An adverse food reaction is defined as an adverse health effect that occurs reproducibly on exposure to a given food or food additive and includes immune-mediated, non-immune-mediated, and intolerant reactions.1–9 Clinical manifestations of adverse food reactions in dogs can include cutaneous and gastrointestinal signs, alone or in combination.9–12 Typically, dogs with CAFR have nonseasonal pruritus affecting the face, ears, perianal region, limbs, and feet.7–14 In addition, CAFRs can affect dogs at any age and has no reported sex predilection.8,14 Overall prevalence of CAFRs within a population of dogs with allergic dermatitis is estimated to be 10% to 25%.15
There are many challenges associated with the diagnosis of CAFR in veterinary medicine. For instance, there are no reliable laboratory tests, including all currently available serum-based IgE and IgG tests, that can refute or confirm the presence of CAFRs in dogs.16 Similarly, intradermal and patch testing modalities have low sensitivity and specificity for the diagnosis of adverse food reactions.16 As such, the current gold standard diagnostic test for adverse food reactions in dogs involves the performance of an EDT.1,13,16 Therefore, veterinarians are dependent on a client-administered diagnostic test to make the diagnosis of CAFR.
An EDT is performed in 2 phases. The first phase involves an owner exclusively feeding a novel commercial or home-cooked diet to the point of clinical resolution. This phase typically lasts for about 8 to 12 weeks. Improvement of clinical signs in this timeframe occurs in 90% of dogs and cats with CAFR if the EDT is performed strictly.13,17 The second phase of an EDT involves a provocative challenge of a specific food item to determine whether it is a trigger for CAFR.13,17 If after resolution of clinical signs (phase 1), signs of CAFR are triggered with a specific food item challenge (phase 2), a diagnosis of CAFR is made.13 Therefore, perfect adherence to a prescribed EDT is essential for the EDT outcome to be clinically useful. Adherence to EDTs arises from a collaborative and mutual relationship between the client and health professional,18–20 and issues with owner adherence to EDTs are attributed to between 3% and 79% of CAFRs not being properly diagnosed in dogs.10,21–26 However, little research has been conducted to understand specific factors affecting owner adherence to prescribed EDTs in veterinary medicine.
The HBM was developed in the 1950s to help explain the widespread failure of people to participate in programs that prevented and detected diseases and is a widely used conceptual framework in health behavior research.27,28 In the HBM, demographic, psychosocial, and knowledge variables (core concepts known as constructs) are assumed to influence health-related behavior.27 Constructs of the HBM include perceived susceptibility and severity, benefits and barriers, and self-efficacy and cues to action. The overall premise of the HBM is that people are likely to engage in a health behavior if they believe that they are susceptible to the condition (susceptibility), the condition could have potentially serious consequences (severity), and a course of action available to them could be of value in reducing health risk (benefits), and if perceived difficulties in performing the behavior are not strong enough to prevent action (barriers). Self-efficacy estimates one's confidence to perform a behavior, and cues to action refers to the social signals surrounding a given behavior (eg, everyone else is doing it).27 The HBM has been used to understand proxy-based health-care decision-making, such as pediatric medical decision-making.27,29
The purpose of the study presented here was to use the HBM to identify factors associated with owner adherence to EDT recommendations by veterinarians for dogs with suspected CAFRs. We hypothesized that at least one of the HBM constructs would be associated with 100% adherence to EDT recommendations. To our knowledge, the study presented here was the first direct application of the HBM in clinical veterinary medicine. We anticipated that results of this study could fill important gaps in knowledge and provide unique tools and insights for generalists and specialists who prescribe EDTs for dogs. Similarly, by addressing reasons for gaps in adherence specific to the diagnosis of CAFR, we suspected that clinical outcomes for dogs with CAFR could improve.
Materials and Methods
Survey participants
To identify dogs treated for suspected CAFRs between April 1, 2012, and April 1, 2017, electronic medical records were searched with the terms food allergy, adverse food reaction, cutaneous adverse food reaction, elimination diet trial, diet trial, novel protein, CAFR, and food trial. Records of dogs identified were then further reviewed to select only those dogs for which medical records contained explicit evidence that an EDT had been prescribed and performed for the purpose of diagnosing CAFR. Owners of such dogs were eligible to participate in the survey if they had sought care for their dog at Veterinary Healing Arts, East Greenwich, RI, between April 1, 2012, and April 1, 2017; had performed an EDT for their dog as indicated by a notation in the medical record during the study period; were ≥ 18 years of age at the time of the survey; provided a valid email address; and consented to survey participation. A cross-sectional study design was employed.
Survey
An online survey was developed on the basis of the HBM. Eligible owners were contacted by email and US Postal Service-delivered mail to request their participation in the anonymous survey about their experience with their dogs’ EDTs. This message contained a link to the survey, which remained open for 30 days, from June 14 to July 14, 2017. All study procedures were approved by the George Mason University Institutional Review Board. Incentives were not given for participation in this study.
Data were collected with a secure online survey assessment tool.a Personal identifying information was not collected from the survey and, therefore, not connected to dogs’ medical records. The survey instrument included questions about demographics, knowledge about diets and CAFRs in dogs, and HBM-driven psychosocial attitudes. The main outcome measure for the study was whether perfectly strict adherence (defined as 100% adherence) to the EDT had occurred and was tied to the question “On a scale of 1 to 5, please rate the overall performance of your dog's diet trial.” Response options comprised a 5-point Likert scale from 1 (poor performance, not strict adherence) to 5 (perfect food trial with no indiscretion or off-limit foods given). Responses were dichotomized into those with a response of 5 (100% adherence) versus those with all other responses.
Owner knowledge regarding diets and CAFRs in dogs was assessed with 6 true-or-false questions (Table 1). Answers were coded as correct (received 1 point) or incorrect (received no point) on the basis of consistency with the most up-to-date scientific information available at the time of survey development. The sum of each respondent's points was calculated for the score, with values ranging from 0 to 6. Modeled after the HBM, the survey collected data on psychosocial attitude constructs of respondents’ perceived susceptibility, severity, benefits, barriers, self-efficacy, and social support related to EDT recommendations.
Measurement characteristics and items (statements), stratified by variable and HBM constructs, that were used in an online survey to assess psychosocial attitudes potentially associated with adherence to EDT recommendations by 192 respondents of 665 (28.8%) invited survey participants who owned dogs prescribed an EDT to diagnose CAFR between April 1, 2012, and April 1, 2017.
| Construct | Items | No. of items | Score range | Cronbach α |
|---|---|---|---|---|
| Knowledge* | A grain-free diet from the pet store is an example of a diet trial. FALSE | 6 | 0–6 | NA |
| Dogs are usually allergic to the protein source (chicken, beef, etc) in their diet. TRUE | ||||
| Dogs are usually allergic to the grain source (wheat, corn, etc) in their diet. FALSE | ||||
| The only accurate test available for the diagnosis of food allergy in dogs is a diet trial. TRUE | ||||
| For a diet trial to be effective at diagnosing food allergy, it must be strict. TRUE | ||||
| Food that is labeled as “natural” or “organic” is appropriate to give during a diet trial. FALSE | ||||
| Perceived | Before starting the diet trial: | 2 | 2–10 | 0.798 |
| susceptibility† | I believed that my dog's signs could be from a food allergy. | |||
| I believed that the chance of my dog's signs being caused by a food allergy was … | ||||
| Perceived severity† | Before starting the diet trial: | 4 | 4–20 | 0.803 |
| My dog's signs were impacting its quality of life. | ||||
| My dog's signs were upsetting to me. | ||||
| I was worried about my dog. | ||||
| I thought that my dog's signs were severe. | ||||
| Perceived benefits† | I understood why a diet trial was recommended by my veterinarian. | 4 | 4–20 | 0.651 |
| I understood that my dog's condition could be diagnosed with this diet trial. | ||||
| I was hopeful that my dog's signs could be fixed with this diet trial. | ||||
| I believed that the diet trial was an important step in figuring out what my dog was allergic to. | ||||
| Perceived barriers† | My hectic lifestyle made the diet trial difficult. | 5 | 5–25 | 0.623 |
| Keeping my dog from eating trash, litter, etc outside made the diet trial difficult. | ||||
| Finding a way to give my dog its medicine without using “off-limits” food or treats was difficult during the diet trial. | ||||
| The cost of the food made the diet trial difficult. | ||||
| Performing the diet trial over a holiday made the diet trial difficult. | ||||
| Self-efficacy† | I found sticking to the diet trial easy. | 6 | 6–30 | 0.735 |
| I felt comfortable calling my veterinarian if I had any problems or questions about the diet trial. | ||||
| I felt confident that I could follow the diet trial strictly. | ||||
| I felt confident that I could follow the diet trial as directed. | ||||
| I had the ability to change my routines as needed to make the diet trial work. | ||||
| I had the ability to change my habits to make the diet trial work. | ||||
| Social support† | Overall, I felt like I had adequate support in my home to perform the diet. | 2 | 2–10 | 0.791 |
| It was easy to convince the other adults in my home to follow the rules of the diet trial. |
Each answer was coded as correct (received 1 point) or incorrect (received no point).
Used a 5-point Likert scale, ranging from 1 (strongly disagree) to 5 (strongly agree).
NA = Not applicable.
A 5-point Likert scale, ranging from 1 (strongly disagree) to 5 (strongly agree), was used for respondents to self-report their perspectives for each item under each construct (Table 1). The construct of perceived susceptibility and social support each had 2 items (statements), and respondents’ scores for each construct with the Likert scale could have ranged from 2 (strongly disagree with each item) to 10 (strongly agree with each item). The constructs of perceived severity and perceived benefits each had 4 items; thus, respondents’ scores for each construct could have ranged from 4 (strongly disagree with each item) to 20 (strongly agree with each item). Similarly, the constructs of perceived barriers and self-efficacy each had 5 and 6 items, respectively, and scores for each could have ranged from 5 to 25 and from 6 to 30, respectively.
Although the HBM had not been previously used in clinical veterinary research, we drew on our combined expertise to develop questions that best captured the core constructs of the HBM to evaluate factors that impact adherence to EDTs in dogs. Care was taken to strike a balance between providing a comprehensive range of questions and limiting the overall number of questions so that the survey was not too burdensome for participants to complete.
Statistical analysis
All analyses were conducted with available software.b Cronbach α was calculated for each HBM-driven construct to determine internal consistency between statements used to assess that given construct. χ2 tests were used to identify differences in responses to demographic variables for 100% versus < 100% EDT adherence. Student t test was used to compare the mean scores for knowledge and HBM-driven constructs between respondents who reported 100% adherence versus < 100% adherence to EDT recommendations. Unadjusted logistic regression was used to identify potential associations between predictor variables (demographics, knowledge, and HBM-driven psychosocial attitudes) and the outcome variable (100% adherence to EDT recommendations). Variables with values of P ≤ 0.05 in the unadjusted logistic regression analyses were included in the multivariable logistic regression analyses that were then used to determine associations between predictor variables and 100% adherence to EDT recommendations, when controlling for all other variables in the model.
Results
Survey participants
A search of the medical records initially identified 1,735 records that were then further reviewed to select only those dogs for which medical records contained explicit evidence that an EDT had been prescribed and performed for the purpose of diagnosing CAFR and that had results of the EDT clearly documented. In total, 665 owners (1 owner per dog) were invited to participate in the survey, 216 provided informed consent and initiated the survey, and 192 completed the entire survey. Although owners of dogs for which EDTs had been prescribed by general practitioners were included, the resulting population of respondents included only those who ultimately sought specialty care.
Of the 192 respondents, 164 (85.4%) reported that they were female and 28 (14.6%) reported that they were male. By age, 39 of the 192 (20.3%) respondents reported being from 18 to 34 years old, 77 (40.1%) from 35 to 54 years old, and 76 (39.6%) ≥ 55 years old. One hundred eighty of the 192 (93.8%) respondents identified themselves as white, and 12 (6.3%) identified themselves as nonwhite. A college degree or higher was reported by 123 of the 192 (64.1%) respondents. In addition, respondents most commonly reported their household as having 2 adults (127/192 [66.2%]), no children (76 [39.6%]), no other dogs (121 [63%]), or no other pets (125 [65.1%]), alone or in combination. Further, only 23 of the 192 (12%) respondents reported that they alone were responsible for feeding the dog.
Bivariate analysis
Of 192 survey participants, 77 (40.1%) reported 100% adherence to EDT recommendations, whereas 115 (59.9%) reported < 100% adherence to EDT recommendations. No meaningful differences were detected for demographic variables or perceived susceptibility between respondents who reported 100% adherence versus < 100% adherence to EDT recommendations.
However, the mean ± SD scores on sections of the survey pertaining to knowledge of diets and CAFRs in dogs, perceived benefits, perceived self-efficacy, and social support were significantly (P = 0.002, P = 0.032, P < 0.001, P = 0.006, respectively) higher for respondents who reported 100% adherence to EDT recommendations (3.56 ± 1.28, 18.79 ± 1.62, 27.97 ± 2.57, and 9.40 ± 1.12, respectively), compared with respondents who reported < 100% adherence to EDT recommendations (2.96 ± 1.37, 18.17 ± 2.13, 25.81 ± 3.43, and 8.68 ± 1.86; respectively). Further, the mean ± SD score on the section of the survey pertaining to perceived barriers was significantly (P < 0.001) lower for respondents who reported 100% adherence with EDT recommendations (10.29 ± 3.76), compared with respondents who reported < 100% adherence to EDT recommendations (12.94 ± 3.98).
Multivariable analysis
In multivariable analyses when controlling for all variables in the model, respondents with higher scores for perceived barriers (reported more barriers) in performing EDTs were significantly (ORa = 0.86, P = 0.010) less likely to report 100% adherence to EDT recommendations than were respondents with lower scores (reported fewer barriers; Table 2). Respondents with higher scores in the knowledge section and the self-efficacy section of the survey were significantly (ORa = 1.30, P = 0.049; ORa = 1.18, P = 0.019; respectively) more likely to report 100% adherence to EDT recommendations, compared with respondents with lower scores in the knowledge section and the self-efficacy section, respectively. In the multivariable model, no meaningful differences in reported perceived benefits or social support were identified between respondents who reported 100% adherence versus < 100% adherence to EDT recommendations.
Results of multivariable logistic regression analysis of data collected from the 192 respondents in Table 1 to identify factors associated with owner adherence to EDT recommendations in dogs.
| Variables | ORc (95% CI) | P value | ORa (95% CI) | P value |
|---|---|---|---|---|
| Knowledge | 1.40 (1.12–1.76) | 0.003 | 1.30 (1.00–1.69) | 0.049 |
| HBM constructs | ||||
| Perceived susceptibility | 1.00 (0.84–1.18) | 0.953 | — | — |
| Perceived severity | 0.99 (0.86–1.15) | 0.930 | — | — |
| Perceived benefits | 1.20 (1.01–1.42) | 0.036 | 1.10 (0.91–1.31) | 0.323 |
| Perceived barriers | 0.83 (0.76–0.91) | < 0.001 | 0.86 (0.79–0.97) | 0.010 |
| Self-efficacy | 1.29 (1.15–1.46) | < 0.001 | 1.18 (1.03–1.35) | 0.019 |
| Social support | 1.39 (1.10–1.77) | 0.006 | 1.16 (0.89–1.50) | 0.272 |
— = Not calculated because the variable was not identified as significant in the model. CI = Confidence interval. ORc = Crude (unadjusted) OR.
Discussion
Peer-reviewed research regarding client compliance or adherence to recommendations by veterinarians is sparse. Previous studies have focused on behavioral interventions,30,31 short-term antimicrobial treatment in dogs,32–36 vaccination compliance and heartworm medication purchase,21,37–40 sporotrichosis treatment,41 and immunotherapy refill adherence.42 In addition, studies20,21,38,43 have focused on owner adherence to recommendations associated with dental and surgical procedures. Although some studies10,21–26 estimate adherence to EDT recommendations, they lack descriptive characterizations regarding why a breakdown in adherence occurred.
Adherence to treatment guidelines impacts health-care outcome. The World Health Organization estimates that only 50% of human patients with chronic illness in developed countries adhere to long-term treatment guidelines.44 Deeper understanding of the factors associated with poor adherence is needed, and investigations with a psychosocial framework, like the HBM, provide much needed information about behavior in making health decisions that may strengthen clinical practice and outcomes.27
To the best of our knowledge, the present study was the first to use the HBM to evaluate demographic, knowledge, and psychosocial factors associated with owner adherence to EDT recommendations in dogs. Results supported our hypothesis that at least one of the HBM constructs would be associated with 100% adherence to EDT recommendations. Knowledge, perception of barriers, and perceived self-efficacy were factors that appeared to substantially impact respondents’ reported adherence to EDTs.
Our study evaluated whether owner knowledge about basic facts related to EDTs, CAFRs, and diets in dogs would have an impact on adherence to EDTs as instructed, and results indicated that respondents with greater knowledge about diets and CAFRs in dogs had greater odds of 100% adherence to EDT recommendations than did respondents with lesser knowledge on the subjects. This result was consistent with a previous study45 in human medicine that shows that knowledge of health risks and benefits of a given health practice is a precondition for change. Similar results also exist in veterinary medicine.23,39 For example, individuals who were minimally informed about the risk of rabies were less likely to vaccinate their pet after an out-break.39 Similarly, when surveyed about challenges associated with atopic dermatitis consultations, dog owners frequently complained about receiving too little information or explanation regarding their dogs’ conditions, suggesting that increased knowledge would improve client involvement and possibly adherence.c
The method by which veterinarians and clients communicate is linked to client knowledge and adherence,20,30,31,46–52 and inadequate veterinary communication is a leading reason for poor client compliance.21 Recent studies20,46–52 describe how specific communication styles used by veterinarians can impact adherence and show that relationship-centered communication styles positively impact adherence. Relationship-centered communication is characterized as a collaborative partnership between a veterinarian and a client that preserves client perspectives and values.53 Although our study did not directly assess how communication methods affected adherence to EDT recommendations, previous studies20,47 show that clients prefer information that is both written and verbal, delivered in a thoughtful and concise manner, repeated in different ways, and provided in understandable language. In addition, encouraging questions from clients can improve adherence.47 Similarly, spending adequate time reviewing the diagnostic plan with a client can lead to a better clinical outcome.21,34,47 Further research is warranted to determine whether these communication strategies would improve owner adherence to EDT recommendations.
Barriers are unique limitations that are defined as possible perceived obstacles to taking action.27 Results of the present study indicated that owners who reported having multiple barriers, including a hectic lifestyle, were less likely to have reported 100% adherence to EDT recommendations. This finding was consistent with that of other studies30,32,35,36 in veterinary medicine. Given that an EDT can require lifestyle adjustments of owners, it is important to discuss how these adjustments are to be made with owners to better ensure ideal adherence conditions are met. When an EDT is recommended for dogs receiving oral medications, it is also important that veterinarians discuss with owners how medications will be given during the EDT. Because medication is often administered in a flavored vehicle, finding an acceptable alternative for such during an EDT could be challenging and perceived by the owner as a possible barrier to adherence. We believe that providing an owner with a plan for medication delivery could improve adherence.
In the present study, the cost of prescribed food was perceived as a barrier that impacted adherence. Although there is conflicting evidence regarding how cost impacts adherence,20,42,54–56 a report21 shows that only 4% of treatments are abandoned because of cost-related factors. Regardless, conversations with owners about the cost of veterinary care should be direct, clear, and personalized.54,55 Veterinarians should aim to make meaningful connections between cost and therapeutic recommendations to mitigate potential cost barriers.21,54,55
Self-efficacy is the conviction or confidence that one can successfully execute a behavior.27 Results of the present study indicated that respondents with greater reported self-efficacy had greater odds of reporting 100% adherence to EDT recommendations. Care should be taken to pay attention to clients when they report unease or poor confidence with performance of an EDT. Providing additional feedback or boosting confidence could improve adherence and EDT outcome. In addition, confidence in performing an EDT strictly and as directed was also related to an owner's reported comfort in contacting their veterinarian should any issues or questions with the EDT arise. The relationship shared between the veterinarian and client is important and will impact clinical outcome.20,52 How well veterinary information is presented is directly related to how well owners adhere to veterinary recommendations.20,21,35,52
Interestingly, the construct of perceived susceptibility was not associated with adherence to EDT recommendations in the present study. From a perspective of health care by proxy, such as that in veterinary medicine, perceived susceptibility is the belief in the likelihood of the patient developing a given condition.27 Thus, whether owners believed their dogs to be susceptible to CAFRs did not impact their ability to perform the EDT with 100% adherence. This finding suggested that when prescribing EDTs, veterinarians could relax their feelings of burden to convince owners that their dogs could be susceptible to CAFR.
The construct of severity, or the belief that a given condition is serious,27 was not associated with adherence to EDT recommendations in the present study. This finding was similar to results of a study32 that shows owner perception of the severity of infection does not impact adherence to antimicrobial delivery. However, in a study30 in cats, adherence was associated with a decrease in severity, which supports the concept that adherence improves once results are demonstrated.31 This was not directly assessed in the present study, and determining the impact of clinical improvement on adherence to veterinary recommendations warrants future research.
There were limitations to the present study. We used data from a homogenous group of individuals who sought specialty-level care for their dogs. Although individuals who had an EDT prescribed by a general practitioner were included in the study, the population of individuals surveyed included only those who had sought specialty care. These individuals often have increased financial means and commitment to diagnosing and alleviating their dogs’ clinical conditions.20 Therefore, we recommend that an investigation similar to the present study be performed on a wider audience of participants from various socioeconomic backgrounds. We also believe that including owners of other species (eg, cats) in such an expanded study would help unveil unique, species-specific factors that affect owner adherence to veterinary recommendations.
Another limitation of the present study was the reliance on self-reporting in our survey model to generate data for analysis. Although self-reporting can be a reliable method for assessment of adherence, there is a risk of over- or underestimation of adherence when self-reporting is used.34 It was possible that the overall adherence to EDT recommendations was actually different from that reported in the present study. Nonetheless, we believe that it is important to understand the rate of and reasons for adherence failure to improve clinical practice outcome.
The retrospective nature of the present study could have resulted in recall bias or memory lapse by respondents. Adherence discussions between veterinarians and owners immediately before and after recommending an EDT could help reduce this apparent flaw. Regardless, attempts to understand owner adherence to recommendations should not be stymied by concerns over self-reporting accuracy. We believe that aside from direct, continuous monitoring of dogs in their home environments, assessment of adherence to EDT recommendations, and other veterinary recommendations for that matter, relies on self-reporting from our clients.
Results of the present study indicated that the HBM could be adapted to understand factors associated with adherence to specific health recommendations in clinical veterinary medicine and that core constructs from the HBM, specifically perceived barriers and self-efficacy, were major determinants of owners’ adherence to EDT recommendations. In addition, findings suggested that client education to improve owner knowledge of diets and CAFRs in dogs could improve adherence to EDT recommendations. When prescribing an EDT, veterinarians should focus on helping owners reduce perceived barriers (eg, discuss medication delivery and lifestyle changes), improve owner confidence (eg, maintain positivity about the owner's ability to make changes needed to perform the EDT properly), and enhance related knowledge (eg, provide written and verbal information related to the EDT, nutrition, food allergies, and CAFRs). Because there is no currently available diagnostic test as accurate or reliable as an EDT, we recommend that veterinarians use the findings of the present study to more carefully and thoughtfully make recommendations when prescribing an EDT for the purpose of diagnosing suspected CAFR. Future research is warranted to design and test the efficacy of targeted interventions that address specific factors associated with EDT adherence.
Acknowledgments
The authors declare that there were no conflicts of interest.
The authors thank Dr. Jessica Mintz, Dr. Craig Griffin, and Dr. Mandy Burrows for their intellectual contributions and technical assistance.
ABBREVIATIONS
| EDT | Elimination diet trial |
| CAFR | Cutaneous adverse food reaction |
| HBM | Health Belief Model |
| ORa | Adjusted OR |
Footnotes
Qualtrics online survey software, version XM, Qualtrics LLC, Provo, Utah.
SPSS Statistics for Windows, version 24.0, IBM Corp, Armonk, NY.
Bensignor EJ, Merven F. Canine atopic dermatitis consultations: what is important for owners of atopic dogs in a specialty practice? (abstr) Vet Dermatol 2017;28:535.
References
1. Verlinden A, Hester M, Millet S, et al. Food allergy in dogs and cats: a review. Crit Rev Food Sci Nutr 2006;46:259–273.
2. Sampson HA. Food allergy. Part 1: immunopathogenesis and clinical disorders. J Allergy Clin Immunol 1999;103:717–728.
3. Cianferoni A, Spergel JM. Food allergy: review, classification and diagnosis. Allergol Int 2009;58:457–466.
4. Turnbull JL, Adams HN, Gorard DA. Review article: the diagnosis and management of food allergy and food intolerances. Aliment Pharmacol Ther 2015;41:3–25.
5. Wills J, Harvey R. Diagnosis and management of food allergy and intolerance in dogs and cats. Aust Vet J 1994;71:322–326.
6. Jackson HA. The pathogenesis of food allergy. In: Noli C, Foster A, Rosenkrantz W, eds. Veterinary allergy. Chichester, England: John Wiley & Sons Ltd, 2014;103–107.
7. Veenhof EZ, Knol EF, Willemse T, et al. Immune responses in dogs with cutaneous adverse food reactions. Vet Q 2012;32:87–98.
8. Carlotti DN, Remy I, Prost C. Food allergy in dogs and cats. A review and report of 43 cases. Vet Derm 1990;1:55–62.
9. Gaschen FP, Merchant SR. Adverse food reactions in dogs and cats. Vet Clin North Am Small Anim Pract 2011;41:361–379.
10. Loeffler A, Lloyd DH, Bond R, et al. Dietary trials with a commercial chicken hydrolysate diet in 63 pruritic dogs. Vet Rec 2004;154:519–522.
11. Paterson S. Food hypersensitivity in 20 dogs with skin and gastrointestinal signs. J Small Anim Pract 1995;36:529–534.
12. Picco F, Zini E, Nett C, et al. A prospective study on canine atopic dermatitis and food-induced allergic dermatitis in Switzerland. Vet Dermatol 2008;19:150–155.
13. Rosser EJ Jr. Diagnosis of food allergy in dogs. J Am Vet Med Assoc 1993;203:259–262.
14. White SD. Food hypersensitivity in 30 dogs. J Am Vet Med Assoc 1986;188:695–698.
15. Olivry T, Mueller RS. Critically appraised topic on adverse food reactions of companion animals (3): prevalence of cutaneous adverse food reactions in dogs and cats. BMC Vet Res 2017;13:51.
16. Mueller RS, Olivry T. Critically appraised topic on adverse food reactions of companion animals (4): can we diagnose adverse food reactions in dogs and cats with in vivo or in vitro tests? BMC Vet Res 2017;13:275.
17. Olivry T, Mueller RS, Prelaud P. Critically appraised topic on adverse food reactions of companion animals (1): duration of elimination diets. BMC Vet Res 2015;11:225.
18. Donovan JL, Blake DR. Patient non-compliance: deviance or reasoned decision-making? Soc Sci Med 1992;34:507–513.
19. Donovan JL. Patient decision making. The missing ingredient in compliance research. Int J Technol Assess Health Care 1995;11:443–455.
20. Kanji N, Coe JB, Adams CL, et al. Effect of veterinarian-client-patient interactions on client adherence to dentistry and surgery recommendations in companion-animal practice. J Am Vet Med Assoc 2012;240:427–436.
21. The path to high quality care: practical tips for improving compliance. Lakewood, Colo: American Animal Hospital Association, 2003;82–83.
22. Tapp T, Griffin C, Rosenkrantz W, et al. Comparison of a commercial limited-antigen diet versus home-prepared diets in the diagnosis of canine adverse food reaction. Vet Ther 2002;3:244–251.
23. Chesney CJ. Food sensitivity in the dog: a quantitative study. J Small Anim Pract 2002;43:203–207.
24. Biourge VC, Fontaine J, Vroom MW. Diagnosis of adverse reactions to food in dogs: efficacy of a soy-isolate hydrolysate-based diet. J Nutr 2004;134(suppl 8):2062S–2064S.
25. Loeffler A, Soares-Magalhaes R, Bond R, et al. A retrospective analysis of case series using home-prepared and chicken hydrolysate diets in the diagnosis of adverse food reactions in 181 pruritic dogs. Vet Dermatol 2006;17:273–279.
26. Proverbio D, Perego R, Spada E, et al. Prevalence of adverse food reactions in 130 dogs in Italy with dermatological signs: a retrospective study. J Small Anim Pract 2010;51:370–374.
27. Skinner CS, Tiro J, Champion VL. The health belief model. In: Glanz K, Rimer BK, Viswanath K, eds. Health behavior: theory, research and practice. 5th ed. San Francisco: Jossey-Bass, 2015;75–94.
28. Painter JE, Borba CP, Hynes M, et al. The use of theory in health behavior research from 2000–2005: a systematic review. Ann Behav Med 2008;35:358–362.
29. Jones CJ, Llewellyn CD, Frew AJ, et al. Factors associated with good adherence to self-care behaviours amongst adolescents with food allergy. Pediatr Allergy Immunol 2015;26:111–118.
30. Casey RA, Bradshaw JWS. Owner compliance and clinical outcome measures for domestic cats undergoing clinical behavior therapy. J Vet Behav 2008;3:114–124.
31. Jobling R, Creighton E. Fostering adherence to horse behavior counseling. J Vet Behav 2011;6:276–286.
32. Bomzon L. Short-term antimicrobial therapy - a pilot compliance study using ampicillin in dogs. J Small Anim Pract 1978;19:697–700.
33. Barter LS, Watson ADJ, Maddison JE. Owner compliance with short term antimicrobial medications in dogs. Aust Vet J 1996;74:277–280.
34. Barter LS, Maddison JE, Watson ADJ. Comparison of methods to assess dog owners’ therapeutic compliance. Aust Vet J 1996;74:443–446.
35. Grave K, Tanem H. Compliance with short-term oral antibacterial drug treatment in dogs. J Small Anim Pract 1999;40:158–162.
36. Adams VJ, Campbell JR, Waldner CL, et al. Evaluation of client compliance with short-term administration of antimicrobials to dogs. J Am Vet Med Assoc 2005;226:567–574.
37. Nelson RS, Mshar PA, Cartter ML, et al. Public awareness of rabies and compliance with pet vaccination laws in Connecticut, 1993. J Am Vet Med Assoc 1998;212:1552–1555.
38. Compliance: taking quality to the next level. Lakewood, Colo: American Animal Hospital Association, 2009.
39. McNeil CS, Nagy S, Moonan C, et al. Community rabies knowledge and pet vaccination practices after a skunk rabies outbreak in Eddy County, New Mexico. J Am Vet Med Assoc 2015;246:1242–1247.
40. Cummings J, Vickers L, Marbaugh J. Evaluation of veterinary dispensing records to measure “clinic compliance” with recommended heartworm prevention programs, in Proceedings. Am Heartworm Soc Symp 1995;183–186.
41. Chaves AR, de Campos MP, Barros MBL, et al. Treatment abandonment in feline sporotrichosis - study of 147 cases. Zoonoses Public Health 2013;60:149–153.
42. Tater KC, Cole WE, Pion PD. Allergen-specific immunotherapy prescription patterns in veterinary practice: a US population-based cohort study. Vet Dermatol 2017;28:362–e82.
43. Miller BR, Harvey CE. Compliance with oral hygiene recommendations following periodontal treatment in client-owned dogs. J Vet Dent 1994;11:18–19.
44. Adherence to long-term therapies: evidence for action. Geneva: World Health Organization, 2003; XIII.
45. Kelder SH, Hoelscher D, Perry CL. How individuals, environments, and health behaviors interact. In: Glanz K, Rimer BK, Viswanath K, eds. Health behavior: theory, research and practice. 5th ed. San Francisco: Jossey-Bass, 2015;159–182.
46. Shaw JR, Adams CL, Bonnett BN, et al. Veterinarian-client-patient communication during wellness appointments versus appointments related to a health problem in companion animal practice. J Am Vet Med Assoc 2008;233:1576–1586.
47. Stoewen DL, Coe JB, MacMartin C, et al. Qualitative study of the communication expectations of clients accessing oncology care at a tertiary referral center for dogs with life-limiting cancer. J Am Vet Med Assoc 2014;245:785–795.
48. Dysart LM, Coe JB, Adams CL. Analysis of solicitation of client concerns in companion animal practice. J Am Vet Med Assoc 2011;238:1609–1615.
49. Shaw JR, Barley GE, Broadfoot K, et al. Outcomes assessment of on-site communication skills education in a companion animal practice. J Am Vet Med Assoc 2016;249:419–432.
50. Shaw JR, Adams CL, Bonnett BN. What can veterinarians learn from studies of physician-patient communication about veterinarian-client-patient communication? J Am Vet Med Assoc 2004;224:676–684.
51. 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.
52. Lue TW, Pantenburg DP, Crawford PM. Impact of the owner-pet and client-veterinarian bond on the care that pets receive. J Am Vet Med Assoc 2008;232:531–540.
53. Shaw JR. Four core communication skills of highly effective practitioners. Vet Clin North Am Small Anim Pract 2006;36:385–396.
54. Coe JB, Adams CL, Bonnett BN. Prevalence and nature of cost discussions during clinical appointments in companion animal practice. J Am Vet Med Assoc 2009;234:1418–1424.
55. Coe JB, Adams CL, Bonnett BN. A focus group study of veterinarians’ and pet owners’ perceptions of the monetary aspects of veterinary care. J Am Vet Med Assoc 2007;231:1510–1518.
56. Dell DL, Griffin CE, Thompson LA, et al. Owner assessment of therapeutic interventions for canine atopic dermatitis: a long-term retrospective analysis. Vet Dermatol 2012;23:228–e47.
