Abstract
This study evaluated the types of items owners consider important to their cats’ quality of life (QoL). We hypothesized that items contributing to QoL in cats are predominantly items requiring mobility. The objectives of the study were to describe the types of items considered important by owners for their cats’ QoL; to describe the proportion of these items that involve mobility; to evaluate what patient factors, including severity of degenerative joint disease (DJD), affect this distribution; and to evaluate whether the proportion of QoL items involving mobility chosen by owners is different in cats presenting for a DJD study compared with a randomly selected population. A total of 830 client-generated items were evaluated. Regardless of DJD status, 40% of items listed by owners involved mobility, while 60% were ‘inactive’ items, rejecting our hypothesis. This highlights the need to assess non-active items that owners consider to constitute QoL to fully assess the impact of diseases like DJD and, therefore, the success of therapeutic interventions.
Introduction
The importance of patient reported outcomes (PROs) in human medicine is widely recognized. 1 These PROs may refer to a large variety of different health data reported by patients, such as clinical signs, functional status, quality of life (QoL) and health-related quality of life (HRQoL). 1 In humans, QoL is a complex, abstract, multidimensional concept defining an individual’s satisfaction with life in domains he or she considers important. The term HRQoL reflects an attempt to restrict this complex concept to those aspects of life specifically related to the individual’s health and which are potentially modifiable by healthcare. 2
Both QoL and HRQoL have been used in veterinary medicine. 3 QoL indices are generally broader and include aspects not necessarily impacted by particular diseases. 3 QoL measures used in veterinary medicine vary from simple scales tied to certain descriptors of behaviors, 4 to broad, unconstrained assessments.5–7 Questionnaires have been developed to assess HRQoL in dogs with cardiac disease, 8 cancer,9–11 chronic pain,12,13 spinal cord injuries14,15 and atopic dermatitis. 16 Some work has also been performed to assess QoL or HRQoL in cats being treated with antiviral agents, 17 and cats with cardiac disease,18,19 cancer 20 and diabetes. 21 Complex QoL scales have also been developed and proposed, with questions covering several different proposed aspects of QoL, including social interaction and opportunities for pleasure.22,23 QoL assessments are often used to capture effects of a treatment that may not otherwise be captured.5–7 However, little is known about what owners are considering when they are asked about their pet’s overall QoL. The aspects or items that owners are focused on when assessing QoL may affect the sensitivity of the measure.
Degenerative joint disease (DJD) is now recognized as being very common in the feline population.24–26 DJD is an umbrella term that includes degeneration in both cartilaginous and synovial joints. It covers both inflammatory causes of joint degeneration, as well as non-inflammatory causes such as osteoarthritis (OA). It has been argued that until more is known about the etiology of joint disease in cats, the broader term of DJD should be used, rather than the more narrow descriptor, OA.24,25 Several studies have investigated the role of pain and mobility in cats with DJD/OA.7,27,28 The assumption has been made that decreased mobility and increased pain negatively affect overall QoL for cats with DJD/OA. However, the importance of mobility, or activities requiring mobility, to a cat owner’s consideration of quality of life has not been established. In order to use QoL assessments to monitor disease progression or therapeutic interventions, particularly interventions designed to increase mobility, it is important to better understand what items cat owners consider to be important when evaluating this measure.
The objectives of the present study were to:
describe the types of items considered important by owners for their cats’ quality of life;
to describe the proportion of these items that involve mobility and the relative importance assigned to these items;
to evaluate what patient factors, including the severity of DJD, affect this distribution;
to evaluate whether the proportion of QoL items involving mobility chosen by owners is different in a population of cats presenting for a DJD study versus those selected randomly for a health screen.
Our hypothesis was that the items owners would identify as important for their cat’s QoL would primarily involve mobility, and that this would be unaffected by the cat demographics and presence of DJD. We tested our hypothesis using data gathered from the administration of a QoL questionnaire, similar to that used in dogs, 14 and data gathered on patient demographics and radiographic DJD scores.
Materials and methods
Data evaluated in this study was collected from three different studies, all of which were approved by the Institutional Animal Care and Use Committee at North Carolina State University.
Subjects
A total of 166 client-owned cats were included in the study — 99 were a randomly selected sample of cats from a single practice, used for a cross-sectional study evaluating the prevalence of DJD 25 (‘prevalence group’) and 67 cats were recruited to one of two mobility studies assessing the efficacy of a feline therapeutic diet 6 or a nutritional supplement (unpublished data) (‘DJD study group’). In each case, owners provided informed consent for study procedures. All owners were 18 years or older and lived in the home with the cat.
Overall study outline
This study was conducted as a prospective, observational study and was divided into two parts. In both parts, owners were asked to list the items that they considered important for their cat’s quality of life (QoL). These items were then divided into appropriate domains of behavior and further designated by the degree of mobility thought to be involved. Once categorized, the distributions of the items were described and related to characteristics of the cat and the degree of radiographic degenerative joint disease (DJD) and joint pain present. Comparisons within, and between, groups of cats were made.
Part I: prevalence group
The population of cats used in this part has been described previously. 25 It consisted of a cohort of 100 cats randomly selected from a single practice database within four equally distributed age groups between 6 months and 20 years. Owners were told the study was a health screen study to avoid introducing bias. One of the owners of these 100 cats did not complete the QoL assessment correctly and so data from 99 cats were included.
The cats’ age, weight, sex and body condition score (BCS) were recorded. An orthopedic evaluation was performed, a pain score assigned to every joint and each segment of the axial skeleton, and a total pain score determined, as described previously 29 and which is detailed in Supplementary material as Appendix 1. Based on the reaction to evaluation, a temperament score was assigned to each cat, as described previously. 30 Under sedation, orthogonal radiographs of all joints and the spine were taken and evaluated as described previously25,29,31,32 and as detailed in Supplementary material as Appendix 1. Based on the data obtained, two extreme groups of cats were identified (n = 22 in each group): musculoskeletally healthy cats with few or no radiographic signs of DJD and no signs of pain on manipulation of any joint or segment of the axial skeleton; and cats with the most severe radiographic evidence of DJD, in which signs of pain were also detected during joint manipulation. 29 These two extreme groups were designated as a ‘normal’ group and a ‘DJD group,’ respectively. Cats in the ‘normal’ group may have had minor indications of radiographic DJD (given the high prevalence of radiographic DJD), but these were considered clinically unimportant.
Part II: DJD study group
A population of 67 cats, not included in part I, that had been specifically recruited to ‘DJD’ studies evaluating the efficacy of a putative therapeutic diet 6 or a nutritional supplement (unpublished data) for pain alleviation and mobility enhancement were included. Cats were recruited initially for these studies on the basis of owner-perceived reduced activity or impaired mobility. They were then screened as described in part I, but radiographs were only taken of the joints where pain was found on orthopedic evaluation. Cats with at least one appendicular joint where manipulation elicited an aversive response and whose radiographs showed the presence of DJD were included in these studies. Cats were stratified as ‘low impairment’ or ‘high impairment’ based on owner assessment of impairment. This method of designation has been described previously and used in previous studies to effect appropriate stratified randomization.6,33
Assessment of QoL
Prior to diagnostic evaluation or discussion of results, the owners of all cats were asked to complete a QoL assessment (modeled after Budke et al 14 ) for their cat (see Supplementary data, Appendix 2). This assessment was performed as follows: first, they were asked to write down the five activities that they believed were important for their cat’s quality of life/enjoyment of life. They were instructed that these ‘activities’ could represent any aspect of daily living. The items owners identified as important to their cat’s quality of life were designated ‘QoL items’. In order to determine the relative importance of each of these items, owners were then asked to rate the importance of each activity they listed, with the sum total importance score (IS) across the five activities being 100.
Characterization of QoL items
A total of 830 QoL items (495 from the population in part I and 335 from the population in part II) were reviewed and classified (see online supplementary material for full list of items). Items were designated into six behavioral domains: active transition (moving from a point A to point B); eating–drinking; grooming–scratching; playing–hunting; resting–observing and social activities. Once categorized into a behavioral domain, each QoL item was also defined as involving motion (active, AA) or not (inactive, IN). Items within the domains ‘active transition’ and ‘playing-hunting’ were considered AA; activities within the ‘resting–observing’ and ‘eating–drinking’ domains were considered IN; activities within the ‘grooming–scratching’ domain were classified as AA or IN depending on the description given by the owners. For some ‘inactive’ QoL items, the description of the activity implied motion to perform the item (eg, ‘sitting on the kitchen table watching the birds’). For these cases, a third category was established and these items were defined as inactive with implied-activity (IA), as decreased mobility might impair the cat’s ability to perform these items as well as the more classically ‘active’ items, such as play. Activities classified as IA were also classified as IN for analysis. Owners were contacted for clarification of the activity involved with some items and categorization was agreed upon by consensus of the authors.
Statistical analysis
Descriptive statistics were used to characterize the demographic composition of both groups of cats. We were interested in evaluating the distribution of the items owners listed; for each cat, the number of AA items, IN items and those with IA were summed, creating frequency scores (fs) for each cat for each subgroup of items — AA(fs), IN(fs) and IA(fs). IA(fs) was a subset of IN(fs). In order to evaluate the proportion of the items that involve mobility, we looked at both the items that were classified as active (AA), as well as the sum of the active and implied activity items (AA+IA), for the analysis. For each cat, importance scores were calculated for the AA, IA and AA+IA categories by the addition of the individual IS for each item within each category, resulting in AA(IS), IA(IS) and AA+IA(IS) values.
For the 99 cats in the prevalence group, descriptive statistics were used to describe the distribution of frequency scores for items in each domain [AA(fs), IN(fs), and IA(fs)] within the population The relationship between various demographic parameters (age, sex, weight, BCS, total DJD score and total pain scores) and the QoL item parameters (frequency and importance) was evaluated using ordinal regression analysis. Age, sex distribution, body weight, BCS, total pain and total DJD scores were compared between the normal and DJD groups using appropriate statistical tests. Frequency and importance scores were compared between the normal and DJD groups.
For the 67 cats in the study group, descriptive statistics were used to describe the distribution of frequency scores for items in each domain [AA(fs), IN(fs), and IA(fs)] within the population. Within this group of cats, frequency and importance scores were compared between the low-impairment (L) and high-impairment (H) cats.
Finally, frequency and importance scores for the QoL were compared between the cats in the prevalance group and those in the DJD study group.
Continuous and ordinal data were expressed as mean or median values, and ranges. Categorical data were expressed as frequencies. Differences were explored using Mann-Whitney U-tests for unpaired samples in both populations. Backwards stepping ordinal and linear regression analysis was used to investigate which variables significantly affected the distribution and importance of the QoL items for the cats in the prevalence group. In all tests, a P value of <0.05 was considered critical. Data analysis was performed using the statistical software GraphPad Prism (Prism 4.0 for windows; GradPad Software and JMP (JMP 9 for the Mac; SAS).
Results
Demographic data for both populations are shown in the Table 1.
Demographic data
Significant difference between normal and DJD groups within the prevalence population
Significant difference between the H and L groups within the DJD study population (see text for details)
Significant difference between the DJD (n = 22) and DJD study population (n = 67)
H = high impairment; L = low impairment
Part I: prevalence group
Demographic characteristics
The normal and DJD groups differed in age (P <0.0001), with the DJD group being significantly older. Additionally, the total pain and DJD scores for the normal and DJD groups were significantly different (P <0.001 and P <0.001, respectively).
QoL items
Four hundred and ninety-five items were reviewed in this population. The distribution of QoL items within the categories of AA, IN and IA, and the distribution of items with different levels of activity in each behavioral domain are shown in Table 2.
Distribution of QoL items based on levels of activity and distribution within the six behavioural domains
H = high impairment; L = low impairment
Frequency scores for the AA Qol items and IN QoL items were significantly different between the normal and DJD groups (P = 0.0022 and P = 0.0022, respectively) (Table 3). Owners of normal cats chose more AA activities and fewer IN items as contributing to their cats’ quality of life. Similarly, the importance score for AA was significantly different between the normal and DJD groups (P = 0.0027) (Table 3). Owners of DJD cats indicated that the overall importance of AA items was less than owners of normal cats. When looking at all activities where mobility was implied (AA+IA), the frequency and importance scores were lower in the DJD group compared with the normal group, though not statistically significant.
Distribution of AA, IN and IA QoL items, and the importance scores associated with AA and IA items in the 99 cats in the prevalence group, and in the normal and DJD groups. Data are expressed as median values (range)
Backwards stepping ordinal logistic regression revealed that the distribution of frequency scores for AA(fs) and IN(fs) was only significantly affected by age (P <0.0001 and P <0.0001, respectively), with AA(fs) decreasing with age and IN(fs) increasing with age. The distribution of IA(fs) was significantly affected by weight (P = 0.0012) with increasing weight being associated with decreased IA(fs). No other variables affected score distribution. Backwards stepping regression analysis revealed that for AA(IS) age was, again, the only variable that significantly affected the values (P <0.0001), with AA(IS) values decreasing with age. No parameters significantly affected the importance score for AA+IA(IS).
Part II: study group
Demographic characteristics
The H and L groups differed in age (P <0.01), (Table 1). Three hundred and thirty-five items were reviewed in this population. Distributions of the frequency of AA(fs), IN(fs) and IA(fs), and the distribution of categories of activity within each behavioral domain are shown in Table 2. The L and H groups within these 67 cats were not significantly different when comparing the frequency or importance scores for the different item categories (Table 4).
Distribution of AA, IN and IA QoL items, and the importance scores AA and IA items in the 67 cats in the DJD Study group, and in the L and H groups. Data are expressed as median values (range)
H = high impairment; L = low impairment
There was no difference between the frequency with which AA, IN or AA+IA items were chosen by owners of cats in the prevalence group (part I) or cats participating in DJD studies (part II) (Table 5). The total importance of items in the AA and AA+IA categories was not significantly different between these two groups of cats.
Comparisons between the 99 cats in the prevalence group and the 67 cats in the DJD study group, showing the distribution of AA, IN and IA QoL items, and the importance scores associated with AA and IA items. Data are expressed as median values (range)
Discussion
This study found that for a cohort of cats randomly selected from a single practice, the owners considered classically ‘active items’ to make up about 40% of the most important items for their cats’ quality of life and ‘inactive items’ to make up about 60%, rejecting our hypothesis that the items owners consider as important to their cat’s quality of life are predominantly activities requiring mobility. Even when accounting for inactive activities that would be hindered by impaired mobility (those that we defined as ‘implied activity’), active items still only made up half of the total items listed. This distribution appeared to be affected by age, with not only the frequency with which active items are selected by owners decreasing with age, but also the overall importance of active items decreasing with increasing age. Overall, owners of older cats appeared to consider active items as less important contributors to QoL than owners of younger cats did. It appears that owners of older cats accept, or expect, that cats become less active with advancing age, irrespective of the reason why. One reason for the decreased importance of activity might be the presence of DJD-associated pain; however, our data do not support this. This may have been a result of the overwhelming effects of age, and the strong association between age and the degree of DJD. 25
Although some interesting work has been performed in cats assessing QoL in association with various diseases,19–21,34 to our knowledge, no QoL studies in cats have been performed that allow owners to detail the activities in their cat’s life that they consider contribute to the cat’s QoL. Insight into what type of activities owners are considering when assessing QoL is important to understand. Owner-generated items were categorized into six domains and further delineated as AA or IN, with the inactive also being able to be further designated as ‘IA. Examples of AA listed by owners include running, playing with toys, etc, while IN included sleeping and being petted, among others. IA included those in which the owner specifically included in their description an activity that would have required action on the part of the cat. The full set of QoL items is available in the supplementary material. We made no restrictions on owners as to how they could describe items and, as such, some items were difficult to categorize. Category allocation of items was agreed upon by consensus of the authors, led by a Boarded behaviourist (MG) and owners were contacted for clarification when necessary. A limitation of the study is the possibility that some categorizations of IA may have been incorrect. For example, eating on the counter was categorized as an activity with implied activity; however, the cat could be picked up and placed on the counter in order to allow it to eat. In fact, many items, such as eating, require mobility for access, but these were included as IA, which may have resulted in an underestimation of the importance of mobility. One possible refinement for future work would be to ask owners if they felt mobility was important to their cat in performing a given activity. However, this could introduce undue bias into the study by focusing caretakers on mobility. Despite this limitation, it is interesting to note that of the items listed by owners as important to their cat’s quality of life, over 50% of the items were in the categories of social activities and rest/observe. These categories include items such as ‘sleeping on the bed with me’ and ‘laying in the sun’. The emphasis owners place on these activities for their cat’s QoL highlights the difficulty in establishing parameters for monitoring therapies designed to increase mobility and also highlights the importance of including the owner in discussions about therapeutic outcomes.
It is our belief that DJD-associated pain negatively impacts a cat’s quality of life and that the use of a quality of life assessment would be advantageous in monitoring patient progress and therapeutic interventions. However, as clinicians and clinical researchers, we may be overly focused on therapies designed to increase mobility, and it seems just as important to consider the effects of DJD-related pain on inactive behaviours when attempting to improve quality of life. Including non-active items in the assessment of DJD-associated pain moves us from the assessment of pure hypersensitivity states to the assessment of affective components of pain. Most of the measurement assays in pain research and clinical pain management are focused on the hypersensitivity sensory states associated with pain, 35 which would include indices such as lameness. However, these measures do not reflect spontaneous pain that has affective effects. QoL and HRQoL questionnaires that include attempts to measure affective aspects are being increasingly discussed in veterinary medicine.8,10–16,19–21 An item concerning mood was included in a study of QoL in diabetic cats 21 with over 20% of respondents saying that this issue was very important to them and their cat. Given the association between mood and pain, 36 inclusion of questions about this affective state may be even more sensitive when used to assess a disease thought to cause spontaneous discomfort, such as DJD.
We expected owners of cats recruited to ‘DJD studies’ to be more focused on mobility problems. This may well have been true, but as far as QoL appears to be concerned, the distribution of active items and inactive items was no different from cats recruited to a health screen study. Additionally, the importance scores associated with these items were not different between these groups of owners.
An important limitation of the present study is the generalizability of the population studied to the greater population of cats. The majority of cats in this study were neutered, indoor cats. In fact, the average percentage of time spent indoors across the cats in the prevalence group was 95%. Activities important to quality of life of outdoor cats and entire cats are unlikely to be represented in this study. However, the majority of the feline patient population in US veterinary clinics is neutered, 37 and the conclusion that non-active aspects of QoL are important to investigate is unlikely to be affected by neuter status. The majority of owners participating in this study were female and the effect of this on the selection of activities contributing to QoL is unknown. A recent study in humans found that in pediatric orthopedic patients, there was no relationship between the sex of the parent and agreement with patient-reported quality of life. 38 Finally, this study represented a cross-sectional view of QoL and it is likely, given the effects of age, that owner perceptions of items that make up QoL change over time. Longitudinal studies are needed to fully explore factors affecting this.
In conclusion, in our randomly selected cohort of cats, QoL items considered ‘active activities’ were chosen less frequently than inactive items. This distribution did not appear to be affected when owners are focused on DJD-related impairment of mobility. Our data suggest that understanding and assessing the non-active items that owners consider constitute QoL may be important to fully assess the impact of diseases such as DJD and therefore the success of therapeutic interventions. It will also be important for fully engaging owners in the treatment of pet cats with DJD.
Supplemental Material
Supplemental Material
Appendix 1 and Appendix 2, giving details of pain and quality of life scoring for owners
Footnotes
Acknowledgements
The authors would like to especially thank Drs B Kirsch (The Cat Hospital of Durham and Chapel Hill, Durham), A Alley (Bowman Animal Hospital and Cat Clinic, Raleigh), B Pondish (VCA Broad Street Animal Hospital, Fuquay Varina), D Holman (Gentle Care Animal Hospital, Raleigh) and K Bush (Kate Bush’s Mobile Veterinary Practice, Raleigh) in North Carolina for their referral of cases for these studies.
Funding
The data described here was gathered during three separate studies, funded by Novartis Animal Health, Royal Canin and Nutramax. The lead author was funded by the Morris Animal Foundation and the Global Fellowship Program, Novartis Animal Health. The collection of this data was not specifically requested by any of the sponsors and none of these sponsors had any role in the design, collection, analysis or interpretation of the data described in this study.
Conflict of interest
BDXL has received honoraria for sponsored CE from all of the sponsors of the work described in this study, and has acted as a consultant for Novartis Animal Health and Nutramax.
Supplementary data
Appendix 1 and Appendix 2, giving details of pain and quality of life scoring for owners, are available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
