References

Bellieni CV. Pain Assessment in Human Fetus and Infants. AAPS J. 2012; 14:(3)456-61

Brondani JT, Mama KR, Luna SPL Validation of the English version of the UNESP-Botucatu multidimensional composite pain scale for assessing post-operative pain in cats. BMC Veterinary Research. 2013; 9 https://doi.org/10.1186/1746-6148-9-143

Calvo G, Holden E, Reid J Development of a behaviour-based measurement tool with defined intervention level for assessing acute pain in cats. J Small Anim Pract. 2014; 55:(12)622-9

Dalla Costa E, Minero M, Lebelt D, Stucke D, Canali E, Leach MC. Development of the Horse Grimace Scale (HGS) as a Pain Assessment Tool in Horses Undergoing Routine Castration. PLoS One. 2014; 9:(3) https://doi.org/10.1371/journal.pone.0092281

Crompton S. Pain assessment and pain scoring models: a review. The Veterinary Nurse. 2010; 1:(1)22-7

Epstein M, Rodan I, Griffenhagan G 2015 AAHA/AAFP Pain Management Guidelines for Dogs and Cats. J Am Anim Hosp Assoc. 2015; 51:(2)67-84

Grant D. Pain Management in Small Animals: a Manual for Veterinary Nurses and Technicians.Philadelphia: Butterworth Heinemann Elsevier;

Keating SC, Thomas AA, Flecknell PA, Leach MC. Evaluation of EMLA Cream for Preventing Pain during Tattooing of Rabbits: Changes in Physiological, Behavioural and Facial Expression Responses. PLoS One. 2012; 7:(9) https://doi.org/10.1371/journal.pone.0044437

Langford DJ, Bailey AL, Chanda ML Coding of facial expressions of pain in the laboratory mouse. Nature Methods. 2010; 7:447-9

WSAVA Guidelines for Recognition, Assessment and Treatment of Pain. 2014. http://www.wsava.org/sites/default/files/jsap_0.pdf (accessed 28th October, 2015)

Nuffield Council on Bioethics. The ethics of research involving animals. 2005. https://nuffieldbioethics.org/wp-content/uploads/The-ethics-of-research-involving-animals-full-report.pdf

Olesen AE1, Andresen T, Staahl C, Drewes AM. Human Experimental Pain Models for Assessing the Therapeutic Efficacy of Analgesic Drugs. Pharmacol Rev. 2012; 64:(3)722-79 https://doi.org/10.1124/pr.111.005447

Sotocinal SG1, Sorge RE, Zaloum A The Rat Grimace Scale: a partially automated method for quantifying pain in the laboratory rat via facial expressions. Mol Pain. 2011; 7 https://doi.org/10.1186/1744-8069-7-55

Pain scoring systems in the canine and feline patient

02 June 2017

Clinical

13 mins read

02 June 2017

Volume 8 · Issue 5

ISSN (print): 2044-0065

ISSN (online): 2052-2959

Abstract

It is important that veterinary professionals are able to recognise and assess pain in patients under their care. Veterinary patients cannot verbalise, or self-report on the amount of pain they are experiencing, so it is the responsibility of the professionals caring for them to be able to recognise the signs of pain, which can differ greatly between species. Once the veterinary professional has recognised the signs of pain during assessment they can then determine and appreciate the type or types of pain the patient is suffering, for example neuropathic, acute and chronic, and create an individual analgesic treatment plan. The assessment of pain in animals is for the most part based on the recognition of behavioural changes in response to pain, which cannot only vary between species but also between individuals within that species. Pain scoring systems can be utilised to quantify pain; putting a number on the level of pain for the purpose of determining whether pain exists, whether analgesia is sufficient, and in order to monitor patient progress in terms of pain management. These take into account the behavioural changes and responses of the patient and guide the user towards the calculation of a score. This article explores some of the key concepts relating to the recognition and quantification of pain, and the reliability and validity of pain scores, before considering the different pain scoring systems available for use in veterinary practice and their relative merits.

Mathews et al (2014: 5) in the World Small Animal Veterinary Association (WSAVA) guideline described pain as the fourth vital sign, and proposed the following definition of pain: ‘a complex multi-dimensional experience involving sensory and affective (emotional) components. In other words, pain is not just about how it feels, but how it makes you feel, and it is those unpleasant feelings that cause the suffering we associate with pain.’ Based on this description, it is clear there are going to be inherent difficulties associated with a veterinary professional being able to quantify pain in animals as veterinary patients are unable to verbalise the location of their pain or indeed how it is making them feel. This problem is similar to that found in human medicine where the professionals are attempting to recognise, assess and quantify pain in (pre-lingual) newborn and infant children, whose ability to verbally communicate is limited or indeed absent (Bellieni, 2012). Bellieni (2012) highlighted that with these pre-lingual patients medical professionals observe and measure a number of physical and behavioural cues, and the responses they record feed into their overall judgement about levels of pain and whether the analgesic plan needs amending. These cues include:

Facial expression — brow bulging, eye squeezing, nasolabial furrow, and an open mouth are all indicators of pain, but are not necessarily exclusively associated with pain. This is similar in animals, where changes in facial expressions in a number of species are thought to be linked to pain, such as in rabbits (Keating et al, 2012), mice (Langford et al, 2010), rats (Sotocinal et al 2011) and horses (Dalla Costa et al, 2014) where ‘grimace scales’ have been developed and tested in research.

Body movements — arm and leg activity, more subtle clenches of their fists and toes are all indicators of pain, however it is recognised that a lack of movement and responsiveness does not mean the infant is not experiencing pain. This also links well to animals as they will often posture in certain ways in order to alleviate the pain they are feeling, or will subtly alter the way they move around and settle down to rest in relation to the pain they are experiencing (Mathews et al, 2014).

Crying — this is thought to be a sign of pain, but again the professionals recognise that it cannot be used as the sole indicator of pain as infants do cry when they are hungry or angry for example. This is similar to animals as many do vocalise when they are experiencing pain, whether that is a cry, a howl, a hiss and so on. However, some animals vocalise in this manner whether they are in pain or not for a number of different reasons. This is where it is important for veterinary professionals to ask the owners about their pet's normal behaviours as they know them best (Epstein et al, 2015), and this baseline data about the patient can inform assessment and consequent nursing interventions. Do they vocalise a lot? When do they vocalise? Have they started vocalising more than usual recently? What is the nature of the change?

Behavioural states — this relates to how quiet or alert the infant is, and also the nature of the transitions between these states. For example, excessive sleep or irritability is generally believed to be indicative of chronic pain states, but if the infant suddenly awakens or cries it can be interpreted as an indicator of acute pain. This can be related to animals as well. The owners are going to be the main source of information here, and the veterinary professionals must ask the right questions to obtain the information they require to make a judgement in relation to pain. The owners might attribute their pet sleeping more or being a little more irritable than normal to the ageing process for example, whereas it could be that the animal actually has a pre-existing condition such as osteoarthritis that is now getting rather painful; it has insidiously worsened over time rather than being an acute incident.

Physiological indicators — heart rate increase, blood pressure increase, oxygen saturation decrease and breathing pattern (rapid, shallow or irregular) variations are all often attributed to the pain response in human infants, and are measured to determine the need for analgesics in many cases. However, as in animal patients, there may well be an underlying disease condition responsible for the physiological changes detected, therefore making them less specific as pain indicators. Underlying conditions should be ruled out and physiological indicators should not be relied on as the sole indicators of pain in any species.

Biological markers — stress hormones such as cortisol and adrenaline can be measured in serum or saliva samples in human infants and animals, which have been proven to be released as part of the pain response. This is good additional data to obtain during pain assessment, however waiting for the sample analysis results to be provided limits the clinical application of these types of tests in practice; professionals cannot wait for these test results to be provided before they alter their analgesic plans for a patient in pain.

Behavioural changes associated with pain in animals can be associated with or influenced by a number of different factors which veterinary professionals must be mindful of, including:

Species — it is essential that the veterinary professional has a good working knowledge of the normal behaviour of a given species in order to be able to recognise any changes that are shown as a result of pain, which links back to the previous point about facial expressions.

Environment — a strange or unfamiliar environment can cause a marked change in an animal's behaviour such as when they are taken to a veterinary practice, which can mask behaviour changes attributable to pain.

Pain location — the actual site of the pain can also influence behavioural changes for instance it is easier to assess a painful joint than abdominal pain.

Previous conditioning — an animal will often over react to an experience that previously it found painful, for example having its nails clipped where the blood vessel was cut as it anticipates that the experience is going to be painful again.

Temperament — the temperament of the animal can also determine its reaction to pain; a naturally quiet dog will tend to become anxious, while a more dominant dog might become aggressive when in pain. Again this is very variable and influenced by many other factors; the veterinary professional must always treat patients as individuals.

Concurrent diseases — both these factors can influence an animal's normal behaviour. Many animals in practice do not potentially have an obvious reason to be in pain, however because of another disease process they are suffering, for example pancreatitis, they are significantly painful so their behaviour alters. The veterinary nurse must be mindful of, and knowledgeable about, the pathophysiology of these types of diseases to ensure they assess the animals thoroughly.

Using the Principle of Analogy is useful in veterinary patients when attempting to recognise and then assess pain; if the professional believes the disease or injury the animal has would be painful if they had the disease or injury themself, it is safe to assume the animal is in pain and analgesia is required. However, anthropomorphism and empathy, while no bad starting point for scoring pain in veterinary patients, may also adversely influence the recognition and interpretation of pain. Vocalisation for example, may lead to higher pain scores being awarded than more subtle indicators, even if the latter represent more severe pain, so veterinary professionals must be mindful of these principles when they are assessing their patients (Nuffield Council on Bioethics, 2005). Epstein et al (2015:69) stated that ‘it is now accepted that the most accurate method for evaluating pain in animals is not by physiological parameters but by observations of behaviour’. Both Mathews et al (2014) and Epstein et al (2015) emphasised that response to appropriate treatment is considered the gold standard in terms of measuring the presence and degree of pain.

Pain scoring: quantifying pain

Increased recognition of the significance of animal pain and the necessity to control it has led to veterinary professionals looking to human medicine for guidance in how to quantify pain, a notoriously difficult task due to the subjectivity of pain experiences. Attitudes towards pain have changed as people are more mindful and concerned about its effects, but still very little work has been done on the scoring and treatment of pain in veterinary species. Veterinary professionals are often responsible for disturbing their patient's natural state and behaviour, and are obliged, then, to return it to normal as promptly as possible.

Pain scoring in animals

The search for an objective pain scoring system for animals remains ongoing. Human medicine considers self-reporting by the patient to be the gold standard against which any other protocol is measured, and clearly this is not an option with animal patients. Problems arise because pain is an experience rather than purely a quantifiable physiological event. Veterinary professionals are attempting to find objective measures for something that is inherently subjective in nature, which will inevitably be flawed due to the previous experience and bias of the observer. The experience of pain is highly variable even with identical stimuli, and varies further between species. With no option of self-reporting, pain in animals can only ever be what the veterinary professional perceives it to be, rather than what the patient perceives it to be. Early attempts to quantify pain in animals were perhaps inevitably based on these protocols used in human medicine, mostly observer-based scales used in prelingual children. These scales tended to rely on subjective evaluation of behaviour without any real proven correlation with behavioural or physiological indicators of pain.

Any pain scoring system based on observation of behaviour relies on the observer being familiar with what is normal. This requires time spent with the patient, time talking to the owner of that patient and also good patient records so that a change of carer is not detrimental. Nevertheless, the accuracy of pain scoring will always be limited by observer bias. In addition to the factors already identified, it is inevitably influenced by personal experience for instance if the observer has undergone a similar procedure themselves, empathy with the patient will be increased, and a higher pain score will be given. The patient itself cannot be subject to the placebo effect but the observer certainly can, for instance if an analgesic has been given, and the observer is confident it will work, lower scores will subsequently be allocated. Humans also tend, when scoring anything, to avoid extremes, so the scores assigned will tend to be clustered in the mid range.

An ideal pain score

There are a few pain scoring systems validated for use in general veterinary practice at the moment, and attempts have been made to devise more pain scoring systems to increase and further improve the recognition of pain and form the basis of guidelines for treatment. The goal, of course, is to reduce the score to as near normal as possible with analgesic interventions.

A scale for assessing pain should, ideally have a means of detecting its presence or absence, as well as of evaluating its magnitude or severity, location and impact on quality of life. Such a scale would also have to be able to account for the differences between visceral, somatic and neuropathic pain, as well as species and age variations. Any intervals on the scale should be equal, and it should be borne in mind that the human ability to differentiate more than 10 to 20 levels of anything is questionable.

Any system that is devised needs to be reliable, sensitive and valid to be useful (Mathews et al, 2014):

Reliability — a reliable scale will give the same score if used in many individuals with similar levels of pain, or the same individual when observed by many people.

Sensitivity — a scale needs to be sensitive in order to detect small variations.

Validity — this can be proven by demonstrating that the system is measuring what it is supposed to, and nothing else.

It is additionally important that any scale is versatile enough to be readily adopted by all members of a practice team, or, indeed, profession. The systems that have been tried have so far relied on correlating different factors such as expected levels of pain, the type of surgery, physiological parameters and spontaneous behaviour on handling, manipulation and interaction. Clearly whatever the system being used, patients should be reassessed regularly, ideally as far as possible by the same person to minimise observer variation.

Types of scoring systems/scales

One (uni)-dimensional pain tools

Simple descriptive scales (SDS), as the name suggests, are the most basic, usually with 3 to 5 grades to choose from, each defined by a short description. The observer watches the patient, handles it and interacts with it, and palpates around a wound for example before assigning a score (Tables 1 and 2). It is perfectly possible to combine more than one SDS scale based on different parameters, behaviours or observations and add the scores together. Scales of this sort are easy enough to use, but are very subjective, and large variations in scores can be given for all the reasons discussed earlier. The quest for user friendliness may also risk over simplification, and consensus about the number of distinct points needed does seem hard to reach.

Visual analogue scales (VAS) employ a line, often 100 mm long, where one end represents no pain at all, the other the worst possible pain (Figure 1). The observer marks a point on the line to correspond with the assessed pain level. No cues are provided for the observer, as the line is not marked off numerically, and the system can be made more relevant by qualifying the one extreme as the worst pain for that particular procedure or condition. The patient is, however, just observed in the kennel or cage, with no interaction or examination involved. This does seem to be a quite sensitive method, and it allows for finer graduation than a SDS, but does still rely on the observer's judgement, which is inevitably subjective. Recording is also apparently affected by the observer's visual and motor co-ordination, which may reduce the ability to accurately place the mark on the line. Observers must be well-trained in the use of such a scoring system, which really provides very little in the way of inbuilt guidance as to what constitutes what (Crompton, 2010).

Numerical rating scales (NRS) are similar to VAS but the observer is asked to choose a number rather than simply marking a point on a line (Figure 2). This system is again more sensitive than an SDS, and may be a good compromise between the lack of sensitivity of an SDS and the unreliability of a VAS, though there still may be considerable variability.

Table 1. Example simple descriptive scale (SDS) for cats

Score	Description
0	No pain, no overt signs of discomfort and no resentment to firm pressure
1	Some pain, no overt signs of discomfort but resentment to firm pressure
2	Moderate pain, some overt signs of discomfort which are made worse by firm pressure
3	Severe pain, obvious signs of persistent discomfort which are made worse by firm pressure

(Grant, 2006)

Table 2. Example simple descriptive scale (SDS) for dogs

Observation	Score	Criteria
Vocalisation	0	No vocalisation
	1	Vocalising, responds to calm voice and stroking
	2	Vocalising, does not respond to calm voice and stroking
Movement	0	None
	1	Frequent position changes
	2	Thrashing
Agitation	0	Asleep or calm
	1	Mild agitation
	2	Moderate agitation
	3	Severe agitation

(Grant, 2006)

Figure 1. An example of a visual analogue scale (VAS).

Figure 2. An example of a numeric rating scale (NRS).

Olesen et al (2012) identified that a limitation with onedimensional pain tools like the VAS and NRS is that they only assess a limited part of the pain experienced by an individual, and suggested that to fully appreciate the complexity of the experience, results from several assessment tools are typically required. Mathews et al (2014) added that the potential for inter-observer variability associated with the use of these tools can limit the reliability, however when used consistently they can be a useful and effective part of any pain assessment protocol. Mathews et al (2014: 9) would recommend the NRS scale in clinical practice due to ‘its enhanced sensitivity over the SDS and increased reliability over the VAS'.

Multi-dimensional/composite pain tools

‘Multi-dimensional pain measurement tools can be used to assess a wider pain experience’ (Olesen et al 2012: 744). The main limitation associated with these, highlighted by Olesen et al (2012) was the time required to complete them when compared with a one-dimensional tool.

Dynamic and interactive visual analogue scales (DIVAS) are a form of modified VAS. The patient is observed at rest, then handled, palpated and walked out before a mark is once more made on a linear scale; hence the ‘dynamic and interactive’ part of its name. This does overcome some of the deficiencies of a simple VAS which may fail to recognise pain in an animal that was just lying still. VAS and DIVAS are considered to be more accurate than a SDS (Crompton, 2010).

Variable rating scales (VRS) include objective physiological data such as heart rate, respiratory rate, pupil size and rectal temperature, as well as spontaneous behaviour, interactive behaviour, response to palpation, mental status and vocalisation. The observer assigns a number from the scale to the patient according to definitions or descriptors provided. These individual scores are then added up to give an overall pain score for that patient (Crompton, 2010) (Table 3).

Table 3. Example variable rating scale for use in dogs

Variable	Criteria	Score
Heart rate	0–10% greater than pre-op value	0
	11–30% greater than pre-op value	1
	31–50% greater than pre-op value	2
	>50% greater than pre-op value	3
Respiratory rate	Normal	0
	Mild abdominal assistance	1
	Marked abdominal assistance	2
Vocalisation	No crying	0
	Crying, responsive to calm voice	1
	Crying, does not respond to calm voice	2
Agitation	Asleep or calm	0
	Mild agitation	1
	Moderate agitation	2
	Severe agitation	3
Response to manipulation	No response	0
	Minimal response, tries to move away	1
	Turns head towards site, slight vocalisation	2
	Turns head with intention to bite, howls

(Grant, 2006)

This does provide a quite sensitive, reliable and valid means of assessing pain in patients, and several academic institutions have given their names to various versions of such a scale. Perhaps the most well known is the University of Melbourne Pain Scale (UMPS), a VRS incorporating physiological and behavioural variables divided into six categories and combining to give a maximum possible score of 27. The Colorado State University (CSU) have produced a number of different pain scoring systems for dogs and cats that combine aspects of the NRS and composite behavioural observations (Mathews et al, 2014), and include:

SDS for dogs

Acute (multi-dimensional) cat

Acute (multi-dimensional) dog

Chronic (multi-dimensional) dog.

The University of Glasgow Veterinary School also has an acute pain scale based on the McGill Pain Questionnaire as used in humans — the Glasgow Composite Measure Pain Scale (GCMPS) and its short form (GCMPS-SF) (Crompton, 2010; Mathews et al, 2014). The GCMPS scale incorporates six behavioural categories, each with four to six descriptors, and again a composite score is given. The words used in the descriptors were suggested by veterinary practitioners. The GCMPS-SF of the scale was produced to make it more user-friendly in practice, and an acute feline scale has recently been tested and proved valid (Calvo et al, 2014). As validity should be paramount to the choice of a scale for use in practice, the fact that both of these Glasgow acute scales have been validated in research should prompt practitioners to choose them for implementation in practice for acute pain assessment in dogs and cats.

Another validated multidimensional pain scale for the assessment of acute pain in cats is the UNESP-Botucatu Multidimensional Composite Pain Scale (MCPS). This is a scale used to assess postoperative pain levels in cats by assessing pain expression, psychomotor changes and physiological variables, which adds up to a score out of 30. The scale was originally devised in Brazilian Portugese, but and English version has subsequently been validated (Brondani et al, 2013). Printable versions of this pain scale are available from the Animal Pain website: http://www.animalpain.com.br/en-us/avaliacao-da-dor-em-gatos.php, where there are explicit directions with regards to the use of the pain scale, and some very useful videos relating to each score criterion to ensure a standardised interpretation of each criterion by users.

Conclusion

The use of pain assessment tools/scales is essential in the optimal management of painful veterinary patients, however their inability to verbalise their pain or discomfort poses additional challenges to veterinary professionals. It is vital, whichever assessment tool or score is used, that patients are reassessed regularly and their response to appropriate treatment documented accurately. The outcome of observations and interactions with patients when these tools are being utilised must be considered alongside the veterinary professional's knowledge of any disease condition the patient has, and its current or historical clinical status to make an informed decision about the overall pain status. There are multiple tools/scales available for use in veterinary practice, and the decision about which to adopt is individual to each practice; it has to be one all staff understand and are able to apply in a standardised manner for it to be useful in managing a patient's pain.

Key points

All veterinary practices should aim to adopt a species-specific and pain typespecific tool for pain assessment in their patients.

One-dimensional pain scales have their place in veterinary practice if they are used consistently, and the numerical rating scale is the recommended scale of this type.

Mulit-dimensional/composite pain scales are preferable over one-dimensional varieties as they are able to assess more facets of the pain experience, and practitioners should aim to use a validated scale.