroix-Fralish, M., Matsumiya, L., Sorge, R.,
Sotocinal, S., Tabaka, J., Wong, D., van den
Maagdenberg, A., Ferrari, M., Craig, K. and Mogil,
J. (2010). Coding of facial expressions of pain in the
laboratory mouse. Nature Methods, 7, 447–449.
10 Sotocinal, S., Sorge, R., Tuttle, A., Marton, L.,
Wieskopf, J., Mapplebeck, J.C.S., Wei, P., Zhan, S.,
Zhang, S., McDougall, J.J., King, O.D. and Mogil, J.S.
(2011). The Rat Grimace Scale: A partially automated
method for quantifying pain in the laboratory rat via facial
expressions. Molecular Pain, 7, 55,
http://www.ncbi.nlm.nih.gov/pmc/ar ticles/PMC316360
2/ (last viewed 30 October 2013).
11Leach, M.C., Klaus, K., Miller, A.L., Scotto di Perrotolo,
M., Sotocinal, S.G. and Flecknell, P.A. (2012). The
assessment of post-vasectomy pain in mice using
behaviour and the mouse grimace scale. PLoS ONE 7(4):
e35656
12 Keating, S.C.J., Thomas, A.A., Flecknell, P.A.
and Leach, M.C. (2012). Evaluation of EMLA cream for
preventing pain during tattooing of rabbits: Changes in
physiological, behavioural and facial expression
responses. PLoS ONE 7(9): e44437
13 European Commission (2012). Working Document on a
Severity Assessment Framework, http://ec.europa.eu/
environment/chemicals/lab_animals/interpretation_en.
htm (last viewed 30 October 2013).
14European Commission (2013). Examples to Illustrate the
Process of Severity Classification, Day-to-day
Assessment and Actual Severity Assessment.
http://ec.europa.eu/environment/chemicals/lab_animal
s/interpretation_en.htm (last viewed 30 October 2013).
Appendix 1.
Commonly used indicators of good and suboptimal welfare in cattle, pigs and sheep
These indicators were listed by breakout groups addressing single species. They have been organised into the
categories set out by the EC Working Group on severity assessment13 with the intention of providing examples of
indicators that may be helpful when reviewing welfare assessment for cattle, pigs and sheep. The ‘procedure-specific’
indicators are usually tailored to particular procedures, species, breeds (where appropriate) and stages of
development but the examples in the tables below apply to a range of different procedure types for each species.
Good welfare – cattle
High level Areas to focus on when Specific indicators to monitor
categories observing animals
Appearance Body condition Good body condition (although this can depend on
the breed, e.g. Holstein-Fresians typically appear to
have lower condition scores than other breeds, so
other indicators should be considered in conjunction
with body scores for these)
Coat and skin condition Shiny coat
Environment Enclosure environment, including Normal faecal consistency
any litter, nesting material, Interest in novel objects, curiosity
enrichment items
Behaviours Social interaction Socialisation and group interaction, including
‘playfulness’ and social grooming
Good response to stockperson
Posture and mobility Good posture
Lies down easily
Other Chewing cud
Coming forward and vocalising for food
Alert
53

Tech-2-Tech
Good welfare – pigs
High level Areas to focus on when Specific indicators to monitor
categories observing animals
Appearance Eyes Bright eyes
Environment Enclosure environment, including Foraging, using toys, playing – with low grunting
any litter, nesting material, Inquisitive, curious – e.g. when person enters the
enrichment items room, in response to noise, new objects or change
in environment
Rooting, exploring with nose and mouth contact
Behaviours Social interaction Interacting with other pigs; snorting and running
together
Positive responses to stimulation
Making eye contact, nose up
Posture and mobility Bouncing (‘pop-corning’)
Relaxed posture
Other Tail wagging when happy
Vocalising before feeding, interest in food
Nest building with new bedding, bouncing into new
beds
Barking when excited or anticipating
Quiet vocalisations, e.g. soft grunt when greeting
Scratching dry skin
Good welfare – sheep
High level Areas to focus on when Specific indicators to monitor
categories observing animals
Appearance Body condition Good health
Body functions Food/water intake Good appetite
Behaviours Social interaction Socialisation with humans
Normal interactions with other sheep
Other Expression of normal range of behaviours
Suboptimal or poor welfare – cattle
High level Areas to focus on when Specific indicators to monitor
categories observing animals
Appearance Body condition Poor body condition (although this can depend on
the breed, see above)
Eyes Sunken eyes
Wide open eyes
Other Bloat – this is an immediate cause for concern
Body functions Respiration Rapid, heavy breathing
Food/water intake Depressed appetite
54

Tech-2-Tech
Other Increased heart rate
Decreased lactation
Environment Enclosure environment, including Changes in defaecation
any litter, nesting material,
enrichment items
Behaviours Social interaction Isolation – this is a cause for concern
Posture and mobility Changes in posture, movement or gait
*Reluctance to rise ot increased lying
Other Escape attempts
Agitation
Specific vocalisations associated with pain or fear
Specific tail movements associated with fear
Pawing the ground
Decreased rumination in adult, including dropping
the cud
Grinding teeth – this is a cause for concern
Procedure- These are identified on the basis Sensitivity to touch
specific of the individual project, its Restlessness
indicators potential adverse effects and Depression
expected indicators of these Decreased maintenance behaviours and social
grooming
General malaise; head and ears down, inappetance
Difficulty in changing position
Hunched posture
Dehydration (skin pinch test)
Salivation
Tail flicking
Aggression (including head thrashing), kicking
For infection studies: respiratory changes,
mastitis, gastrointestinal signs e.g. scour, bloat,
dropped cud, blood in faeces
Suboptimal or poor welfare – pigs
High level Areas to focus on when Specific indicators to monitor
categories observing animals
Appearance Body condition Poor body condition
Body functions Food/water intake Poor food consumption or reduced drinking
Body temperature Problems thermoregulating, e.g. panting, blotchy skin,
skin colour changes
Environment Enclosure environment, Diarrhoea
including any litter, nesting Reduced interest in nesting material or toys; not
curious – or even neophobic
Behaviours Social interaction No eye contact, do not come to pen door
Isolation from group
Nervous of other pigs
Undesirable behaviours Fighting, tail biting
Bar biting or chewing, e.g. surfaces,
nipple drinkers, tails, ears
55

56
Tech-2-Tech
Posture and mobility Posture – tail position and carriage, ear position
Other Reduced foraging
Quiet, with reduced vocalisation; or barking when
startled, angry or afraid
Procedure- These are identified on the basis ‘Dog-sitting’
specific of the individual project, its Apathy
indicators potential adverse effects and Restlessness, e.g. postoperatively, shows that
expected indicators of these animal is uncomfortable
Laboured respiration after surger y
Bark or squeal reaction to acute pain, e.g.
intramuscular injection
Changes in skin colour – could be stress or problem
thermoregulating
Regurgitation
Persistent vocalisation is a serious concern
Suboptimal or poor welfare – sheep
High level Areas to focus on when Specific indicators to monitor
categories observing animals
Appearance Body condition Poor body condition score
Coat and skin condition Poor fleece or skin condition
Other Ear position
Body functions Respiration Panting
Food/water intake Inappetance
Other Disease prone
Behaviours Social interaction Isolation from other sheep
Undesirable behaviours Stereotypies
Posture and mobility Changes in gait or lameness
Prolonged lying down
Other Vocalisation
Procedure- Aggression
specific Behaviour signs of fear
indicators Unresponsiveness