All conversations were coded according to the manual for MITI 4.2.1 at MIC Lab AB, a Swedish center of expertise for quality assurance, coding, and education in MI. Reliability of the coders is verified through regular controls. To minimize inter-rater bias, all conversations were coded by the same coder. In addition to the coding, written feedback from the coder was received from the coder for each conversation, with suggestions for improvements from an MI-perspective.

In total, eight veterinarians were recruited to the study through convenience sampling, seven women and one man with an average age of 41 years, ranging between 31 and 52 years. All participants graduated as veterinarians during the period 2000-2018 and had been working as clinical small animal veterinarians for 10 years on average, ranging between 3 and 20 years. Three of the veterinarians had not participated in any further education in dentistry, while the remaining participants had completed at least one course within veterinary dentistry after completing their veterinary education. None of the veterinarians had participated in any form of communication training.

Each veterinarian participated in one telephone conversation with a standardized client played by the same actor (male) employed by MIC Lab AB. The actor was trained for this purpose and had experience in role-play scenarios using MI. The actor did not use a script but was provided with a client profile with background information about the dog owner and the dog. The client profile was developed based on literature studies and clinical experience, and in collaboration with Lars Forsberg at MIC Lab AB. In short, the scenario was: A 6-year-old female cocker spaniel visited the veterinary clinic for professional dental cleaning and was diagnosed with periodontitis. Several teeth had been deemed unsalvageable because of severe periodontitis and extracted. To stop disease progression, implementation of daily tooth brushing is now necessary. The dog owner's willingness to start brushing the dog's teeth was scored nine on a ten-point scale, and his confidence in his ability to succeed was scored two on a ten-point scale.

General information about the nature of the study, a communication study with an actor playing a dog owner, and background information about the scenario, was sent in advance to participating veterinarians via e-mail.

Before each conversation, the actor telephoned the veterinarian to confirm that the veterinarians had received information about the study, that the call would be recorded, and that they consented to participate in the study. The actor then telephoned again as the fictional dog owner and recorded the call using a digital voice recorder. The calls were directed by the actor to be ~20 min long to allow recommended coding time. After each conversation, the actor uploaded the audio file to the website of MIC Lab AB, from where the MITI-coder could access them.

For each recorded call, a coding protocol was obtained from MIC Lab AB. Based on the obtained coding protocols mean values ± standard deviation for the global scores and behavior counts were calculated. The coder's written comments were compiled by the author E.J. and presented in text.

Descriptive statistics (mean ± standard deviation) of the global scores and behavior counts are presented in Table 1. The global scores were ≤ 3 (reference default value) in all coding protocols. Cultivating Change Talk was scored lowest possible (score one) in all conversations. The Softening Sustain Talk scores were either two (6/8) or three (2/8). The Partnership scores were also either two (6/8) or three (2/8). Empathy was scored two in all (8/8) coding protocols.

The veterinarians' consultation approaches were predominantly characterized by Giving information, Persuasion, and Questions (Table 1). On a few occasions, the veterinarians used Persuasion with Permission. All veterinarians made at least one Simple or Complex reflection in response to client statements. The participants made few attempts to Affirm the client, and few were Seeking Collaboration. None of the veterinarians were Emphasizing Autonomy during the calls. The behavior Confront was not shown (Table 1). Overall, the results were very similar for all participating veterinarians.

In the written comments, the coder noted both the clinical strengths of the veterinarians and potential areas for improvement in terms of MI, by commenting how behaviors in the behavior count affected the global scores.

The coder identified that most of the veterinarians were asking Questions trying to understand the client's situation and thoughts, positively affecting the Empathy score. Actual examples include:

The coder noted that Reflections, such as “Just as you said, daily tooth brushing…” also positively affected the Empathy score. Another clinical strength was that the veterinarian in a few situations Affirmed the dog owner by acknowledging the client's achievements, which positively affected the Partnership score.

The coder also suggested areas for improvement from an MI-perspective. Regarding Cultivating Change Talk, the score would have been higher if the veterinarians would have asked the client about the reasons why the change was needed. This was noted in 8 of 8 conversations.

The Partnership scores were adversely affected by the veterinarians (8/8) giving advice without the client's permission (Persuade). Additionally, all the veterinarians (8/8) dominated the conversations having most of the speaking time. The Partnership scores would have been positively affected if the veterinarians had involved the client more in the conversation and Sought Collaboration to a greater extent. In more than half of the protocols (6/8), the coder also mentioned that the veterinarian could have Emphasized Autonomy to strengthen the Partnership with the client.

Another comment noted by the coder regarded the behavior Affirm. To achieve a higher score in Partnership, the veterinarians could have affirmed the client when expressing ambitions to make a change and acknowledged the client's efforts and achievements. To achieve a higher score in Empathy, the veterinarians could have made more Simple or Complex reflections on the client's utterances.

To the authors' knowledge, this is the very first time that the degree of spontaneous use of MI in small animal veterinarians without prior training has been investigated.

In the present study, the global scores were consistently low (≤ 3). Cultivating Change Talk was scored the lowest possible in all conversations, which implies that the veterinarians showed no explicit attention to, or preference for, the client's language in favor of changing according to the MITI 4.2.1 manual (32). Regarding Softening Sustain Talk, the majority (6/8) of the veterinarians unfortunately mainly chose to explore, focus on, or respond to the client's language in favor of the status quo. However, some (2/8) of the participating veterinarians made attempts to shift focus away from the sustain talk. The scores would likely increase if the veterinarians became more aware of sustain talk and how to respond in favor of change.

Partnership and Empathy have been identified as critical elements of a relationship-centered consultation approach (10). In addition, partnership and shared decision making (i.e., healthcare professional together with client reaching a decision about care) in veterinary care is favored by both veterinarians and pet owners according to recent research (33, 34). However, the veterinarians in the present study predominantly superficially responded to opportunities to collaborate, resulting in low Partnership scores. Although some veterinarians (2/8) incorporated the client's contributions to some extent. Moreover, Empathy was equally scored two in all coding protocols meaning that the veterinarians only made sporadic efforts to explore the client's perspective. Since low advisor empathy has been identified as damaging to the advisor relationship and associated with poorer patient outcomes (35), the authors suggest empathy as an area for improvement.

Based on the behavior counts, the conversations were dominated by the veterinarians relying predominantly on Giving information, Persuasion, and Questions in their communication with the dog owner. According to the MITI 4.2.1 manual, this implies that the veterinarians focused primarily on providing information and educating the client, making overt attempts to change the client's attitude, opinions, or behavior, and eliciting information by asking questions (32). Occasionally, the veterinarians used Persuasion with Permission, which means that they included an emphasis on collaboration or autonomy support while persuading. This behavior is considered an MI-adherent behavior, in contrast to Persuasion. These results are in accordance with the findings of previous studies (4, 11, 12, 19) and suggest that a paternalistic communication style was adopted by the participating veterinarians taking on the role of an expert paid to provide a service of advice and solutions. In terms of health related problem appointments, a paternalistic approach was used by the veterinarian in 85% of the cases, and by 54% of wellness appointments (4). The results presented in the present study may indicate that the paternalistic communication style is frequently used also among Swedish small animal veterinarians, even though this persuasive approach has been proven to be ineffective and also increases resistance to change in ambivalent clients (6). Persuasion is more likely to elicit reactions against, rather than in favor of, change (36). This phenomenon is known as psychological reactance and has been a frequent subject for research since the reactance theory was first described by Brehm (37). In addition, studies in medical communication have shown a positive association between the use of relationship-centered care and aspects of clinical outcomes, such as patient satisfaction, patient health outcomes, physician satisfaction, and reduction of malpractice complaints (4).

The veterinarians in the study relied heavily on Questions, but only one of the questions that the coder highlighted was an open question. The remaining were closed questions with limited answer options such as “yes,” “no,” “maybe,” or “do not know.” In contrast to closed questions, open questions stimulate reflection and exploration in favor of change (10). By reformulating the questions into open questions, there is a potential to evoke change talk, strengthen partnerships, and thus, promote behavior change.

In general, the core skills of MI are not considered to be regularly used in professional conversations (10), which was also true in the present study. The veterinarians practiced reflective listening to a very small extent. However, all veterinarians made at least one Simple or Complex reflection in response to client statements. Likewise, the veterinarians made few attempts to Affirm the client by accentuating something genuinely positive about the client's strengths, efforts, intentions, or worth (32). These results suggest that there is potential for improvement regarding several behaviors. Encouragingly, all these behaviors can be improved by communication training.

Furthermore, few of the veterinarians were Seeking Collaboration, which implies making attempts to share power or acknowledge the expertise of the client (32). Without seeking collaboration, the veterinarian controls the conversation supposing and communicating to have the best solutions to the client's problems, resulting in unequal power distribution between veterinarian and client. Such an approach has been shown to be ineffective and to increase resistance to change in ambivalent clients (6). Contrarily, a more client-centered communication style may be beneficial to achieve behavior change (14). According to research guided by the self-determination theory (SDT), autonomy is one of three innate psychological needs, which when satisfied, enhances self-motivation, and when thwarted leads to diminished motivation (38). When comparing people whose motivation is authentic with those who are simply externally controlled for action, the former, relative to the latter, typically have more interest, excitement, and confidence. This in turn leads to enhanced performance, persistence, and creativity. However, none of the veterinarians in the present study were Emphasizing Autonomy by making utterances that highlight the client's freedom of choice and right to make his own decisions about his dog. This may lead to decreased client motivation and shows that there is room for improvement in veterinary communication.

Similar to the present study, a predominance of MI-nonadherent behaviors has been reported in studies of food safety by health and environmental inspectors (39) as well as cattle veterinarians (12, 19). Thus, MI-training is required to be able to apply the method in professional counseling (10). Studies have shown that considerable time is required to learn MI (10, 39). To read or hear about the method is seldom enough; to develop MI-skills, practical training, including feedback, is needed. However, communication training seems to be in demand among veterinarians. In a survey among veterinary practitioners in the United Kingdom and the United States in 2012/2013, 40% answered that they would be interested in further veterinary communication skills training, with the preferred methods being simulated consultations and online training (40). In a recently published study, Svensson et al. (18) evaluated a 6-month training program in MI for veterinarians involved in veterinary herd health management. After completing the training program, all participating veterinarians had significantly improved their MI-skills, regarding at least one parameter. Veterinarians with higher MI-skills have in turn been associated with increased expression of so-called change talk from the client (21). Moreover, the participating veterinarians perceived their new skills and knowledge of MI as highly relevant in their work (18). Likely, MI-training of small animal veterinarians would have similar effects.

Encouragingly, no veterinarians showed Confronting behaviors such as disagreeing, arguing, shaming, blaming, criticizing, moralizing, or warning the client which are considered MI-nonadherent behaviors (32), and could be very counterproductive in terms of behavior change.

The predominance of Giving information, Persuasion, and Questions, characteristic of paternalism, in the consultation approach among veterinarians, might create psychological resistance, and may be contributing to the low uptake of dental home care recommendations reported. One previous study showed that adherence to dental homecare instruction in dogs that had undergone periodontal surgery was low, with only 24% of dog owners still brushing daily after 6–20 months post-op (41). In addition, we have in a recent survey of preventative dental home care in dogs in Sweden shown that <4% of Swedish dog owners brushed their dog's teeth daily, and that adherence to veterinary recommendation about tooth brushing was low. Nevertheless, the majority of the dog owners answered that they might consider brushing, indicating a great potential motivation to perform dental home care (31). In human dentistry, MI is recommended in the national guidelines as a method for effective dental care communication (17). Similarly, the implementation of MI in veterinary practice may be a tool to improve adherence regarding tooth brushing in dogs.

The field of application for MI in veterinary practice is extensive. MI may be applied in every situation where a change in the behavior of the animal owner is desirable, such as weight loss, medication, and rehabilitation, as well as dental home care. We suggest that MI may be helpful in increasing adherence to veterinary recommendations in all these fields.

In the present study, role-play scenarios with a professional actor were used instead of real clinical situations to standardize the degree of difficulty and conditions for practicing MI as far as possible. By using role-play scenarios with the same actor in all conversations, the variability was minimized, and the reliability increased. In turn, the validity of the obtained results increased. Moreover, all calls were coded by the same coder at MIC Lab AB to avoid the risk of inter-rater bias. The intra-rater reliability was considered high as the coders at MIC Lab AB undergo continuous training and regular controls.

However, the MITI scores may not only vary depending on the external conditions but also on the veterinarian's daily form. This implies that the estimate of the veterinarians' true MI-skills would have been improved if multiple calls with each veterinarian had been coded. Furthermore, the number of calls in this study was very limited, thus further studies with a larger scope are required to be able to characterize the communications styles of Swedish small animal veterinarians in general. In addition, the sample of veterinarians was not chosen at random which implies potential bias due to over- or under-representation of subgroups in the sample compared to the target population. For these reasons, the sample of the present study may not be representative of the population of Swedish small animal veterinarians, which should be considered when interpreting the results.

The dog owner profile was created in collaboration with the experienced MI-researcher Lars Forsberg at MIC Lab AB, based on literature review and clinical experience of a typical dog and owner. There is likely a variation in the veterinarian's communication depending on the pet owner, hence the results may have been different if the dog owner profile, or the actor, would have been another. It should, however, be noted that the conversations with the actor were experienced by all participating veterinarians as completely natural and resembling a real clinical situation.