Assessment of companion animal quality of life in veterinary practice and research



Objectives: Quality-of-life (QOL) assessment is a novel field in veterinary medicine. It shares similarities with the more established fields of human QOL assessment and animal welfare science and could prove similarly useful.

Methods: This paper draws on existing literature on human QOL assessment, animal welfare, philosophy and animal QOL assessment.

Results: It provides practical recommendations for QOL assessment in veterinary practice and in veterinary research on what should be assessed, how it should be assessed and who should assess it.

Clinical Significance: These should assist veterinarians intending to develop their use QOL assessment.


Animal quality-of-life (QOL) assessment is a central component of veterinary practice, but it is still part of the “art” of veterinary practice. More structured approaches to QOL assessment are uncommon in veterinary clinical work, and QOL assessment is a novel discipline in veterinary practice. The academic literature is limited to fundamental theoretical analysis and methodological tools in early stages of development. There is limited guidance available in what is an inevitably multifaceted, complex and controversial matter.

More structured methods of QOL assessment can be used within veterinary practice to inform the clinical decision making and in research to make the results more meaningful in their application in decision making. These require the following five steps: deciding what is important, working out what parameters can be assessed, assessing them, inferring what is important from the parameters and making a decision about actions or recommendations. The second and third can be informed by scientific methods in animal welfare science and human QOL literatures. The first, fourth and last steps require value judgements. Hence, QOL assessment in animals needs to draw upon the disciplines of human QOL assessment, animal welfare, ethics and philosophy to provide recommendations for practitioners and researchers hoping to include more structured QOL assessment in their case management or research.

A fundamental desire to improve an animal’s QOL is a central goal of veterinary practice for all parties involved and is a major concern for owners (Mellanby and others 2003, Chang and others 2006). Yet veterinary work may not always truly achieve such a goal. Interventions dealing with improving issues for owners, such as incontinence treatment for hygiene reasons, aim more at client satisfaction than animal QOL. The monitoring of conditions using biological parameters, such as biochemical results, risks treatments being directed at improving that parameter rather than the QOL for the animal. Furthermore, in all veterinary work, more QOL improvements might be achieved by considering other aspects of the animal’s life that may be more important for the animal than the presenting problem. The use of specific QOL assessment approaches can increase the scope of considerations to cover all the aspects of an animal’s life that are important to that animal. Additionally, formalising QOL assessment can help describe the benefits of treatments in a meaningful way both for individual cases and in scientific reports.

This paper discusses how to increase the benefits of QOL assessment for animals by ensuring that its scope and methods truly reflect what is important for the animal. The paper considers four important questions: when to assess QOL, what to assess, how to perform assessments and who can best assess it.


Human QOL assessment is considered a primary value in decision making by clinicians in individual cases, clinical audit and clinical governance (Sanders and others 1998). It is used to ensure that treatment goes beyond improving health and is actually beneficial to the patient (Bradley 2001), avoid paternalism (Detmar and others 2002), provide prognostic indicators (Rumsfeld and others 1999, Kaplan and others 2007) and assess client perceptions of illness and treatment. Furthermore, studies on human QOL assessment have suggested that it can improve clinician-client relations and client satisfaction with treatment (Kumar and others 2007). For example, Griffiths and others (2007) used a participatory tool on human patients and found that it built “relationship and confidence in the service user that they are being listened to”. Perhaps as a result, QOL assessment can improve compliance with treatment recommendations (Anderson and others 1999, Awad and Voruganti 1999, Kumar and others 2007).

There are several cases in veterinary practice for which QOL assessment is likely to be especially useful. QOL assessment can be used to screen for issues, that is, identify QOL concerns that had not been perceived by an owner. This may be particularly useful as a component of routine geriatric screening. Indirectly, QOL assessment may also be a good way for practices to legitimately encourage owners to consider matters such as obesity, parasite control, analgesia, geriatric concerns, separation-related behaviours and exercise.

QOL assessment may be useful in monitoring changes in QOL over time, especially chronic cases where subtle changes may benefit from more formal approaches. Treatment choices that are likely to significantly impact on QOL, such as chemotherapy and analgesia, may benefit from explicit assessment of QOL. In veterinary practice, euthanasia decisions often involve elements of QOL assessment, specifically in assessing whether the animal’s life is worth living or not. It may also be a sincere way to increase customer involvement and satisfaction.

In addition, the process of assessment may also have beneficial effects on patient QOL. The assessment itself (not just the result) can raise awareness of the broad scope of issues that may affect an animal’s QOL. As QOL aims to identify aspects of the animal’s life that is important for it, this may be different or encompass more issues than are considered by the health care professional whose primary focus may be the specific health-related outcome they believe they are tasked to solve.

In veterinary research, QOL assessment can be useful to monitor both longitudinal changes over time and comparisons between treatment groups. For example, Craven and others (2004) observed a highly significant association of owner-rated QOL and outcome, concluding that “owner perceptions are a valid aspect of disease monitoring”. QOL assessment additionally assists in ensuring that research provides meaningful patient-related outcome measurements beyond the traditional clinical and pathological assessments. This allows results to be represented to clinicians and clients in ways that are directly relevant to their decision making.

Research areas where QOL assessment will be important include all in which an intervention may be hypothesised as having deleterious or beneficial QOL effects. It is especially appropriate for those that are palliative or expected to have relatively low chances of effecting a cure, such as chemotherapeutic protocols (Mellanby and others 2003, Tzannes and others 2008), or those where the aim is to improve QOL rather than life expectancy. The evaluation of novel surgeries or treatment regimes should include assessment of anticipated QOL effects and screening for unexpected negative QOL effects from the novel treatment protocols. This may become especially significant in clinical trials.


The question of what is QOL is a deep philosophical matter that has been debated since ethics began. The authors suggest that animal QOL can be best approached practically by firstly deciding what is important for the animal, and secondly by working out what parameters can be assessed for use in making decisions. This is a combination of assessing those qualities of a life from the animal’s point of view and the assessment from the observer’s point of view. This combination is necessary: the first half alone could lead to futile attempts to know absolutely everything from the animal’s point of view, whereas the latter half alone could lead to reducing the concept of QOL to what can be observed from the outside rather what actually matters to the animal.

Characterising QOL assessment in this way establishes the kind of questions that need to be asked. These are complex questions but assistance is available from more established paradigms. Many authors argue that one can meaningfully compare human and animal QOL (Appleby and Sandøe 2002, McMillan 2003). Others consider that animal QOL and animal welfare are similar or equivalent (DeGrazia 1998, Fraser 1998, Mullan and Main 2007, Wojciechowska and Hewson 2005). The approaches used by human QOL and animal welfare science cannot be applied unthinkingly. Human QOL uses many anthropocentric terms (Bowling 2005, pp. 125-126), such as happiness, subjective well-being and life satisfaction (for definitions see Diener 2006), the uncritical use of which is unsuitable for animal assessments. Animal welfare science has been appropriated by scientific and economic demands in a way that makes it less suitable for clinical QOL assessment. Furthermore, there is no universally agreed method of assessing either human QOL (Bowling 2005) or animal welfare (Appleby and Sandøe 2002, Nordenfelt 2006).

These considerations, and analysis of the human QOL and animal welfare literatures, suggest two main types of approach to animal QOL assessment: mental state approaches and external parameter approaches. The first is more subjective, the latter more objective, but both approaches involve subjective and objective elements, and this represents one balance that needs to be appropriate to the purpose of the specific assessment. The ideal approach may be concerned with mental states but, in practice, external parameters can be the only assessable QOL parameters.

Mental state approaches concentrate on attempting to evaluate the feelings and emotions of the animal. Important experiences include pain and suffering, and some veterinary research has assessed such indicators (Hellyer and others 2007). Other more positive experiences, such as pleasure and enjoyment and relationships and control, are also valuable (Yeates and Main 2008). This can involve questions aimed specifically at tapping the respondent’s empathy. For example, Tzannes and others (2008) asked the owner how they thought their cat would rate its QOL during therapy, and Mullan and Main (2007) asked owners how willing they would be to take on the life their pet is now living.

The main advantage of mental state approaches is that they address what directly matters from the animal’s point of view. This approach is the central goal of human QOL assessment (Schwartz and Rapkin 2004), and animal welfare science is also increasingly making attempts to infer animals’ experiences (Duncan 1996, Spruijt and others 2001). Furthermore, mental states approaches, by encouraging reflection and empathy, may alter assessors’ behaviour, leading to improvements in the animal’s QOL. This can make QOL assessment as both a measurement and an intervention.

One disadvantage of mental state approaches is that, because animals cannot self-report as human patients can, they necessitate the assessor having to “mind-read” the animal and convert their observations into linguistic or quantifiable parameters. This necessitates using proxies, such as owners, veterinarians, nurses, or other professionals. Different proxies are likely to have different response styles (Jokovic and others 2004, Davis and others 2007), and it might be expected that they will assess subjective elements of animal QOL differently. Many human QOL tools, especially those designed for children, are designed in self- and proxy-assessment formats (David and others 2006), and comparing results from each reveals significant disagreements between assessments. In different cases, proxies rated QOL higher or lower than patients to varying extents (Sprangers and Aaronson 1992, Chang and Yeh 2005, Jozefiak and others 2008). Differences can depend on the illness (Varni and others 2006, Wilson-Genderson and others 2007) and the facets considered (Eiser and Morse 2001, Hexdall and Huebner 2007). One veterinary research paper incidentally reported some differences between proxies in assessing demeanour (Graham and others 2002). Such discrepancies may represent significant concerns about the validity of either group of responses, or it may be that disagreements are over-represented in literature. Their existence in veterinary contexts is unstudied. It may, however, be significant in research, and the use of a single consistent assessor warranted for reliability.

External parameters approaches focus on parameters that can be directly observed from the outside. External parameters may concern the provision of “inputs” to an animal’s QOL and “outcomes” other than experiences (Appleby and Sandøe 2002). Important inputs might include levels of care (Marinelli and others 2006) and therapy (Föllmi and others 2007) and the animal’s environment (Ng and others 2005) and its social interactions (Bengtsson-Tops and Hansson 1999, Marinelli and others 2006). Inputs also include the characteristics of the individual including its breed, physical condition and health (Marinelli and others 2006, Wojciechowska and Hewson, 2005). Non-experiential outcomes include achievements (Parfit 1984, Keyes and others 2002), (un)naturalness (Rollin 1996, Alrøe and others 2001) and biological functioning or productivity (Barnett and Hemsworth 1990, Broom and Johnson 1993, Moberg and Mench 2000, Gregory 2004).

The use of external parameters may be more useful for researchers, especially in cross-sectional studies. External parameters are directly observable and there is likely to be greater agreement between individual proxy assessors. In practice, they are cognitively less demanding for assessors, and directing assessors to look at specific inputs may actively encourage better provision of those inputs. Some veterinary studies have usefully assessed particular facets of QOL. Graham and others (2002) assessed QOL by the use of two parameters, demeanour and activity. Their study considered diabetes mellitus, for which these are especially relevant facets. Lod and Podell (1999) asked about changes in the animal’s attitude and activity level since before treatment, and Chang and others (2006) also assessed particular aspects of QOL by asking owners about changes in their pet’s level of activity and interaction with family members.

However, it must be remembered that external parameters are indicators; they should not be taken to be QOL. To redefine QOL as what we can observe from the outside would be erroneous. External parameter approaches involve controversial assumptions and subjective judgements in deciding what inputs or outcomes are important. There is a risk that assessors choose external parameters that are less important, or even irrelevant, from the animal’s point of view simply because they are more objective and assessment of overall QOL would be consequently skewed by the underlying assumptions.


The central question considered in this paper is how QOL assessment can be approached. General approaches range from being almost totally unstructured and qualitative to being completely formalised, quantitative measures.


Owners often unreflectively evaluate their animal’s QOL based on prereflective empathy and intuition and their first-hand experience of their animal. This “gut feeling” can be easily tapped by qualitative assessments in a non-systematic way, without using predetermined structures. One example is the use of open questions such as “How do you think your [animal] would rate its quality of life?” (Mullan and Main 2007, Tzannes and others 2008). In practice, these questions can also be less explicit and questions such as “How is your dog at the moment?” and “How’s your cat getting on?” may generate useful responses.

These can be quantified in different ways. Craven and others (2004) recorded overall QOL for dogs with inflammatory bowel disease as rated by their owners on a one to 10 numerical scale. Similarly, Tzannes and others (2008) asked owners to record their cat’s QOL on a 10-point scale, with anchors of “could not be worse” and “could not be better”. They also asked separate questions about whether each cat had always good days, always bad days or more good days than bad days or vice versa and how the owner thought the cat would rate its QOL during therapy, thereby providing three parameters to triangulate. Mellanby and others (2003) recorded owner-rated QOL in a three-point scale of “the same as before the condition, “acceptable but not as good” and “unacceptable” and asked owners whether or not therapy improved their dogs’ QOL and about side effects.

A different style was used by Chang and others (2006), who asked owners to record how well they agreed with the statement “My dog’s quality of life is good” on a visual analogue scale and whether QOL had changed. Similarly, Lod and Podell (1999) assessed owner’s assessment of their animal’s overall QOL, attitude and activity by asking owners the extent to which they agreed with statements such as “[their dog] is leading an unacceptable poor quality of life”. While phrasing such as “unacceptable” does not preclude this response being selected, such emotive and value-laden terms might be thought to reduce its likelihood. An owner answering that they think their animal’s QOL to be “unacceptable” could make them feel vulnerable to censure, guilt and even prosecution for not having had their animal euthanased despite it having an “unacceptable” QOL. Indeed, between two studies in which it was used only one respondent gave this response (Lod and Podell 1999, Mellanby and others, 2003).

Broad open questions can be further used to prompt further questions to explore the animal’s mental state and external parameters. If the owner reports a dog as “a bit off colour”, more specific questions can be used that focus in on the relevant mental states and observable behaviours, such as “What behaviours have you seen that makes you think that?” or “Do you think the dog is in pain?”. Similarly, if the owner describes a certain token of behaviour such as “slower during walks”, the discussion can focus on the possible mental or physical states, for example whether it is because of pain (mental state) or weakness (that is a physical parameter). Ideally, the practitioner can think what mental state aspects are important, a relevant question formulated about relevant external parameters, the answers gathered and an inference about the mental state made. This is especially useful in practice, but can also be used in research. Item response theory, as described below, can allow follow-up questions within formal tools. Less formally, qualitative questions by researchers can add meaning for practitioners. For example, Chang and others (2006) provided additional information for practitioners by recording reasons stated for deterioration in overall QOL.

This approach can be highly specific to the individual animal and allow the owner to use information and intuition at their disposal. A disadvantage is that the method is highly subjective and the question itself is vague and ambiguous both in terms of which facets should be included and how they should be subjectively aggregated. Consequently, the assessment of each animal will vary with the response style of each owner because each will understand and approach the question differently, and each owner may have different relationships with their animals.


The frameworks presented in Table 1 can be used in consultations or to provide appropriate research questions. These can structure the closing of open questions or to generate questions in the first instance.

Table 1. Possible approaches to QOL assessment
Three concepts framework (Fraser and others 1997)
 This approach distinguishes three different conceptualisations of animal welfare
  Physical – biological states of animals are measured (for example physical, physiological and pathological characteristics)
  Mental – emotional states of animals are considered
  Natural – the states of an animal (for example behaviour) is compared with its wild counterparts
Five Freedoms (Farm Animal Welfare Council 1993)
 This approach combines elements of the three different concepts of animal welfare
  Freedom from hunger and thirst by ready access to fresh water and a diet to maintain full health and vigour
  Freedom from discomfort by providing an appropriate environment and a comfortable resting area
  Freedom from pain, injury and disease by prevention or rapid diagnosis and treatment
  Freedom to express normal behaviour by providing sufficient space, proper facilities and the company of the animal’s own kind
  Freedom from fear and distress by ensuring conditions and treatment which avoid mental suffering
Input assessment
 This approach concentrates on the external parameters
Affective assessments (McMillan 2003, Morton 2007)
 These approaches concentrate on the experiences of animals
 Positive feelings
 Negative feeling

The three concepts (mental, physical and natural) used with the various definitions of animal welfare (Fraser and others 1997) can be useful for asking fundamental questions about an animal’s QOL. The mental aspect implements the mental state approach above. The physical and naturalness aspects are external parameters. Physical states refer to the function of the animal in terms of biology and health. Naturalness assesses the external parameters in relation to how well they correspond to an animal’s nature. An intervention may affect each differently and these assessments may conflict. For example, reducing an animal’s anxiety with a tranquilliser in a particular environment may have a beneficial effect on the animal’s mental state, but the animal’s inability to respond normally may reduce its ability to cope biologically and is less natural than an animal responding appropriately. Thus, the approach does not provide a definitive answer but is more a means to generate considerations and relevant questions.

The “Five Freedoms” are a well-established framework (Farm Animal Welfare Council 1993) and, although developed for farm animals, they can be applied to companion animals. They represent a combination of both mental state and external parameter approaches. It can be useful for practitioners to informally consider each freedom separately to establish what the animal’s needs are and whether they can still be fulfilled. In assessing the impact of sudden onset blindness, for example, it can remind veterinarians to consider whether access to food and water are an issue, whether the blindness leads to fear or distress and whether it prevents the animal’s normal behaviour, such as companionship with people and other dogs, and exploratory behaviour during walks. The Five Freedoms can be related to care of the animal by considering what resources are necessary to ensure satisfaction of the identified needs.

An input-assessment framework can assist in focusing on whether the animal’s QOL can be improved by providing different resources. Such resource factors that can influence QOL can be related to the (1) carer (for example knowledge, skills and attitude), (2) environment (for example food, lying area and space) and (3) animal itself (for example breed and temperament). This approach can be usefully combined with the Five Freedoms as it is useful to examine what carer, environment and animal factors are necessary to provide for each QOL aspect defined within the Five Freedoms. For example, improvement of separation-related distress may be achievable by different handling or training by the carer, altering the environment in which the animal is separated and altering the personality of the animal itself.

The affective assessment frameworks, which consider the negative and positive experiences of an animal, provides an assessment of what matters directly. McMillan (2003) has produced a quick list of questions that relate to positive and negative feelings. Consideration of positive experiences can be easily overlooked during the management of chronic illnesses as the focus of the intervention is normally focused on managing the negative consequences such as pain. Yeates and Main (2008) proposed a framework for considering positive experiences based on (a) pleasures (eating, play, tactile pleasure, exercise, sleep, and thermal comfort), (b) engagement (curiosity, novelty, confidence, optimism and contact with other animals, humans, objects) and (c) realisation (achievement, space and opportunity).


More objective assessment methods may be more suitable for researchers or clinicians who require qualitative data. Assessments of different facets can be structured, usually in a questionnaire-style format, and these can be aggregated together to assess overall QOL either by the subjective interpretation of the veterinarian or by predetermined formal methods. These can be based on responses from a single proxy or can incorporate the responses from owners and members of the veterinary team.

There can be considerable work involved in administering more formal tools for use in human QOL assessment: selecting or designing tools, distributing tools, assisting respondents where necessary, collecting responses, validating responses, data entry, analysing and interpreting results and determining further action (Bliven and others 2001, Jones and others 2007). In a veterinary context, the use of tools might also indirectly add to workload by increasing requests from owners for further information. On the other hand, appropriate use of QOL tools may decrease veterinarians’ workload. Preconsult screening by owners in the waiting room or in nurse clinics may be used to increase efficiency during history taking.

The pain assessment tools (Holton and others 2001, Wiseman-Orr and others 2004) are intended for both practical and research settings. Their scope is largely restricted to pain assessment. Assessors give scores on number of external parameters based on their observations of the animal’s behaviour, and the overall assessment provides a summative inference about the animal’s mental state. By making assessment consistent, the tool increases both intraobserver and interobserver reliability. Validity has been demonstrated in several settings. One main advantage is that it allows quantitative assessment. One main disadvantage is the time costs.

Morton (2007) proposed a numerical scoring system that involved assessing behavioural parameters that may infer positive and negative mental states. The obvious advantage is that the quantitative system could be used to measure longitudinal change in an individual animal or cross-sectional variation in research protocols. A weakness is that it may not engage the assessor and, until it is validated, the possibility of subjective variation may cause scepticism over the validity of the numerical scoring system.

The QOL screening tool (Mullan and Main 2007) has different sections for assessors to rate inputs, subjective assessment of mental states and behavioural parameters. It includes a summary of the pain assessment tool of Wiseman-Orr and others (2004). It is not intended to compare between dogs and is therefore inappropriate for cross-sectional research. In the practice setting, it may be useful for screening for issues and encouraging reflection on how the animal’s care could be improved. Its benefit is increased by its combining a number of approaches from a wide scope of concerns, but this comes with significant time costs.

A possible future development for veterinary QOL includes the use of item banks and item response theory (Hays and others 2000, Ware and others 2000, Bjorner and others 2007, Edelen and Reeve 2007, Martin and others 2007, Reeve and others 2007). In such methods, follow-up questions are selected from a stock, based on the respondent’s responses to previous questions and according to an algorithm. This allows particular respondents to give detailed information without the same questions being asked of respondents for whom they would be irrelevant.

The choice or design of tool involves balancing different demands. Although those who develop tools will favour those tools (Hyland 2002), there is – and can be – no such thing as an absolute, perfect QOL approach. Each tool has its limitations (Scientific Advisory Committee of the Medical Outcomes Trust 2002). Tools should be chosen for a balance of sensitivity (responsiveness for longitudinal tools and discrimination for cross-sectional tools), range, reliability, subjectivity, specificity (for example to individual); universality and feasibility that is appropriate for the use.

As with the general approach, the choice of tool should depend on the purpose of the assessment. The existence of over a thousand different human patient-related outcome measures (Bielli and others 2004) is partly because of the increasing recognition that tools must be appropriate to the specific needs of the clinician. Approaches should be appropriate to the patients (breed, age and conditions), to the owners (feasibility, cognitive abilities, demographics of client, response style, including any shift in responses, etc.), to the administrator (feasibility and knowledge) and to the purpose of the assessment and use of the information once it is gathered.


The use of a more interactive participatory assessment tool is rare in veterinary practice but may be a development of the frameworks and formal tools. These can maximise the usefulness of the first-hand experience and empathy of the owner, while increasing and directing their reflection. Studies have shown that participatory methods based on dialogue or pictorial responses can be valid and reliable (Griffiths and others 2007) and can improve reflection and clinician-client dialogue and relations. In addition “think-aloud” techniques can allow researchers and clinicians to understand the bases and significances of the owner’s prereflective assessments (Jobe 2003). No such tool currently exists for veterinary practice, but work is currently underway to develop such approaches.


The impossibility of self-assessment, the unavoidability of subjectivity and the disagreements between proxies mean that the identity of the QOL assessor is an important question for veterinarians. Candidate groups of proxies include the owners, veterinarians, nurses and researchers. The disagreements between these groups should not be taken to mean that any one group of assessors is wrong. Each proxy provides separate, independent information (Higashi and others 2005, Varni and others 2007) and each will have strengths and limitations.

Where objectivity or reliability is especially desirable, trained clinical staff may be the most suitable proxies. Veterinary surgeons will generally have more experience of more structured QOL assessment than owners. Veterinary nurses can spend more time than surgeons with inpatients, but rarely with outpatients. Assessment by veterinary surgeons and nurses can also help them to avoid over-treatment and improve the meaningfulness of their advice to owners.

In comparison, owners will have more experience of the individual animal (McMillan 2003). They often know the animal’s history and spend more time with it in a greater range of environments and activities than the veterinary staff could feasibly achieve. This allows them both more objective knowledge of the animal’s external parameters and more awareness of how these matter from the individual animal’s point of view. This may be especially advantageous when greater specificity or range is required, such as screening or assessing longitudinal changes in individual cases. Assessment by owners may increase their level of reflection concerning their animal. This can help owners to perceive problems or improvements, which may lead to desirable behaviour changes. For example, parental QOL assessment can determine whether children receive medical care (Janicke and others 2001) and the same is plausible in the veterinary context.

The limitations of each proxy might be partially offset by combining the strengths of different proxies in various forms of joint assessment (Föllmi and others 2007). This can be performed by one proxy with some strengths formalising the style of response remotely, for example by acting as a tool designer, and another proxy completing the specific assessment. This approach has been used in determining pain in dogs (Holton and others 2001, Wiseman-Orr and others 2004). Alternatively, joint assessment can involve face-to-face dialogue between parties, which may improve mutual respect and cooperation. Graham and others (2002) used this approach for assessing patients’ activity, which was scored by the veterinarian after discussion with the owner.


Although QOL assessment is now fairly well established in human medicine, the introduction and development of any QOL assessment method in veterinary practice and research is dependent upon demonstrating its feasibility and benefits to potential users, in terms of time, money and ease. If a method is not considered feasible, then it is unlikely to be implemented. Another barrier may be scepticism about the benefits. This may be because of unawareness of the possibilities. In some cases, this is deliberate, as some vets consider themselves to be concerned with health, rather than QOL, or they may equate QOL with health. Other practitioners may consider that their informal methods of QOL assessment, developed over hundreds of consultations, cannot be improved. Other veterinary surgeons may be concerned about QOL but perceive limited value in the existing measures and feel unable to develop novel ones themselves.

These problems will hopefully improve as time goes on. Vets may become increasingly aware that their area of expertise and responsibilities to animals go beyond health-related matters. In other aspects of practice, such as consulting and public relations skills, more vets are realising that unreflective or informal methods can be improved and that consulting room skills can be learnt by methods other than trial-and-error, and this mindset may encourage vets to develop QOL assessment skills and to share their experiences with others. For example, Mellanby and others (2003) were usefully explicit about ways that their assessment could have been improved. Perceived value may be improved as more approaches are developed and propagated.

However, the idea of an absolutely perfect QOL measure that takes no time is completely comprehensive and is without controversial assumptions is unrealistic (Hyland 2003). Despite the extensive research field, there is no universally agreed method of assessing human QOL (Bradlyn and others 1995, Clarke and Eiser 2004). It is therefore important that an approach to animal QOL assessment has a balance of limitations and benefits appropriate to the purpose of the assessment. This balance is of primary importance in deciding what, who and how to assess animal QOL.

The impossibility of an absolute, perfect QOL tool should not prevent clinicians or researchers from using and developing existing tools. Many of the advantages of QOL assessment can be achieved by any QOL assessment. Asking both staff and owners simple questions such as “How is [your] dog’s QOL?” may encourage subjective, individual-based reflection, improve client satisfaction and ensure that medicine is always in the interests of its patients. The feasibility of simpler methods can offset the possible lower statistical reliability and objectivity. The advantages of more sophisticated tools can enhance veterinary QOL assessment further and clinicians and researchers designing or selecting tools to maximise the benefits of QOL assessment should use the best theoretical bases and methodological approaches available.

However it is performed, companion animal QOL assessment represents an excellent opportunity for small animal veterinarians to reflect upon what is important for animals, and what is important for veterinarians to assess.


The authors thank Martin Owen and Michael Stevenson for reading drafts of this article and Petsavers which has funded James Yeates’ residency in animal welfare, ethics and law.

Petsavers Dog Quality of Life Project

Advising owners on how to improve their animal’s lives is a fundamental desire of veterinary practice. James Yeates, the Petsavers funded resident in animal welfare, ethics and law will soon be contacting BSAVA members regarding a newly funded Petsavers project. This project will investigate important issues, as perceived by vets, and look at ways of improving these issues. BSAVA members will shortly be contacted by email and provided with a link that will lead them to a two-page questionnaire (which should take about four minutes to complete). Vets will be asked to share their opinions on issues concerning particular species. If you are not a BSAVA member but would like to contribute to the discussion you can contact James Yeates - - for further information.