The practice of veterinary medicine has changed considerably over the past 20 years. A major determinant of success and professional satisfaction in veterinary practice is the ability to address the needs of patients and their owners with empathy and compassion. Social concerns about animal well-being have focused attention and research on the concept of QOL (ie, having adequate food, water, and shelter and being free from signs of pain or distress).1–3 The terms QOL, welfare, and well-being are used interchangeably in veterinary medicine. It is very challenging to define QOL for animals because we can only infer psychological states in our patients. Thus, it is more difficult to assess the impact of health problems, environmental conditions, husbandry factors, nutritional adequacy, and other changes on a patient's QOL. Freedom from hunger and thirst, freedom from physical and thermal discomfort, freedom from pain and injury, freedom from fear and distress, and freedom to express normal behavior (ie, the so-called 5 freedoms3) are considered essential to define a good QOL for animals.4
McMillan5 states that unpleasant experiences including fear, anxiety, boredom, loneliness, separation anxiety, grief, depression, frustration, anger, hypoxia, pain, thirst, hunger, cough, dizziness, full bladder, constipation, nausea, and pruritus should be considered to negatively impact animal QOL. Pleasant experiences are defined as joy, play, social companionship, mental stimulation, physical contact, gustatory sensation, nurturing young, and sexual activity.5 The balance between these pleasant and unpleasant aspects determine the QOL of the animal.5
Increasingly, owners of animals afflicted with life-threatening illness opt for treatments that offer the potential for prolonged survival time. In canine cancer patients that are treated with chemotherapy, associated transient adverse effects might be considered a reasonable trade-off for increased survival duration. Thus, an important factor in selecting a treatment is to consider the QOL of the animal during the treatment period.6 Owners may excessively anthropomorphize or, conversely, may deny anthropomorphic considerations when making decisions about treatment.7–9 Some owners may elect to prolong their pet's life, regardless of QOL consideration. Veterinarians may be called upon to assume the role of animal advocate and attempt to interpret what might be important from the perspective of the animal.10,11 It is also crucial that veterinarians establish relationships with owners who facilitate communication and establish agreement on QOL goals.12–16
The medical treatment of dogs with disseminated cancer is largely palliative. Making decisions regarding the type and duration of cancer treatment often requires that the clinician and owner assess patient QOL. This assessment is typically based on inference and assumption, which may or may not be a true reflection of the patient's status. Thus, it is important to have an efficient, objective, and accurate tool to assess QOL of the companion animal. Comparatively little research has been conducted to evaluate QOL in companion animals treated for specific disease entities. Several groups have attempted to include QOL measures,5,17–21 but no comprehensive and objective tool exists for evaluating QOL in a clinical veterinary oncology setting. The purpose of the study reported here was to address the need for a reliable, objective, and easy-to-use QOL instrument in veterinary medicine. A major goal was to develop and validate an instrument that would not only identify the QOL changes of canine cancer patients as perceived by the owner, but also be appropriate for use in a longitudinal evaluative process. The present study was designed to be a longitudinal evaluation of QOL in a small animal oncology setting. Patients were evaluated both by owners and an attending clinician at the time of their first visit to the clinic and then at 3 and 6 weeks after the initiation of chemotherapy. It is challenging or even impossible to determine the QOL of another person or animal, despite the best efforts of the proxy. Thus, in an effort to identify whether the owner's QOL might impact their perception of their pet's QOL, as is the case in human pediatric proxy assessment, we included queries about the owner's QOL over the same time period.
Caregiver effect in QOL assessment is recognized in pediatric oncology to reflect the emotional and physical impact of the child's disease on the parent's perception of the child's QOL.22,23 We assessed a caregiver effect in this study to determine the impact that terminal disease in the pet has on the owner's QOL, decision making, and perception of the animal's well-being. The present study was also designed to meet the need for a QOL instrument to enhance owner-clinician communication to facilitate owner decision making and support the human-animal bond in a challenging medical setting.
Materials and Methods
Literature describing human pediatric and oncology proxy questionnaires as well as literature detailing veterinary studies on animal QOL were reviewed to develop a survey instrument that featured 6 key components. These components were the canine patient's QOL 6 months before cancer diagnosis and during and after treatment, the caregiver effect and the owner's QOL during the dog's chemotherapeutic treatment, impact of caregiver effect on decision making, clinician's assessment of owner's QOL, and the compatibility between the owner and clinician's proxy QOL assessments. First, owners were queried about the canine cancer patient's QOL at a point in time when the animal was healthy and free of disease. Each subject thus served as its own control before and after diagnosis of cancer and during the initial phase of chemotherapy. Questions were subsequently addressed to assess the canine cancer patient QOL since the manifestation of neoplastic disease. Further questions were posed regarding the canine cancer patient's QOL with regard to response to treatment. Canine owner QOL and priorities in decision making at the time of diagnosis and during the initial phase of chemotherapy were investigated through directed questions. The attending clinician's impressions of the owner's (proxy) QOL and the owner's ability to prioritize decisions for the canine patient were recorded. Finally, the attending clinician was solicited to provide impressions about the canine cancer patient's QOL.
The survey was designed to be administered to owners of canine patients admitted to the Michigan State University Animal Cancer Care Clinic for evaluation and treatment of metastatic neoplasia.a The project was initially reviewed and approved by the Michigan State University Institutional Review Board human research protection program to assure compliance with applicable federal regulations and university policies. Accordingly, study objectives and methods were presented both in writing and by oral explanation to owners and clinicians at the initial visit. All human study participants then signed a consent form.
Dogs were enrolled in the study on the basis of the sole inclusion criteria of diagnosis of disseminated metastatic cancer requiring chemotherapy. Dogs and their owners were enrolled in the study at the initial referral visit. Canine patients with life-limiting nonneoplastic diseases or with acute leukemia were excluded on the basis of short anticipated survival time. Dogs that underwent prior chemotherapy for any reason were excluded.
Relevant clinical data regarding the patient's physical and laboratory findings along with responses to treatment and potential adverse effects of the treatment were obtained from each patient's medical record at each visit. These data were subsequently correlated with the results of the QOL survey to identify any significant predictors of the dog's QOL.
Two survey instruments were used.a One survey form was completed by the owner, and the other was designed to be completed by the attending veterinary clinician. Additionally, relevant information was compiled from the medical record. After the initial visit survey, the questionnaire had only 4 parts, omitting the retrospective assessment of QOL before cancer diagnosis.
The second evaluation occurred 3 weeks after the initiation of chemotherapy. During this visit, another survey form, slightly modified from the first by omitting the original 6 months of retrospective history, was distributed to assess changes that had occurred since the start of treatment. Surveys were again distributed to both the owner and the attending clinician, and the patient's medical record information was integrated into the study documents.
During the third evaluation 6 weeks after the initial visit, the patient was reevaluated by the attending clinician to determine response to treatment. On the basis of this reevaluation, changes in treatment plan were implemented. A survey form identical to that used during the second evaluation was distributed to both the owner and the attending clinician to assess changes in the interval between the 3- and 6-week evaluations. The patient's medical record information was again updated.
The human study participants were given the opportunity to respond electronically or in a written paper format for all 3 visits. To facilitate online participation, computer terminals were provided for client use. The medical record information for each patient was included in each patient's study data sheet. Chemotherapy adverse effects were interpreted on the basis of the Veterinary Cooperative Oncology Group criteria.24 A 5-point Likert-type scale was used to grade chemotherapy-associated adverse effects.25 Data obtained through the 1–5 Likert scale used for the QOL instrument were assigned as 1 being the best measure, 3 being a neutral measure, and 5 being the worst measure of outcome.
To evaluate for a correlation between the QOL of the dogs as assessed by the dog owners and by the attending clinicians, the data were ranked in order of size and calculations were based on the ranks of corresponding values. The Spearman P value was calculated. For all comparisons of QOL parameters between sequential visits to the clinic, the Wilcoxon test for paired samples was used. Multiple regression analysis was used to analyze the relationship between the dog's QOL as perceived by the owner through the variables analyzed in the questionnaire. Multiple regression analysis was performed for each visit, with the variables entered in a forward manner. The following comparisons were made with the Wilcoxon signed rank test: comparison between QOL prior to diagnosis and at the time of diagnosis, comparison between QOL prior to diagnosis and at the time of the third chemotherapy treatment, comparison between QOL at the time of cancer diagnosis and at the time of the second chemotherapy treatment, and comparison between QOL at the time of cancer diagnosis and at the time of the third chemotherapy treatment. For these comparisons, the following data were assessed: the dog's QOL, signs of anxiety, appetite, happiness, mobility, play activity, perceived pain level, activity level, whether the dog was “bothered by cancer,” whether the dog had clinical signs of illness, perceived enjoyment of the dog's favorite activity, and level of owner's worry about their dog's health issues.
All reported P values are 2-sided. Values of P < 0.05 were considered significant. Statistical analyses were performed with standard software.b
Results
Surveys were distributed to all dog owners visiting the Michigan State University Center for Comparative Oncology Animal Cancer Care Clinic between March 2008 and August 2008. During their initial visit, 81 clients completed questionnaires. During their second visit, 36 clients completed questionnaires. Prior to the second visit, patients were lost to follow-up (n = 29), clients declined chemotherapy (8), chemotherapy was not recommended as a treatment option (4), patients were euthanatized (3), or cancer was not diagnosed (1). Thirty-one clients completed questionnaires during their third visit. Between the second and third visits, 2 clients did not complete the questionnaire in a timely manner and completed both the second and third questionnaires at the same time. These clients’ retrospectively recollected observations were not included in the statistical analysis because of the potential for incorrect data or bias. One client moved out of state, 1 patient was euthanatized, and 1 owner did not submit the questionnaire.
Completed questionnaires were obtained from the assigned clinicians during each visit. Twenty-one questionnaires were collected at the time of treatment initiation, 19 were collected 3 weeks later, and 16 were obtained during the sixth week of treatment.
Two medical records could not be located until 6 months after the end of the study, and the questionnaires corresponding to these patients were excluded from the statistical analysis, resulting in 29 dogs ultimately evaluated. The median age of these patients was 8.3 years (range, 3 to 15 years). The patient population included 14 males and 15 females. The following breeds were represented: mixed (n = 7), Golden Retriever (6), Boxer (3), pitbull-type dog (3), German Shepherd Dog (2), Labrador Retriever (1), Flat-Coated Retriever (1), German Shorthaired Pointer (1), Basset Hound (1), Bernese Mountain Dog (1), Brittany (1), Standard Schnauzer (1), and Yorkshire Terrier (1).
Diagnoses included lymphoma (n = 17), osteosarcoma (5), hemangiosarcoma (2), mast cell tumor (1), thymoma (1), mammary gland carcinoma (1), pancreatic carcinoma (1), and histiocytic sarcoma (1). At initial diagnosis and throughout the first 6 weeks of treatment, metastases were noted within the thorax (n = 6) and abdomen (12) and affected the regional lymph nodes (2), skin (1), and bone (1). Two dogs with lymphoma were also being treated for comorbid conditions, including diabetes mellitus and lifelong seizure activity.
Treatment protocols were not consistent across the patient population because the diagnosis differed among patients. Chemotherapeutic drugs used included carboplatin, doxorubicin, vincristine, cyclophosphamide, L-asparaginase, chlorambucil, mustargen, pamidronate, lomustine, and gemcitabine. Of the dogs that were treated with chemotherapy, 6 dogs underwent surgery before initiating chemotherapy and 2 dogs were concurrently treated with radiation therapy. Of the 29 dogs evaluated, 7 dogs were treated in an adjuvant setting and 22 dogs were evaluated for response to chemotherapeutic treatment. Of these 22 dogs, 6 dogs had progressive disease, 12 dogs achieved complete clinical remission, and 4 dogs achieved partial remission. Thus, 72.7% responded to treatment, and 27.2% failed to respond.
On the basis of owner survey responses for 29 dogs, 23 (79%) dogs were acquired as puppies. Overall, 27 (93%) owners described the main role of the dog in the household as a family member, and 17 (59%) owners indicated that the main role was as a companion. Owners were allowed to respond in multiple categories when describing the role of the animal in the family. Additionally, 4 (14%), dogs were classified as guard dogs, 4 (14%) as hunting dogs, and 1 (3%) as a show dog. No dogs were reported to be working dogs.
Owners described their dogs’ preferred activities as playing for 25 (86%), interacting with humans for 25 (86%), eating for 24 (83%), walking for 23 (79%), other exercise for 22 (76%), riding in the car for 19 (66%), and staying at home for 18 (62%) dogs. Sixteen (55%) of the dogs were reported to enjoy all of these activities, and 8 (28%) were reported to enjoy other activities that were not provided as options, as designated by another response option. Owners were allowed to respond in multiple categories.
Our results indicated that 20 (69%) owners selected the multiple-choice option “extremely worried” when queried about their dogs’ cancer diagnosis at initial evaluation. Owners reported a significantly (P < 0.001) decreased level of worry during the sixth week, compared with the time of initial diagnosis. Owners responded to the following multiple-choice options when queried about the most prominent cause for concern: 21 (72%) reported that their worry arose from the incurable nature of the disease, 19 (66%) from fear of decrease in the dog's QOL, and 12 (41%) from concern regarding the potential chemotherapy adverse effects.
Sixteen (55%) owners evaluated our QOL questionnaire as being extremely easy and 12 (41%) as being very easy to complete. Of the 29 owners who completed the questionnaire, 23 (79%) indicated that they enjoyed the opportunity to evaluate their dog's QOL, 5 (17%) owners enjoyed this opportunity moderately, and 1 (3%) enjoyed this opportunity somewhat. These results were representative for the questionnaires obtained during all 3 consecutive visits, and the question regarding enjoying this opportunity to evaluate their dog's QOL was repeated at the end of the questionnaires of the first, second, and third visits.
Comparisons between QOL prior to diagnosis and at the time of diagnosis—Owners were asked to select discrete variables in response to specific questions. Owners selected the multiple-choice option that their dog had better QOL 6 months prior to cancer diagnosis than at the time of enrollment in the study (P = 0.007). Similarly, at 6 months prior to diagnosis, dogs were specifically reported to be more active (P < 0.001), happier (P = 0.001), and not bothered by their disease (P < 0.001); to have fewer signs of anxiety (P = 0.002); and to have a better appetite (P < 0.001). Dogs had better mobility (P < 0.001), elicicted fewer signs of pain (P < 0.001), were more playful (P < 0.001), had fewer clinical signs associated with illness (P < 0.001), and had a better QOL (P < 0.001) according to their owners at the prediagnosis assessment.
Comparisons between QOL prior to diagnosis and at the time of the third chemotherapy treatment—Compared with behaviors noted 6 weeks after initiation of chemotherapy, owners specifically selected multiple-choice options which reported that 6 months before cancer diagnosis, dogs had fewer signs of anxiety (P < 0.001), had a better appetite (P < 0.001), appeared happier (P = 0.007), did not appear to be bothered by cancer (P = 0.007), did not have clinical signs of illnesses (P = 0.015), had better mobility (P < 0.001), were more playful (P < 0.001), and had a better QOL (P < 0.001).
Comparisons between QOL at the time of cancer diagnosis and at the time of the second chemotherapy treatment—Owners selected responses that indicated they worried less about their dogs’ health (P < 0.001) and found completing the questionnaire easier (P = 0.003) during the second visit, compared with during enrollment into the study. Improvements were noted in the following areas during the second visit, compared with the time of diagnosis: activity level (P = 0.015), clinical signs (P = 0.007), mobility (P < 0.001), perceived pain level (P < 0.001), play activity (P = 0.015), and appetite (P = 0.002). Owners also reported that their dogs appeared less bothered by the neoplastic disease (P = 0.003). In addition, owners responded that their own activities were less limited at the second visit, compared with the first visit (P = 0.001; caregiver effect).
Comparisons between QOL at the time of cancer diagnosis and at the time of the third chemotherapy treatment—When comparing QOL parameters noted at the third chemotherapy treatment with those documented at the time of enrollment in the study, owners selected discrete response options that described improvements in the following areas: activity level (P = 0.013), motility (P = 0.031), play activity (P = 0.013), appetite (P = 0.017), perceived enjoyment of favorite activity (P = 0.002), and perceived pain level (P < 0.001). Dogs were reported to be less bothered by the clinical signs associated with cancer (P > 0.001), vocalize less (P = 0.031), and have fewer clinical signs of illness (P = 0.003). During the third visit, owners reported that the questionnaire was easier to complete (P < 0.001) and that they worried less about their dogs’ health issues, compared with the time of enrollment (P < 0.001).
Multiple regression analysis performed on data derived from the survey administered during the initial visit indicated significant predictors of the owner's perception of QOL to be whether the dog appeared ill (standardized coefficient β, 0.41; P < 0.001) and whether the dog was playful (standardized coefficient β, 0.46; P < 0.001). The regression model was significant (P < 0.001), with the significant predictors explaining 85.21% of the variability in the dog's QOL.
Multiple regression analysis of the dogs’ QOL as perceived by owners during the second visit indicated the following significant predictors: the happiness of the dog (standardized coefficient β, 0.39; P = 0.016) and the dog's playfulness (standardized coefficient β, 0.61; P < 0.001). The regression model was significant (P < 0.001), with the significant predictors explaining 82.56% of the variability in the dog's QOL.
Multiple regression analysis of the questionnaire results obtained during the third visit indicated significant predictors of owner's perception of QOL to be the illness of the dog (standardized coefficient β, 0.33; P = 0.014) and the dog's playfulness (standardized coefficient β, 0.44; P < 0.005). The regression model was significant (P < 0.001) with the significant predictors explaining 55.53% of the variability in the dog's QOL.
For all 3 visits, significant correlations were found between the responses of owners and clinicians to questions regarding QOL. The greatest agreement was noticed during the third visit (β = 0.8133; P < 0.001) and the first visit (β = 0.807; P < 0.001). During the second visit, it was noticed that there was still significant (β = 0.562; P = 0.017) agreement, but the percentage of agreement was 56%.
Discussion
In the present study of owners of 29 dogs undergoing chemotherapy for various types of cancer, we identified 3 significant predictors of QOL: changes in play behaviors, clinical signs of disease, and canine happiness as perceived by the owner. Furthermore, the QOL instrument we developed for the present survey was easy to use and enhanced client perception of patient care and clinician concern. The significant predictors of QOL that were identified may guide treatment decisions and assessment of canine patients undergoing chemotherapy.
In this study, we used a QOL instrument that we developed to assess chemotherapy-treated canine cancer patients with the intent to correlate owner and clinician perception of the dog's QOL with the impact of caregiver affect experienced by the owner. Thus, this represented the first attempt in veterinary medicine to address the concern for caregiver negative affect in QOL assessment. A significant finding of the present study was that there was no correlation between the owner's level of stress and his or her assessment of the dog's QOL parameters. According to our study results, there were 3 significant predictors of owner proxy QOL assessment of chemotherapy-treated canine cancer patients (changes in play behaviors, clinical signs of disease, and canine happiness as perceived by the owner). Another unique aspect of this study was that it used the same dog while healthy as an internal control, rather than dogs from a healthy-matched population as controls. Continued fluid QOL proxy assessment of the canine cancer patient was considered to be very important because such ongoing assessment could help owners realistically evaluate the impact that the specific disease process and correlated treatment had on their pet's QOL. The survey QOL tool created for this endeavor identified that canine cancer patients treated with cytotoxic chemotherapy had a better QOL when healthy and free of neoplasia, compared with their QOL at the initial visit or at the time of the second and third hospital visits. Subjects had a better QOL at 3 and 6 weeks after initiation of chemotherapy treatment than they did at the time of cancer diagnosis (first visit). Thus, treatment appeared to improve QOL but did not restore the original QOL level 6 months prior to the diagnosis of cancer.
We covered 3 main functional domains (physical, psychological, and social functioning) in the questionnaire developed for this study, as has been done in most human and animal QOL questionnaires.21,26–39 Physical functioning assesses signs associated with disease, treatment, and ability to function in daily activities. Psychological functioning assesses psychological distress and a positive sense of well-being.40
Social functioning assesses aspects of social relationships and interactions.21,41 Psychological functioning was addressed by the owner's response to questions about their dog's happiness. Social functioning in our study was identified as human-animal interactions and play behaviors.
A critical aspect of this QOL instrument was the use of each animal as its own internal control. Comparing the animal's baseline QOL, as assessed by the owner 6 months before cancer diagnosis, with its QOL at the time of diagnosis and throughout treatment might ultimately help identify potentially beneficial treatment modifications. Thus, by focusing on what is considered normal for each individual animal, it may be possible to determine whether a specific treatment protocol affects the animal's QOL positively or negatively, compared with the proxy baseline assessment.42 For example, if an animal has a life history of being described by the owner as a picky eater, we would not automatically attribute slight anorexia to chemotherapeutic treatment. This supposition suggests that compromised QOL noted at cancer diagnosis and during treatment in an animal previously described as healthy is induced by the disease process and treatment rather than extraneous causes. Accordingly, any improvement in QOL attained throughout the treatment course may be attributable to the positive impact of the therapeutic interventions. To our knowledge, this QOL instrument is the first in canine cancer medicine to use the animal's prediagnosis status as an internal control.
Overall43 asserts that fear is a critical aspect to be evaluated in animals undergoing treatment for neoplasia because fear can compromise an animal's QOL. Animals cannot predict the beneficial outcome of treatment, and they may perceive only a threat, especially during painful procedures such as injections or catheter placement.43 Our survey instrument included domains addressing fear and pain in the animal subjects (ie, “Your dog experiences pain/discomfort now: a. Never, b. Seldom, c. Sometimes, d. Often, e. Always” and “How often does your dog experience signs of anxiety and fear [shivering, increased salivation, panting, whimpering, howling, barking, and growling] during your visits to MSU VTH? a. Never, b. Seldom, c. Sometimes, d. Often, e. Always”). Although our survey addressed these issues, pain and fear were not identified as significant predictors of QOL. Signs of illness might include pain and fear, but overlap in these categories may have obscured their relevance to QOL in our study. This aspect of QOL research may become more important over time because increased survival rates associated with advanced cancer treatments in veterinary medicine may increase the incidence of chronic tumor-induced pain.21,44,45
There are many similarities between the proxy assessment of pediatric patients, who are unable to self-report, and QOL assessment of small animal patients. Our survey instrument was constructed to provide for proxy assessment by the most prominent caregiver serially throughout the duration of the study. In pediatric medicine,46 QOL might be greatly affected by the physical and psychological sequelae experienced by both cancer survivors and their parents.47 Caring for a family dog with neoplastic or terminal disease can also be highly distressing for owners, contributing to caregiver negative affect in the survey response.
We attempted to assess caregiver negative affect by inclusion of specific owner QOL questions. We wanted to determine the extent to which the owner's report had been biased by the impact of the dog's disease on the owner's lifestyle and QOL. Specifically, the owner was asked how much worry or anxiety he or she was experiencing because of different variables such as chemotherapy and adverse effects, the dog's QOL, the incurable nature of the disease, financial issues, perception of others, and time and scheduling concerns. Also, the owner was asked about the negative lifestyle impact that the animal's condition was inducing to his or her own QOL. These identical questions were asked in the first, second, and third survey (visit) to identify potential serial changes. We also created validating questions to specifically solicit information about the impact of the pet's health status on the owner's QOL from the perspective of the attending veterinarian. To our knowledge, this is the first survey in veterinary medicine to address the aspect of caregiver negative affect directly and indirectly through veterinary secondary proxy assessment.
In the case of proxy decision-making for ill children and for pets, parents as well as owners are the ultimate arbitrators of changes in treatment modalities. However, a relatively smaller contingent of treatment approaches is offered to companion animal owners because of practical limitations in the level of cancer care available for the treatment of companion animals. For example, potentially curative treatment such as stem cell transplantation may be offered to pediatric cancer patients, and the parent must weigh the risk of treatment failure, complication, decrement in the child's short- and long-term QOL, and potentially even the child's death due to transplant complications against the potential for life-saving curative treatment. Stem cell transplantation treatment is in its infancy in veterinary oncology, and curative treatment for most animal cancer patients that require chemotherapy treatment is virtually nonexistent. Thus, the level of owner stress in decision making may not approach the level of stress engendered in the parent of a critically ill child. It would be helpful in future studies to have representatives from the field of human psychology to assist in clarifying issues such as caregiver negative effect.
Another difference between human and companion animal medicine is the potential therapeutic decision of euthanasia for a pet, which adds considerable stress to the companion animal owner who is not present for the parent of a cancer-stricken child. Given both the similarities and differences between companion animal and pediatric oncology concerns, it is of interest that negative caregiver bias was not identified as a significant factor in QOL assessment for companion animals in our study.
Veterinarians always have to work through a financially responsible third party that is able to make fiscal and medical decisions for the animal. This situation again is quite similar to the areas of pediatrics, gerontology, and psychiatry in human medicine. Owners must weigh the cost of treatment against the potential benefits for and risks to their companion animal, whereas in human medicine in the United States, availability of third-party payment may render cost of treatment a less significant concern for parents in many cases.
Our findings suggested that both owners and clinicians were able to objectively evaluate the QOL of the animal and proceed with appropriate decisions regarding the individual animal's case management. There is always the possibility that a subjective component, a placebo effect, or even influences of anthropomorphism exist in the assessments performed by both the clinician and the owner in inferring the emotional and psychological state of an animal. If that were the case in our study, we would expect to have a disagreement between the owner and the clinician because we would not expect both individuals (owner and clinician) to make identical observations. Also, clinicians are individuals trained to make objective assessments. If veterinarians were affected by excessive anthropomorphism or subjective projection of feeling states in assessing their patients, their ability to make correct clinical decisions and appropriate treatment plans for their patients would be impaired. In this case, clinicians and owners did not make identical observations, but the questionnaire seemed to be helpful in clarifying the objective assessments of both parties and was able to assist in identifying a common ground for treatment decisions.
General considerations for errors in owner reporting in our study include the fact that, although painful and debilitating conditions such as neoplasia could affect the behavior of dogs, owners could interpret such behavioral changes in overly optimistic or pessimistic ways. Some owners might react under stress and misinterpret their animal's behavior after cancer diagnosis as indicating excessive animal suffering. Other owners might deny negative aspects of the animal's behavior and be inappropriately optimistic considering the fact that the animal is under treatment in the absence of appreciable tumor response (placebo effect).
Continued fluid QOL proxy assessment of the canine cancer patient was also considered to be very important because such ongoing assessment would help the owners understand the impact the specific disease process was having on their pet's QOL.6 When the owners were aware of these changes, we considered it less likely that they would euthanatize prematurely because of the misinterpretation of a stressful event or that they would inappropriately postpone the decision for euthanasia to avoid the sadness they anticipate upon the pet's death. This is a very critical issue because both the assigned veterinarian and the owner may share responsibilities in decision making, but the owner will make the ultimate decision. Valid and accurate QOL assessment may be invaluable in clarifying the owner's decisions regarding necessary patient care. Further, longitudinal QOL assessment may not only help to facilitate decision making for treatment changes and appropriate end-of-life decisions, but may also support owners during the difficult grief period.6
The decreased QOL observed during treatment, compared with the time of the dog's baseline state of health, might be attributable to the relatively brief duration of observation after initiation of treatment. Longer duration of follow-up might be required to identify a complete restoration of the animal's QOL level. Alternately, it is possible that treatment with chemotherapy was responsible for the decrement in QOL from baseline, and evaluation after conclusion of chemotherapy might be better able to assess a return to normal baseline QOL. Another study with a more protracted evaluation period would help answer this question.
Another possible reason for decreased QOL while the dog was undergoing chemotherapy treatment might be that the animal's QOL improved as a result of the treatment, but given that the animal was not expected to be completely cured from the neoplastic disease, it might be perceived to have lower QOL than while healthy. A possible explanation would be that the cancer impact was not completely controlled by the treatment modalities used, resulting in improved but not optimal QOL. The animal might have improved QOL because the treatment proved effective, but the animal might still have a degree of compromised QOL attributable to the chemotherapy's adverse effects. Also, owners may subconsciously react to the lack of curative treatment for their pet and thus may interpret behaviors and physical signs in a more negative way than if the diagnosis was of a curable disease rather than cancer.
A question that was revealed to be a significant predictor of the dog's QOL addresses the dog's playfulness and activity level. This is an easily identifiable change, in that a dog that feels physically well is more likely to be playful and to behave as it did at baseline before the cancer diagnosis. Most owners and clinicians would assume that a dog that experiences pain and depression would be unlikely to be highly active or to participate in play behaviors. Play behavior and activity level are observable and quantifiable, making them more reliable indicators of the animal's QOL. The significant association of play behaviors with QOL reinforces the notion that QOL questions should aim to address changes that the proxy can identify, compare, and quantify. The findings of the present study seem to agree with those of Boissy et al,48 who argued that play behaviors appear to be one of the “most promising and convenient indicators” of good QOL in laboratory and farm animals. The results of our study support the view that play behavior is a good QOL indicator for canine cancer patients as well.
The 3 domains identified by multivariate analysis as significant QOL predictors (regarding illness, playfulness, and happiness) should be included in future studies of QOL for canine cancer patients as perceived by the owners.
One question in our study regarding barking and vocalization behaviors was found to be confusing to owners. Owners appeared uncertain as to how to correlate the level of vocalization and barking with the animal's QOL. Barking as an expression of excitement and happiness might reflect normal behavior for the dog, whereas other dogs might bark or vocalize as a sign of distress. In future efforts, the owner's perceived reason for vocalization should be factored into any question that attempts to solicit information about this aspect of canine behavior.
Despite the fact that the clinician's questionnaire was very brief, we did not have 100% study compliance. This occurred despite the fact that all clinicians signed consent forms and agreed to participate in the study from the outset. Compliance with any survey-based study is always challenging, and 100% participation is never expected. However, a robust reminder system and making the survey easy to complete would be expected to improve clinician compliance. Clinicians are busy, and other priorities relative to patient care make survey completion a low priority for them. However, thorough explanation of the importance of achieving the study goals might help to improve clinician participation and hence improve compliance with questionnaire completion.
In contrast, we had 29 (100%) animal owners who agreed to participate in the study. These owners expressed that they enjoyed the opportunity to participate and appreciated the opportunity to evaluate their dog's QOL. They reported that completing the questionnaire was easy. In our study, 19 (66%) responding owners were extremely concerned about their dog's QOL while pursuing chemotherapeutic treatment. Allowing them an opportunity to complete a QOL study might have been perceived as evidence of the clinic acknowledging and being supportive of those concerns. A well-designed QOL tool can facilitate owner's emotional clarity and might even help with closure for owners during the grief period after the loss of a pet.
It is of great interest that even though only two-thirds of dogs included in this study ultimately responded to their treatment and one-third of the dogs failed to respond, the owners and clinicians perceived the QOL of the dogs that received chemotherapy to be overall improved. This finding suggests that either the dogs benefited from the chemotherapeutic treatment in some manner not associated with clinical remission, or that the signs of the neoplastic disease were improved even in patients that did not achieve complete remission. It is possible that both clinicians and owners projected a placebo effect into the animal's response, and they subjectively misperceived that the treatment benefited the animal.
The QOL questionnaire developed for this study had several innovative aspects. To our knowledge, this is the first QOL instrument developed in the field of veterinary medicine to assess, through longitudinal evaluation, the QOL of canine cancer patients treated with chemotherapy as perceived by both the owner and the attending clinician. A major finding of our study was that both owners and clinicians had significant agreement regarding the QOL of each individual canine cancer patient when assessed over 3 consecutive visits. In future studies, we would recommend inclusion of the 3 significant predictors of QOL as identified in this study. We believe this approach would prove beneficial in assessing QOL in chronic medical conditions such as diabetes or thyroid conditions, particularly through extension through a longer duration of study.
The present study was a pilot project and thus had several limitations. First, the time constraints of conducting the study necessitated a small time window for accrual; therefore, we had a small sample size. We achieved an adequate sample size as projected originally in our study design, but the diseases represented were relatively heterogeneous. A larger sample size would allow more comprehensive understanding of the proxy assessment of canine cancer patients treated with chemotherapy. With a small sample size, caution in interpretation of results is necessary because the findings might not completely apply to the general population of dogs treated with chemotherapy. Another limitation of this study is that we did not include age- and breed-matched control dogs, but rather we used a retrospective measure of the same animal as its own internal control. In this case, we cannot be sure of the veracity of the QOL information given retrospectively by owners, who might misrepresent the animal's actual QOL in the past because of owner bias.
In the future, larger prospectively designed case control studies, with longer duration of follow-up, might provide additional valuable insight into the QOL impact of chemotherapy on dogs with selected cancer types. It is possible that the QOL impact of cancer treatment would differ between dogs with diseases as disparate in clinical manifestation as lymphoma and hemangiosarcoma and that any potential impact would be additionally affected by disease stage. We believe that availability of QOL data would further verify the beneficial aspects of cancer treatment for dogs. These data could be used to support the veterinarian's discussion of cancer care with animal owners.
ABBREVIATION
| QOL | Quality of life |
Copies of the survey instrument are available from the corresponding author upon request.
Medcalc, version 12.3.0, MedCalc Software, Mariakerke, Belgium.
References
1. Farm Animal Welfare Council. Advice to ministers on the need for further controls over the mutilation of farm animals. Br Vet J 1982; 138: 7–10.
2. The Council of the European Union at Europa EUR-Lex website. Council Directive 98/58/EC of 20 July 1998 concerning the protection of animals kept for farming purposes. Available at: eur-lex.europa.eu/smartapi/cgi/sga_doc?smartapi!celexapi!prod!CELEXnumdoc&lg=EN&numdoc=31998L0058&model=guichett. Accessed Dec 26, 2012.
3. European Union website. European commission. Health and consumers: animal welfare on the farm. Available at: ec.europa.eu/food/animal/welfare/farm/index_en.htm. Accessed Dec 25, 2012.
4. New council for animal welfare. Vet Rec 1979; 104: 271–272.
5. McMillan FD. Quality of life in animals. J Am Vet Med Assoc 2000; 216: 1904–1910.
6. Villalobos A. Caring for geriatric cancer patients: what is needed aside from medical care. In: Kaplan L, ed. Canine and feline geriatric oncology: honoring the human-animal bond. Hoboken, NJ: Wiley-Blackwell, 2007; 23–40.
7. Lee BH. Managing pain in human neonates—applications for animals. J Am Vet Med Assoc 2002; 221: 233–237.
8. Cornell KK, Kopcha M. Client-veterinarian communication: skills for client centered dialogue and shared decision making. Vet Clin North Am Small Anim Pract 2007; 37: 37–47.
9. Shaw JR, Bonnett BN, Adams CL, et al. Veterinarian-client-patient communication patterns used during clinical appointments in companion animal practice. J Am Vet Med Assoc 2006; 228: 714–721.
10. Hewson CJ. How might veterinarians do more for animal welfare? Can Vet J 2003; 44: 1000–1004.
11. Akashi A. Ten tips for veterinarians dealing with terminally ill patients. Vet Clin North Am Small Anim Pract 2011; 41: 647–649.
12. Rollin BE. The regulation of animal research and the emergence of animal ethics: a conceptual history. Theor Med Bioeth 2006; 27: 285–304.
13. Adams CL, Frankel RM. It may be a dog's life but the relationship with her owners is also key to her health and well being: communication in veterinary medicine. Vet Clin North Am Small Anim Pract 2007; 37: 1–17.
14. Shanan A. A veterinarian's role in helping pet owners with decision making. Vet Clin North Am Small Anim Pract 2011; 41: 635–646.
15. Shaw JR, Lagoni L. End-of-life communication in veterinary medicine: delivering bad news and euthanasia decision making. Vet Clin North Am Small Anim Pract 2007; 37: 95–108.
16. Chun R, Garrett LD. Communicating with oncology clients. Vet Clin North Am Small Anim Pract 2007; 37: 1013–1022.
17. Tzannes S, Hammond MF, Murphy S, et al. Owners ‘perception of their cats’ quality of life during COP chemotherapy for lymphoma. J Feline Med Surg 2008; 10: 73–81.
18. Hartmann K, Kuffer M. Karnofsky's score modified for cats. Eur J Med Res 1998; 3: 95–98.
19. Brown DC, Boston R, Coyne JC, et al. A novel approach to the use of animals in studies of pain: validation of the canine brief pain inventory in canine bone cancer. Pain Med 2009; 10: 133–142.
20. Wiseman-Orr ML, Nolan AM, Reid J, et al. Development of a questionnaire to measure the effects of chronic pain on health-related quality of life in dogs. Am J Vet Res 2004; 65: 1077–1084.
21. Yazbek KV, Fantoni DT. Validity of a health-related quality-of-life scale for dogs with signs of pain secondary to cancer. J Am Vet Med Assoc 2005; 226: 1354–1358.
22. Price MR, Bratton DL, Klinnert MD. Caregiver negative affect is a primary determinant of caregiver report of pediatric asthma quality of life. Ann Allergy Asthma Immunol 2002; 89: 572–577.
23. Boruk M, Lee P, Faynzilbert Y, et al. Caregiver well-being and child quality of life. Otolaryngol Head Neck Surg 2007; 136: 159–168.
24. Veterinary Cooperative Oncology Group. VCOG—common terminology criteria for adverse events (VCOG-CTCAE) following chemotherapy or biological antineoplastic therapy in dogs and cats. Ames, Iowa: Blackwell Publishing Ltd, 2011.
25. Rogers MP, Orav J, Black PM. The use of a simple Likert scale to measure quality of life in brain tumor patients. J Neurooncol 2001; 55: 121–131.
26. Rebok G, Riley A, Forrest C, et al. Elementary school-aged children's reports of their health: a cognitive interviewing study. Qual Life Res 2001; 10: 59–70.
27. Vogels T, Verrips GH, Verloove-Vanhorick SP, et al. Measuring health-related quality of life in children: the development of the TACQOL parent form. Qual Life Res 1998; 7: 457–465.
28. Fraser D. Understanding animal welfare: the science in its cultural context. Wiley-Blackwell, 2008.
29. Duncan IJ, Petherick JC. The implications of cognitive processes for animal welfare. J Anim Sci 1991; 69: 5017–5022.
30. Bono G, De Mori B. Animals and their quality of life: considerations ‘beyond mere welfare’. Vet Res Commun 2005; 29(suppl 2):165–168.
31. Kaasa S. Using quality of life assessment methods in patients with advanced cancer: a clinical perspective. Eur J Cancer 1995; 31A(suppl 6):S15–S17.
32. D'Eath M, Walls M, Hodgins M, et al. Quality of life of young people with an intellectual disability in Ireland. Interim report. Oranmore, Galway, Ireland: National Federation of Voluntary Bodies, 2003.
33. Cella D, Chang CH, Lai JS, et al. Advances in quality of life measurements in oncology patients. Semin Oncol 2002; 29: 60–68.
34. Carlson LE, Bultz BD, Morris DG. Individualized quality of life, standardized quality of life, and distress in patients undergoing a phase I trial of the novel therapeutic Reolysin (reovirus). Health Qual Life Outcomes 2005; 3: 7.
35. Seitzman RL, Glover DA, Meadows AT, et al. Self-concept in adult survivors of childhood acute lymphoblastic leukemia: a cooperative Children's Cancer Group and National Institutes of Health study. Pediatr Blood Cancer 2004; 42: 230–240.
36. Edwards B, Ung L. Quality of life instruments for caregivers of patients with cancer: a review of their psychometric properties. Cancer Nurs 2002; 25: 342–349.
37. Eiser C, Morse R. Quality-of-life measures in chronic diseases of childhood. Health Technol Assess 2001; 5: 1–157.
38. Eiser C, Morse R. A review of measures of quality of life for children with chronic illness. Arch Dis Child 2001; 84: 205–211.
39. Connolly MA, Johnson JA. Measuring quality of life in paediatric patients. Pharmacoeconomics 1999; 16: 605–625.
40. Cox S, O'Donoghue AC, McKenna WJ, et al. Health related quality of life and psychological wellbeing in patients with hypertrophic cardiomyopathy. Heart 1997; 78: 182–187.
41. Sprangers MA, Aaronson NK. The role of health care providers and significant others in evaluating the quality of life of patients with chronic disease: a review. J Clin Epidemiol 1992; 45: 743–760.
42. Gosling SD, Kwan VS, John OP. A dog's got personality: a cross-species comparative approach to personality judgments in dogs and humans. J Pers Soc Psychol 2003; 85: 1161–1169.
43. Overall KL. Clinical behavioral medicine for small animals. St Louis: Mosby, 1997.
44. Pacharinsak C, Beitz A. Animal models of cancer pain. Comp Med 2008; 58: 220–233.
45. Mellanby RJ, Herrtage ME, Dobson JM. Owners' assessments of their dog's quality of life during palliative chemotherapy for lymphoma. J Small Anim Pract 2003; 44: 100–103.
46. Knoester H, Grootenhuis MA, Bos AP. Outcome of paediatric intensive care survivors. Eur J Pediatr 2007; 166: 1119–1128.
47. Eiser C, Morse R. Can parents rate their child's health-related quality of life? Results of a systematic review. Qual Life Res 2001; 10: 347–357.
48. Boissy A, Manteuffel G, Jensen MB, et al. Assessment of positive emotions in animals to improve their welfare. Physiol Behav 2007; 92: 375–397.
