In veterinary medicine, nutrition is identified as the fifth vital assessment along with a patient's temperature, pulse, respiration, and pain.1,2 Guidelines promote that a nutritional assessment and a nutritional recommendation be part of every animal's visit to a small animal veterinarian and advise that a detailed diet history is a key aspect of conducting a comprehensive nutritional assessment.1,2 Obtaining a detailed diet history is also key to a veterinarian's ability to develop and deliver an acceptable and appropriate nutritional recommendation.3 Continuing to identify evidence-based communication strategies and techniques that can assist veterinary professionals in obtaining a comprehensive diet history is important to the health of veterinary patients.
To maximize prevention of diet-related adverse conditions (eg, obesity)4 or management of chronic conditions (eg, food allergies),5 communication with a pet's caregiver, including solicitation of a comprehensive diet history, is crucial. A comprehensive diet history includes information about the animal's main diet and eating behavior, the client's and others' feeding practices (including provision of treats or dietary supplements), and the animal's environment and exercise.3 Previous observational research by our group has found that the process of gathering a diet history and proposing a long-term dietary change can be a delicate and complex process requiring specific attention to the communication processes involved.6,7
Results of an observational study6 of veterinarian-client-patient interactions that assessed the effects of question design on the dietary information solicited from dog and cat owners revealed that the most common approach veterinarians use to initiate a discussion of a patient's diet is a simple what-prefaced question (eg, “What kind of food is he [or she] on?”). These questions resulted in the reporting of restricted information by clients about their pet's diet. In 57 of 64 (89%) instances, the client reported only 1 or 2 food items provided to the animal, with information about treats or human foods disclosed in only 5 of 64 (8%) interactions. The study6 results indicated that both veterinarians and clients perceived the what-prefaced question as a specifying, closed-ended question for which only a brief response was expected; further, in 48 of 64 (75%) interactions that began with a veterinarian's question prefaced in this manner, the veterinarian did not ask about additional food items, which suggested that participating veterinarians were soliciting limited diet histories. Client underreporting of items fed can potentially have negative health consequences associated with an item's nutrient excesses, deficiencies, or imbalances.6,8 As a result, it was proposed that veterinarians explore new communication approaches for gathering complete and comprehensive diet histories from pet caregivers.6
Communication best practices encourage veterinary professionals to begin gathering information from caregivers with open-ended questions or statements.9 Questions or statements framed in an open-ended manner allow clients considerable freedom in their responses and invite them to expand on their answers. One suggestion, specific to the solicitation of veterinary patient dietary information, is to use questions that are more open, rather than more targeted questions that request specific, concise pieces of information.6 On the basis of information obtained in a study10 that involved conversation analysis of dietitians' solicitation of human diet histories, the statement “Tell me about everything [pet name] eats throughout a day, starting first thing in the morning right through to the end of the day” is 1 type of open-ended statement proposed to solicit a pet's diet history from the caregiver.6 Further, the possibility was raised that lexical features in the design of veterinarians' what-prefaced questions, including reference to food in the singular form (instead of the plural foods), may contribute to the limited diet-related information gathered through this question form, and it was proposed that simply minimizing the specificity of question scope by use of the plural term foods might result in more diet information being obtained from clients.6 Researching ways to encourage clients to maximize the amount of diet-related information given with each question asked can assist veterinary professionals in providing optimal nutritional advice.
Although identifying the prevalence of what-prefaced diet history questions by veterinarians during consultation with clients was a strength of the aforementioned study,6 the lack of variation in the design of veterinarians' questions prevented systematic investigation of the impact that different types of question design might have had on the amount of diet-related information elicited from clients. The primary objective of the study reported here was to examine the effect of 3 different diet history questions identified in the aforementioned study6 on gathering diet-related information from veterinary clients. We also aimed to assess whether the amount and type of information gathered in person with these approaches would differ from the amount and type of information gathered by use of a diet history survey filled out independently by each client.
Materials and Methods
The prospective, cross-sectional evaluation study was performed from June 29 to August 21, 2015. A convenience sample of cat and dog owners who had scheduled appointments with participating veterinarians at 4 small-animal veterinary clinics within a 2-hour driving radius of Guelph, ON, Canada, was invited to participate in the study. The study protocol was reviewed and approved by the University of Guelph Research Ethics Board (REB No. 15MY032).
Training for delivery of diet history questions
One of the authors (KAJ), in the role of primary research assistant, was trained to a standardized process for delivering 3 diet history questions and 1 history question related to physical activity. The 3 diet history questions, created on the basis of findings in a previous study6 of diet history questions by our group, were as follows: “Tell me everything he [or she] eats throughout a day, starting first thing in the morning right through to the end of the day”; “What kind of food does she [or he] eat?”; and “What kind of foods does he [or she] eat?” The physical activity–related history question was, “Tell me about the kinds of physical activity he [or she] gets during a typical day.”
Training to standardize the primary research assistant's presentation of the 4 questions involved interactions with a convenience sample of 7 pet owners known to the researchers. The research assistant was viewed from an observation room through a 2-way mirror by other members of the research team (JBC and REO) during each interaction. Between interactions, feedback and coaching were provided to the research assistant to standardize verbal delivery of the 4 questions as well as nonverbal communications (eg, voice inflections and hand gestures). Training proceeded until the delivery of the study questions was deemed consistent and standardized. The training session was also used to evaluate the overall study protocol and to train 2 additional research assistants in recruiting pet owners for study participation and obtaining informed consent.
Participant recruitment
One of the 2 research assistants trained to recruit participants was responsible for approaching dog and cat owners in the waiting room of the small animal veterinary clinic before the client's scheduled appointment. The research assistant used a standardized script to explain the study in detail, answered questions about the study, and obtained written informed consent from individuals who agreed to participate. Participants were informed that the purpose of the study was to describe veterinarian-client-patient interactions in companion animal practice and to examine the effect of communication between veterinarians and their clients, including communication regarding companion animal nutrition and physical activity. Clients who had appointments scheduled with the participating veterinarian were approached by the research assistant, except for pet owners attending their pet's scheduled euthanasia, pet owners with accompanying children, pet owners with a litter of newborn animals, or pet owners accompanied by a pet other than a cat or dog.
Video-recorded study interactions
At each of the 4 veterinary clinics, a video camera was set up in a dedicated examination room and left running continuously during the data collection visit. Following recruitment, the primary research assistant approached each participant in the examination room prior to their interaction with the attending veterinarian.
Using a standardized script, the primary research assistant introduced herself and informed the client that she was working with the attending veterinarian and would be asking the client 2 questions. The first question posed to the participant was always 1 of the 3 differently worded diet history questions, which had been randomly assigned to participant identification numbers at the start of the study with an online random number generator.a The question about the pet's daily physical activity was used as a standardized distractor and was posed to all participants after they had answered the first question. All questions during this stage of the study were asked by the primary research assistant in a standardized manner, with the research assistant paying particular attention to controlling the use of minimal encouragers (eg, “okay,” “mm hm,” or nodding) in response to the client's reply. The primary research assistant did not provide assistance to participants with answering the questions and did not ask participants to elaborate on their responses. If a client asked for clarification, the research assistant was trained to repeat the original question. Once the client answered both questions, the research assistant explained that the attending veterinarian would be present shortly and left the room.
Postinteraction diet history survey
A postinteraction diet history survey was developed on the basis of diet history forms used at the time of the study by the Clinical Nutrition Service at the Ontario Veterinary College Health Sciences Centre11 and published by the World Small Animal Veterinary Association.12 Additional questions regarding participant demographics were generated by the authors.
An online version of the diet history survey was pretested among 16 pet owners (known to the authors) to determine the ease and flow of the survey format, to clarify survey questions, to identify whether relevant questions were missing, and to obtain general feedback about the survey. Once these individuals filled out the survey online, 1 author (KAJ) followed up with the participants by telephone to discuss their experience with the survey and record their feedback. This feedback was used to make minor revisions to questions to enhance the survey's clarity and respondent comprehension. The final survey comprised questions to gather information about the pet or pets involved in the recorded interaction, including nutrition information such as specifics of the main diet (type of food, wet or dry, amount per feeding, and frequency), treats (type of treat, amount per feeding, and frequency), and supplements (type of supplement, amount per treatment, and frequency), and an open-ended question related to anything else the pet ate regularly. Information was also collected on the pet's body weight and owner-identified body condition, owner feeding behaviors (who fed the pet, availability of food, number of meals per day, and source for determining the amount fed), pet's eating behavior (attitude toward food changes, stealing food from other pets, and unmonitored sources for food), and pet's physical activity (nature of activities, frequency of walking, frequency of play with people, and changes in activity level over the past year).b A copy of the survey is available upon request from the corresponding author (JBC).
When a participant's appointment with the attending veterinarian was completed, 1 of the 2 research assistants responsible for recruitment intercepted the participant exiting the examination room and invited the individual to complete a paper version of the postinteraction diet history survey prior to leaving. Data were taken from the survey for the same pet for which the diet history question had been answered. If clients expressed time restrictions, they were offered the option of providing an email address so that the research assistant could send them an email message within 24 hours of the visit, providing a unique identification (participant) number and a link to the online version of the diet history survey managed through an electronic survey program.c
Transcription and coding of video recordings
Participants' video-recorded responses to the randomly assigned diet history questions were transcribed verbatim by 1 research assistant. The diet history question posed to the participant was omitted from the transcript so that coders of participants' answers would be blinded to the particular question asked. All verbatim transcripts were subsequently reviewed for accuracy by one of the authors (REO) while listening to each video recording. At the time of transcription, the duration of the participant's answer to the randomly assigned diet history question was also measured with the media player.d The duration of a participant's response was determined by measuring the time in seconds to the hundredth of a second from the beginning of the client's verbal response to the question until the end of the last word of their completed response.
A coding form managed in the survey platform softwarec and a codebook were developed to code the transcripts of participants' responses with regard to the number, type, and specific details of diet-related items elicited by the 3 different diet history questions. The coding form and codebook were developed on the basis of preexisting diet history forms11,12 and findings from previous research.6 The coding framework was pretested by 3 of the authors (JBC, REO, and CM), who independently coded 10 participant transcripts. Coding for each transcript was then discussed by the 3 authors as a group; inconsistencies in coding were clarified in the codebook to ensure consensus, and additional codes were added to the coding framework to improve the comprehensiveness of the coding.
The diet-related item categories for the 3 nutrition-related questions were defined as main diet, treats, human foods, and medications or dietary supplements. For the main diet, the types of information coded included package description, brand implied, brand name given, food form (eg, wet vs dry), targeted life stage or health concern, owner's reasons for feeding an item, ingredients, food intended for another species, quantity fed and frequency of feeding, free-choice feeding, raw dietary items, home-prepared dietary items, and mention of water. The same types of information were coded for treats, except that free-choice feeding and mention of water did not apply to this category and a code for description of the treat was added. The information types coded for human foods included single-ingredient items, multiple-ingredient items, quantity given, and frequency given. For medications or dietary supplements, the information types coded included brand name given, targeted health concern, owner reasons for administration, form (eg, tablet, liquid), quantity administered, and frequency of administration.
Two individuals who were blinded to the study protocol, to the particular question asked of participants, and to the participants' responses to the diet history survey were enlisted to code the verbatim transcripts of participants' responses by use of the established coding form and definitions provided in the codebook. For training purposes, one of the authors (REO) explained the codebook definitions to both coders and provided examples for each code. One of the coders had prior experience working in a veterinary clinic as a veterinary assistant, and the other was a licensed small animal veterinarian. Initially, these 2 individuals coded 5 transcripts individually and subsequently met to compare coding and to discuss disagreements. The coders remained unaware of the diet history question that led to the participant's response recorded on the transcripts. Once the coders were familiar with the codebook, they individually coded 10 transcripts at a time, each time followed by a meeting to go over coding and discuss discrepancies, until all transcripts were coded. Discrepancies were resolved by consensus between the coders. If an agreement could not be reached or there was uncertainty as to how an item should be coded, the opinion of one of the authors (JBC or REO) was sought.
Using the same coding form and codebook established for coding the verbatim transcripts, one of the authors (REO) and the veterinarian who assisted with coding of the transcripts each independently coded participants' responses to the postinteraction diet history survey. The 2 coders followed the same method of coding and process for addressing discrepancies that were used for coding the verbatim transcripts.
Statistical analysis
For responses to each diet history question and for written survey responses, summary variables were created for each participant to represent the total number of main diet items disclosed, number of treat items disclosed, number of human food items disclosed, and number of medication or dietary supplement items disclosed. In addition, a variable representing all diet-related items disclosed by participants was created by adding the number of main diet items, treat items, human food items, and medication or dietary supplement items together. Descriptive statistics were calculated for demographic information and for response variables of primary interest (ie, sums for the number of main diet items, treat items, human food items, and medication or dietary supplement items; overall items from the video transcripts and from the diet history survey; and duration of the client's verbal response to each diet history question). Frequencies were calculated for categorical variables, and mean, SE, median, minimum and maximum values were determined for continuous variables. Response variables of interest were assessed for normality with the Kolmogorov-Smirnov and Shapiro-Wilk tests and by visual examination of histograms. All analyzed variables were determined to have nonnormal distribution.
Tests for independence (χ2 analyses) were conducted to investigate the association between the diet history questions and the frequency of participants' reports on brand name, diet form (eg, dry vs wet), quantity given, and frequency of feeding for main diet items. First, a comparison across the 3 questions was performed, and Pearson exact P values were reported. If a significant (P < 0.05) association was found, pairwise χ2 tests for independence were conducted that included 2-way comparisons of all 3 questions. Fisher exact 2-tailed P values, ORs, and exact Clopper-Pearson 95% CIs were reported.
Kruskal-Wallis tests were conducted to compare the number of main diet items, treat items, human food items, and medication or dietary supplement items and total number of diet-related items disclosed by participants as well as the duration of participants' response (in seconds) among the 3 diet history questions. If the Kruskal-Wallis tests yielded significant P values, pairwise Mann-Whitney U tests were conducted among the 3 groups. For Kruskal-Wallis and Mann-Whitney U tests, a Monte Carlo simulation of P values (n = 100,000) was used.
Wilcoxon signed rank tests were used to compare the diet-related information gained through the postinteraction diet history survey (ie, number of main diet items, treat items, human food items, and medication or dietary supplement items and the total number of diet-related items) with the information gained with each of the 3 randomly assigned diet history questions individually.
Sums for the total number of main diet items, treat items, human food items, and medication or dietary supplement items identified in the transcripts of the video-recorded interactions and the postinteraction diet history survey were calculated for each interaction in a commercially available spreadsheet.e All other analyses were performed with standard statistical software.f,g Values of P < 0.05 were considered significant.
Results
Study population
Overall, 106 of 165 (64.2%) pet owners approached during the study agreed to participate in the videotaped portion of the study. Data from 7 of these individuals were excluded because of interference by a member of the veterinary team (n = 4) or failure to video record the transaction (3). Therefore, a total of 99 video recordings were transcribed and included in the analysis. Of the 99 participants, 80 (80.8%) completed the participant demographics portion and 77 (77.8%) completed the diet history portion of the postinteraction survey.
Of the 80 participants who completed the postinteraction survey, 48 (60%) completed it in hard-copy form at the end of the appointment, and 32 (40%) completed it online. Eighteen of 80 (23%) respondents were male, and 58 (73%) were female; 4 (5%) did not provide gender information. Age was provided by 73 respondents, with a mean ± SE of 48 ± 1.38 years (median, 48 years; range 26 to 71 years). Twenty-three of 80 (29%) individuals had a bachelor's degree, 17 (21%) indicated a college diploma without further specification, 14 (18%) had a high school diploma or equivalent, and 9 (11%) reported some college or university education. A master's degree (4/80 [5%]), professional degree (eg, MD; 4 [5%]), and vocational or technical school training (2 [3%]) were less commonly reported. One (1%) respondent indicated other education, and none reported having other doctoral (eg, PhD) degrees; 6 (8%) did not answer this question. Reported annual household income was stratified in $20,000 increments. Most individuals reported an annual household income of $100,000 to $119,999 (11/80 [14%]), followed by $140,000 to $159,999 (10 [13%]), $80,000 to $99,999 (9 [11%]), and $40,000 to $59,999 (9 [11%]). The remaining respondents reported incomes of > $160,000 (8/80 [10%]), $60,000 to $79,999 (8 [10%]), $120,000 to $139,999 (6 [8%]), or $20,000 to $39,999 (2 [3%]). Seventeen of 80 (21%) individuals did not answer the question.
Overall, 76 of 99 (76.8%) video-recorded study interactions involved diet-related information about a participant's dog, and 23 (23.2%) involved diet-related information about a participant's cat. Of the 99 participants in the study, 33 (33.3%) were assigned the “Tell me…” question, 40 (40.4%) were assigned the “What kind of food…” (singular) question, and 26 (26.3%) were assigned the “What kind of foods…” (plural) question.
Information types shared during video-recorded interactions
The frequencies of each type of information disclosed by participants regarding main diet items, treats, human foods, and medications or dietary supplements during the 99 video-recorded study interactions were stratified by diet history question and summarized (Tables 1 and 2). Frequency of feeding for main diet items was the most commonly reported detail provided by participants in response to the “Tell me…” question. In contrast, brand name was the most commonly reported information provided by participants in response to the “What kind of food…” and “What kind of foods…” questions.
Descriptive results for disclosure of information pertaining to main diet and treat items fed to pets by 99 participants in a study to examine the effects of 3 diet history questions on diet-related information gathered from pet owners and to assess whether the information obtained with each question in person differed from information obtained with a written diet history survey.
| Question | ||||
|---|---|---|---|---|
| Information type | Item category | 1 (n = 33) | 2 (n = 40) | 3 (n = 26) |
| Description of item | Main diet | — | — | — |
| Treats | 4 (12) | 1 (3) | 0 (0) | |
| Package description | Main diet | 0 (0) | 3 (8) | 3 (12) |
| Treats | 0 (0) | 0 (0) | 0 (0) | |
| Brand implied | Main diet | 5 (15) | 3 (8) | 8 (31) |
| Treats | 1 (3) | 0 (0) | 0 (0) | |
| Brand name given | Main diet | 20 (61) | 33 (83) | 15 (58) |
| Treats | 7 (21) | 0 (0) | 1 (4) | |
| Form (eg, dry vs wet food) | Main diet | 19 (58) | 20 (50) | 4 (15) |
| Treats | 0 (0) | 0 (0) | 0 (0) | |
| Targeted life stage or health concern | Main diet | 11 (33) | 11 (28) | 11 (42) |
| Treats | 0 (0) | 0 (0) | 0 (0) | |
| Owner reasons for providing item | Main diet | 6 (18) | 11 (28) | 7 (27) |
| Treats | 1 (3) | 1 (3) | 1 (4) | |
| Ingredients | Main diet | 7 (21) | 11 (28) | 8 (31) |
| Treats | 1 (3) | 0 (0) | 1 (4) | |
| Different species (food intended for another species) | Main diet | 0 (0) | 0 (0) | 0 (0) |
| Treats | 0 (0) | 0 (0) | 0 (0) | |
| Quantity fed | Main diet | 18 (55) | 2 (5) | 1 (4) |
| Treats | 10 (30) | 1 (3) | 0 (0) | |
| Frequency of feeding | Main diet | 24 (73) | 4 (10) | 1 (4) |
| Treats | 11 (33) | 1 (3) | 0 (0) | |
| Free-choice feeding | Main diet | 3 (9) | 2 (5) | 0 (0) |
| Treats | — | — | — | |
| Raw dietary item | Main diet | 1 (3) | 0 (0) | 0 (0) |
| Treats | 0 (0) | 0 (0) | 0 (0) | |
| Home-prepared dietary item | Main diet | 1 (3) | 0 (0) | 0 (0) |
| Treats | 0 (0) | 0 (0) | 0 (0) | |
| Mention of water | Main diet | 7 (21) | 0 (0) | 0 (0) |
| Treats | — | — | — | |
Each participant was randomly assigned 1 diet history question that was asked by a trained research assistant during a video-recorded interaction that immediately preceded a scheduled veterinary appointment with a dog or cat; data represent the number (%) of study interactions where the information type was provided in response to the respective question. Question 1 was“Tell me everything he [or she] eats throughout a day, starting first thing in the morning right through to the end of the day.”Questions 2 and 3 differed from each other only by use of the singular term food versus the plural term foods; question 2 was“What kind of food does she [or he] eat?”and question 3 was“What kind of foods does he [or she] eat?”
— = Not applicable.
Descriptive results for disclosure of information pertaining to human foods and medications or dietary supplements given to pets by the same 99 dog and cat owners in Table 1 during video-recorded interactions.
| Question | |||
|---|---|---|---|
| Item and information type | 1 (n = 33) | 2 (n = 40) | 3 (n = 26) |
| Human foods | |||
| Single-ingredient item | 14 (42) | 1 (3) | 4 (15) |
| Multiple-ingredient item | 5 (15) | 0 (0) | 0 (0) |
| Quantity given | 8 (24) | 1 (3) | 3 (12) |
| Frequency given | 12 (36) | 1 (3) | 4 (15) |
| Medications or dietary supplements | |||
| Brand name given | 0 (0) | 0 (0) | 1 (4) |
| Targeted health concern | 3 (9%) | 0 (0) | 1 (4) |
| Owner reasons for administration | 0 (0) | 0 (0) | 1 (4) |
| Form (eg, liquid vs tablet) | 5 (15) | 0 (0) | 0 (0) |
| Quantity administered | 3 (9) | 0 (0) | 0 (0) |
| Frequency of administration | 5 (15) | 1 (3) | 0 (0) |
See Table 1 for key.
The frequency of reporting the brand name for main diet items was not significantly associated with the diet history question asked. However, there was a significant (P = 0.009) association between the question asked and the reporting of diet form (eg, dry vs wet food). The odds of a participant reporting the diet form were significantly (P = 0.004) greater when the “Tell me…” question was asked (OR, 5.7 [95% CI, 1.5 to 23.5]) than when the “What kind of foods…” question was asked. Furthermore, the odds of a participant providing this type of information were significantly (P = 0.02) greater when the “What kind of food…” question was asked than when the “What kind of foods…” question was asked (OR, 4.2 [95% CI, 1.2 to 16.8]).
The diet history question asked was significantly (P < 0.001) associated with the reporting of the quantity fed for main diet items. Participants who were asked the “Tell me…” question had greater odds of providing this information than did those who were asked the “What kind of food…” question (OR, 22.8 [95% CI, 4.4 to 217.4]; P < 0.001) or the “What kind of foods…” question (OR, 30.0 [95% CI, 3.8 to 1,250.0]; P < 0.001). Similarly, a significant (P < 0.001) association was detected between the diet history question asked and the reporting of daily feeding frequencies for main diet items. Participants who were asked the “Tell me…” question had greater odds of providing feeding frequency information, compared with those who were asked the “What kind of food…” (OR, 24.0 [95% CI, 5.9 to 113.6]; P < 0.001) or “What kind of foods…” (OR, 66.7 [95% CI, 7.9 to 2,500.0]; P < 0.001) question.
Number of items disclosed during video-recorded interactions
Descriptive statistics for the number of main diet items, treat items, human foods, and medications or dietary supplements disclosed by participants during the video-recorded study interactions are provided (Table 3). The number of items disclosed in the main diet and treat categories differed significantly (P < 0.001 for both comparisons) between participants asked the “Tell me…” question and those asked the “What kind of food” or “What kind of foods” questions. The number of human food items disclosed by participants differed significantly (P < 0.001 for all comparisons) among the 3 diet history questions, but no significant difference was detected in the number of medications or dietary supplements disclosed.
Proportions of participants who disclosed ≥ 1 item and comparison of the number of main diet, treat, human food, and medication or dietary supplement items disclosed during video-recorded interactions for the 99 dog and cat owners in Table 1.
| Question | |||
|---|---|---|---|
| Variable | 1 (n = 33) | 2 (n = 40) | 3 (n = 26) |
| Main diet | |||
| Participants who disclosed ≥ 1 item (No. [%]) | 33 (100) | 40 (100) | 26 (100) |
| No. of items (mean ± SE) | 1.55 ± 0.12 | 1.15 ± 0.06 | 1.04 ± 0.04 |
| No. of items (median [range]) | 1 (1–3)A | 1 (1–2)B | 1 (1–2)B |
| Treat items | |||
| Participants who disclosed ≥ 1 item (No. [%]) | 13 (39) | 1 (3) | 3 (12) |
| No. of items (mean ± SE) | 0.52 ± 0.13 | 0.05 ± 0.05 | 0.08 ± 0.05 |
| No. of items (median [range]) | 0 (0–3)A | 0 (0–2)B | 0 (0–1)B |
| Human foods | |||
| Participants who disclosed ≥ 1 item (No. [%]) | 18 (55) | 1 (3) | 6 (23) |
| No. of items (mean ± SE) | 0.97 ± 0.21 | 0.08 ± 0.08 | 0.35 ± 0.18 |
| No. of items (median [range]) | 1 (0–4)A | 0 (0–3)B | 0 (0–4)C |
| Medications or supplements | |||
| Participants who disclosed ≥ 1 item (No. [%]) | 6 (18) | 1 (3) | 3 (12) |
| No. of items (mean ± SE) | 0.48 ± 0.23 | 0.03 ± 0.03 | 0.15 ± 0.09 |
| No. of items (median [range]) | 0 (0–5)A | 0 (0–1)A | 0 (0–2)A |
| Total diet-related items | |||
| No. of items (mean ± SE) | 3.52 ± 0.37 | 1.30 ± 0.14 | 1.77 ± 0.21 |
| No. of items (median [range]) | 3 (1–9)A | 1 (1–6)B | 1 (1–5)B |
Mean ± SE and median (range) data represent the number of items in the specified category disclosed by participants in response to the respective question.
Within a row for a given item category, results with different superscripted letters are significantly (P < 0.001) different among questions.
See Table 1 for remainder of key.
The total number of diet-related items (ie, the combined number of main-diet items, treat items, human food items, and medications or dietary supplements) disclosed by participants asked the “Tell me…” question was significantly (P < 0.001 for both comparisons) greater than the number disclosed by those asked the “What kind of food…” or “What kind of foods…” question. The total number of items disclosed did not differ between participants asked the “What kind of food…” and “What kind of foods…” questions.
Response duration during video-recorded interactions
The duration of participants' responses differed significantly (P < 0.05 for all comparisons) among the 3 questions. Individuals asked the “Tell me…” question spent more time responding (mean ± SE, 43.57 ± 5.62 seconds; median, 33.23 seconds [range, 9.93 to 167.82 seconds]) than those asked the “What kind of food…” (mean ± SE, 15.04 ± 1.74 seconds; median, 11.20 seconds [range, 1.98 to 42.88 seconds]) or “What kind of foods…” (mean ± SE, 27.01 ± 5.47 seconds; median, 20.75 seconds [range, 3.78 to 153.13 seconds]) question. Individuals asked the “What kind of foods…” question spent more time responding than those asked the “What kind of food…” question.
Comparison of information provided in response to verbal diet history questions and information obtained with the postinteraction diet history survey
Of the 33 participants assigned the “Tell me…” question, 21 (64%) completed the postinteraction diet history survey regarding the same pet (Table 4). Comparison of the quantity of diet-related information disclosed by participants during the video-recorded study interaction with the quantity of diet-related information disclosed by participants on the postinteraction diet history survey revealed that the number of treat items disclosed on the written survey was significantly (P = 0.02) greater than the number disclosed in the video-recorded interaction. No significant difference was detected for any other item category.
Comparison of diet-related information obtained in response to each of 3 verbal diet history questions during video-recorded interactions with information of the same type obtained with written postinteraction diet history surveys for 77 of the 99 pet owners in Table 1 who provided information by both methods.
| Question 1 (n = 21) | Question 2 (n = 34) | Question 3 (n = 22) | ||||
|---|---|---|---|---|---|---|
| Variable | VRI | Survey | VRI | Survey | VRI | Survey |
| Main diet | ||||||
| Participants who disclosed ≥ 1 item (No. [%]) | 21 (100) | 21 (100) | 34 (100) | 34 (100) | 22 (100) | 22 (100) |
| No. of items (mean ± SE) | 1.48 ± 0.13 | 1.29 ± 0.10 | 1.12 ± 0.06 | 1.32 ± 0.09 | 1.05 ± 0.05 | 1.14 ± 0.08 |
| No. of items (median [range]) | 1 (1–3) | 1 (1–2) | 1 (1–2) | 1 (1–3)* | 1 (1–2) | 1 (1–2) |
| Treat items | ||||||
| Participants who disclosed ≥ 1 item (No. [%]) | 9 (43) | 16 (76) | 1 (3) | 30 (88) | 2 (9) | 17 (77) |
| No. of items (mean ± SE) | 0.62 ± 0.19 | 1.19 ± 0.26 | 0.06 ± 0.06 | 1.32 ± 0.15 | 0.09 ± 0.06 | 1.55 ± 0.29 |
| No. of items (median [range]) | 0 (0–3) | 1 (0–5)* | 0 (0–2) | 1 (0–4)* | 0 (0–1) | 1.5 (0–6)* |
| Human foods | ||||||
| Participants who disclosed ≥ 1 item (No. [%]) | 11 (52) | 13 (62) | 1 (3) | 11 (32) | 6 (27) | 8 (36) |
| No. of items (mean ± SE) | 0.95 ± 0.26 | 0.95 ± 0.2 | 0.09 ± 0.09 | 0.5 ± 0.16 | 0.55 ± 0.24 | 0.64 ± 0.22 |
| No. of items (median [range]) | 1 (0–4) | 1 (0–3) | 0 (0–3) | 0 (0–4)* | 0 (0–4) | 0 (0–4) |
| Medications or supplements | ||||||
| Participants who disclosed ≥ 1 item (No. [%]) | 5 (24) | 6 (29) | 1 (3) | 3 (9) | 3 (14) | 9 (41) |
| No. of items (mean ± SE) | 0.71 ± 0.35 | 0.71 ± 0.33 | 0.03 ± 0.03 | 0.09 ± 0.05 | 0.18 ± 0.11 | 0.73 ± 0.24 |
| No. of items (median [range]) | 0 (0–5) | 0 (0–5) | 0 (0–1) | 0 (0–1) | 0 (0–2) | 0 (0–4)* |
| All diet-related items | ||||||
| No. of items (mean ± SE) | 3.76 ± 0.55 | 4.14 ± 0.51 | 1.29 ± 0.16 | 3.24 ± 0.25 | 1.86 ± 0.28 | 4.05 ± 0.56 |
| No. of items (median [range]) | 3 (1–9) | 4 (1–10) | 1 (1–6) | 3 (1–7)* | 1 (1–5) | 3.5 (1–9)* |
Of the 40 individuals assigned the “What kind of food…” question, 34 (85%) completed the postinteraction diet history survey for the same pet (Table 4). The number of main diet items (P = 0.02), treat items (P < 0.001), and human food items (P = 0.002) and the total number of diet-related items (P < 0.001) disclosed on the written survey were each significantly greater than the number of items in the same category disclosed in the video-recorded interaction.
Of the 26 individuals assigned the “What kind of foods…” question, 22 (85%) completed the postinteraction diet history survey for the same pet. The number of treat items (P < 0.001), number of medications or dietary supplements (P = 0.02), and total number of diet-related items (P = 0.001) disclosed on the written survey were each significantly greater than the number of items in the same category disclosed in the video-recorded interaction.
Discussion
Results of the present study revealed the effects that different designs of a diet history question can have on the diet information gained from a pet owner. Given that a comprehensive diet history is a central part of a comprehensive nutritional assessment, it is important for veterinary professionals to consider both the content and the process of communication about nutrition. The present study found that the process of communication with pet owners (eg, the use of closed-ended vs open-ended questions) can influence the content of communication (eg, information about the amount and form of food given to the pet) gathered from pet owners.
In the present study, the open-ended or telling form of question (“Tell me everything he [or she] eats throughout a day, starting first thing in the morning right through to the end of the day”) elicited disclosure of a significantly greater number of diet-related items than the closed-ended, more specifying form of question (“What kind of food does she [or he] eat?”) or the modified version of this question (“What kind of foods does he [or she] eat?”) in which the plural term foods was used. Given the broad role nutrition has in maintaining and managing health, ascertaining a detailed diet history from owners is critical to diagnostic and therapeutic success in veterinary medicine.
Further supporting the value of telling-form questions in gathering a comprehensive diet history, the telling form of question in the present study resulted in disclosure of a broader range of diet-related information from participating clients than did both specifying-form questions, including the main diet form, quantity of main diet items fed, frequency of feeding, treats fed, and human foods fed. This suggested that, to accomplish detailed diet histories, veterinary professionals should begin diet history inquiries with broad questions or statements that invite expansion from clients, subsequently reserving the use of more targeted specifying questions to fill in information gaps when specific, concise pieces of information have yet to be obtained.
It was previously proposed that by simply minimizing the specificity of the “What kind of food…” question by employing the plural term foods, comparatively more diet-related information might be obtained.6 When this modification was used in the present study, there was little statistical difference in the information gathered between the 2 what-prefaced questions, except that the question containing the plural term foods elicited disclosure of more human food items than did the question containing the singular term food.
The use of telling-form questions, also commonly referred to as open-ended questions or open-ended inquiry in the literature,9 to initiate nutritional information gathering allows the veterinary professional to explore the area of nutrition without unintentionally shaping or focusing the client's response from the onset. When practitioners consider how to formulate their own telling-form questions, it is important to recognize that the telling-form question used in the present study had 2 key structural components. First, the statement invited clients to share everything the pet ate throughout a day, highlighting an interest in eliciting a broad narrative about the animal's dietary intake. Second, the statement requested a temporally structured response, which likely aided a client's recall of the requested information by requesting that the client account for items consumed from the start of the morning until the end of the day.
Participating clients in the present study spent significantly more time in providing their response to the “Tell me…” and “What kind of foods…” questions, compared with the time spent answering the narrower specifying “What kind of food…” question. Time pressures have been cited as presenting a communication challenge for veterinarians in practice13; therefore, the extra time taken by clients to respond to open-ended or broader inquiries is likely to be an important practical concern for veterinary professionals. It is important to recognize that, although clients in the present study spent median times of 22 and 12 seconds longer to respond to the “Tell me…” question than to the “What kind of food…” and “What kind of foods…” questions, respectively, this inquiry elicited reporting of significantly more main diet, treat, and human food items and in many cases a broader scope of information than did the other 2 questions. Although not assessed in the present study, the authors believe that it would take an equal or greater amount of time to obtain a similar amount and breadth of information through the use of a series of specifying or closed-ended questions. Communication best practices for information gathering include the use of an open-to-closed cone interviewing approach to enhance information-gathering effciency.9 This involves structuring the gathering of information in a way that starts with broad exploratory open-ended inquiry, followed by more focused open-ended inquiry to explore areas of subsequent interest and an eventual move to closed-ended inquiry to elicit important remaining details that are often of a very specific nature.
By investing in a comprehensive diet history, veterinary practitioners may be better positioned to save time by gaining an up-front understanding of the challenges or barriers a client may have to a potential nutrition recommendation. Research suggests that client resistance to veterinarian-initiated proposals for long-term dietary change often stems from diet-related or proposal-relevant factors that are revealed by clients after the veterinarian's recommendation has been made.h This information could be gained in advance of making a dietary recommendation by obtaining a detailed diet history. This approach also sets the foundation for shared decision-making, which has been described as a process in which the veterinarian and client participate jointly in the recommendation or proposal phase of the interview.13 By sharing in the decision-making process, both the veterinarian and client may develop greater investment in the final decision, which could engender greater commitment from the client in carrying out the nutrition recommendation or proposal.
The use of a written diet history form to aid the gathering of diet-related information from clients has been advocated in the veterinary literature.1–3 It has been suggested that having clients complete the form in advance of an appointment may assist veterinary professionals in managing the time challenges that can be associated with gathering a comprehensive diet history.3 In the present study, although the written diet history survey captured significantly more information in several diet-related categories and significantly more diet-related information overall than did the 2 what-prefaced questions, there was little statistical difference found between the amount of information obtained with the diet history form and that received in response to the telling-form question. The only area where the postinteraction diet history survey gathered significantly more information from participants than the telling-form question was in relation to treat items. This is important, as it may highlight that gathering information on treats is a sensitive area of diet history taking that specifically requires direct probing. It has been suggested that clients are likely to omit diet items that they feel sensitive about.6 Alternatively, it may also be that some treats are not provided on a daily basis but are instead given occasionally or regularly over longer periods of time, so that a question inviting the reporting of food intake over the course of a single day may fail to elicit such information. Both of these possibilities further support the use of an open-to-closed cone questioning approach to diet history gathering, recognizing there are specific areas that will warrant further probing and exploration by the use of more focused open-ended inquiries. Regardless of the initial approach to gathering a diet history, veterinary professionals should be encouraged to further probe and explore important patient-specific, diet-related, feeding management, and environmental information that owners may have believed was not relevant, failed to recall, or felt embarrassed to raise in their initial responses.
The present study was restricted to the responses of a small sample of pet owners accessing 4 companion animal practices in Ontario, Canada. In addition, the 3 diet history questions were posed to participants in advance of their interaction with the veterinarian, and this approach may have introduced contexts or nuances that influenced participants' responses. To minimize this impact, the research assistant received training to standardize the approach to participants, and each participant was randomly assigned to 1 of the 3 diet history questions. In addition, the study was carried out in veterinary clinics where clients were bringing their pets for scheduled appointments to produce a more natural clinic-like context for the interaction. In the analysis performed to compare the diet-related information gained by each of the 3 study questions with that obtained through the postinteraction diet history survey, it was possible that significant differences were not identified owing to the small number of participants included in each subanalysis. Further, the amount of time taken by a client to complete the diet history survey and time needed for the veterinarian to read this information were not examined. This limited the ability of the present study to assess the efficiency of this approach to gather a comprehensive diet history. This is a potential area for future research to inform evidence-based approaches to managing the time challenges veterinarians can experience. The present study only examined clients' responses to an initial diet history question and did not explore the extent of diet information that could be gained from clients with further exploratory and probing questions. Examining the effectiveness of follow-up diet history questions and probes in the context of a veterinary clinical encounter is an area for future research consideration.
Results of the study reported here indicated that question design has an influence on the yield of diet history information obtained from dog and cat owners and supported the use of open-ended questions to initiate gathering a comprehensive diet history for veterinary patients. In accordance with communication best practices,9 gathering diet history information from clients in a way that starts with broad exploratory open-ended inquiry, such as use of the telling question of the present study, and then transitioning to more focused open-ended inquiry to explore areas of interest with an eventual move to closed-ended inquiry to elicit any remaining details of importance, may optimize the information gained.
Acknowledgments
Funding for the study was provided by a grant from Royal Canin Canada. The funding source did not have any involvement in the study design, data analysis and interpretation, or writing and publication of the manuscript.
Dr. Coe regularly receives honoraria and research funding from various commercial pet food companies. Dr. Verbrugghe is the Royal Canin Veterinary Diets Endowed Chair in Canine and Feline Clinical Nutrition at the Ontario Veterinary College, University of Guelph. Dr. Verbrugghe also serves on industry-related scientific advisory boards and has received research grants in association with various pet food companies.
Presented in abstract form at the 8th International Conference on Communication in Veterinary Medicine, Hockley Valley, ON, Canada, March 2018.
The authors thank Kim Lambert, Janet Cutler, Caroline Graefin von Waldburg-Zeil, and Helen Toner for research assistance and William Sears for statistical support.
ABBREVIATIONS
| CI | Confidence interval |
Footnotes
Research Randomizer 4.0, Geoffrey C. Urbaniak GC and Scott Plous. Available at: www.randomizer.org. Accessed Jun 25, 2015.
A copy of the survey is available from the corresponding author upon request.
Qualtrics, Provo, Utah.
VLC media player software, version 2.2.0. VideoLan Organization, Paris, France.
Excel, version 14.3.7, Microsoft Corp, Redmond, Wash.
SAS, version 9.3, SAS Institute Inc, Cary, NC.
SPSS, version 22.0, IBM Corp, Armonk, NY.
MacMartin C, Wheat HC, Coe JB. Veterinarian-initiated long-term dietary recommendations: practitioners' management of clients' responses (poster presentation). Waltham International Nutr Sci Symp, Portland, Ore, October 2013.
References
1. Baldwin K, Bartges J, Buffington T, et al. AAHA nutritional assessment guidelines for dogs and cats. J Am Anim Hosp Assoc 2010;46:285–296.
2. WSAVA Nutritional Assessment Guidelines Task Force, Freeman L, Becvarova I, et al. WSAVA nutritional guidelines. J Small Anim Pract 2011;52:385–396.
3. Michel KE. Using a diet history to improve adherence to dietary recommendations. Compend Contin Educ Vet 2009;31:22–24.
4. Laflamme DP. Companion animals symposium: obesity in dogs and cats: what is wrong with being fat? J Anim Sci 2012;90:1653–1662.
5. Verlinden A, Hesta M, Millet S, et al. Food allergy in dogs and cats: A review. Crit Rev Food Sci Nutr 2006;46:259–273.
6. MacMartin C, Wheat HC, Coe JB, et al. Effect of question design on dietary information solicited during veterinarian-client interactions in companion animal practice in Ontario, Canada. J Am Vet Med Assoc 2015;246:1203–1214.
7. MacMartin C, Wheat HC, Coe JB, et al. Conversation analysis of veterinarians' proposals for long-term dietary change in companion animal practice in Ontario, Canada. J Vet Med Educ 2018;45:1–20.
8. Bartges JW, Kirk CA, Lauten SD. Nutrition in disease. In: Morgan RV, ed. Handbook of small animal practice. 5th ed. St Louis: Saunders-Elsevier, 2008;1176–1186.
9. Adams CL, Kurtz SA. Skills for communicating in veterinary medicine. Parsippany, NJ: Dewpoint Publishing, 2017;90–100.
10. Tapsell LC, Brenninger V, Barnard J. Applying conversation analysis to foster accurate reporting in the diet history interview. J Am Diet Assoc 2000;100:818–824.
11. Ontario Veterinary College Clinical Nutrition Service. Diet history form for pet owners. Available at: ovc.uoguelph.ca/hsc/en/resources/451244Nov2013ClinicalNutritionDietHistoryForm.pdf. Accessed May 1, 2015.
12. WSAVA Global Nutrition Committee. Short diet history form. Available at: wsava.org/sites/default/files/Diet%20History%20%20Form.pdf. Accessed May 1, 2015.
13. Coe JB, Adams CL, Bonnett BN. A focus group study of veterinarians' and pet owners' perceptions of veterinarian-client communication in companion animal practice. J Am Vet Med Assoc 2008;233:1072–1080.
