Effects of training podiatrists to use imagery-based motivational interviewing to improve self-care for people with diabetes-related foot disease: A mixed-methods pilot study

Background Self-care in diabetic foot disease (DFD) is challenging and can contribute to poor outcomes. Motivational Interviewing (MI) engages people in self-care and integrating imagery may further improve its outcomes. No previous studies have trained podiatrists in using MI to address DFD self-care. This was the rst study on training podiatrists to conduct imagery-based motivational interviewing (MI) when treating people with diabetes-related foot disease (DFD), and to examine impacts on MI related skills, job satisfaction and subjective experiences in a mixed-methods pilot study. Methods Eleven recruited podiatrists (Median age 35, 9 female) received two 4-hour training sessions and three received later mentoring. MI and imagery skills were rated using validated tools during two clinical sessions per participant at baseline, and 2- and 12-weeks post-training. Job satisfaction was assessed at baseline and 12 weeks. Semi-structured interviews at 12 weeks were analysed using the framework approach. Results Signicant improvements over time ( p =.006-.044) with substantial effect sizes (η 2 =.50-.67) were found in three of four global MI related communication skills and two of four MI behaviours. However, effects on these indices were not sustained to 12 weeks, and imagery was rarely used. Job satisfaction was high at baseline and unchanged at follow-up ( p =0.34, η 2 =.100). In qualitative interviews, MI training and skills were valued, but signicant challenges in using MI when treating people with DFD were reported. Conclusion Training podiatrists in MI may have potential but more training, observation and mentoring appear needed to obtain sustained communication changes in practice.

motivate and guide their self-management. An early form of this approach was tested in people with type 2 diabetes and showed substantial acceptability (26). A multi-session version has shown stronger outcomes than standard MI in dieting, exercise and weight reduction (27,28).
Effective application of imagery-based MI requires that practitioners are su ciently trained to apply it routinely with high delity in their work (19). MI skills can be acquired by nurses or general practitioners who care for people with diabetes, albeit to varying success (22,(29)(30)(31)(32)(33), depending in part on the length of MI training and availability of supervision (34). However, training podiatrists in using MI to address DFD self-care has not been investigated yet. Podiatrists are in a prime position to engage people with DFD in carrying out self-care, and training podiatrists in imagery-based MI could be a valuable addition to their clinical skills (35). An observational study of podiatrists working with people with DFD identi ed that 90% perceived their role in promoting self-care as vital, but only 25% reported use of "MI type" skills (36). Increasing this level represents an acute need.
Our aim was to train podiatrists to conduct imagery-based MI when treating people with DFD, and to evaluate effects on related skills, job satisfaction, training experiences and implementation attempts.

Methods
This mixed-methods pilot study involved podiatrists who primarily treated people with DFD. Its quantitative component

Participants
Participants for this study were consenting podiatrists who were registered by the Australian Health Practitioner Regulation Agency and worked within government funded DFD Services in Brisbane, Australia; they primarily treated people with DFD (i.e. with a diagnosis of diabetes and a history of foot ulceration). Exclusion criteria included prior training in MI. Potential participants were informed about the study by recruitment emails and posters displayed in clinical rooms.
Participating podiatrists recruited people with DFD to assist with outcome assessment. These people also provided voluntary written consent to participate in recorded clinical sessions during which podiatrists' skills in imagery-based MI were assessed.

Intervention
The training was planned in collaboration with the manager of the podiatry team, who gave the project strong support, linked training sessions to team meetings and actively participated in the training sessions. The intervention training program was jointly developed and delivered by two authors: a clinical psychology researcher (DK) and a senior podiatrist in the trained team (TK). DK is a co-developer of FIT and has led multiple controlled trials using MI or FIT (27,37,38). TK was trained in MI and FIT before study commencement and had experience applying these skills in her own clinical sessions, allowing her to incorporate detailed descriptions of her experience in training.
The program comprised two 4-hour training sessions in imagery-based MI, separated by 2 weeks. Training was delivered in groups of 5-6 participants. The rst session included group discussion of issues participants had experienced in attempting to facilitate DFD-related self-care, and of advantages for morale of de ning success in terms of delivering a favourable context for change, rather than reaching behavioural or clinical targets. It outlined core concepts of MI and provided video examples of MI, followed by group discussion of potential application in podiatry sessions. A written manual gave a script for use in consultations. To maximise compatibility with existing sessions, trainers suggested incorporating elements of MI when assessing self-management, conducting physical treatment and collaboratively planning future self-care.
The second session described the use of imagery and home-practice in podiatry sessions, with demonstrations, roleplayed practice and feedback to build skills and con dence. Following feedback from participants, an abbreviated version of the manual was provided. Key elements of the manual were later converted into a owchart placed in the clinic rooms, to prompt the application of skills in sessions.
All participants were offered follow-up peer support opportunities from TK to observe and feedback on consultation sessions or provide peer-mentoring via email, phone or face-to-face. Such feedback and advice were positive and practical, recognising incremental improvements and helping participants solve challenges in their application of MI related skills. Emails reminding participants about key elements of the intervention were sent 4 and 10 weeks after the workshop training.

Outcomes of interest
The outcomes of interest for the quantitative component included MI related skills, use of imagery and job satisfaction. MI related skills and imagery were assessed using audio recordings from two clinical consultation sessions by each podiatrist at baseline, and 2-and 12-weeks post-training. Sessions were selected opportunistically, based on availability of the participants and researcher. No data on characteristics of the people with DFD or outcomes of sessions were collected, since the focus was on podiatrists' conduct of sessions. If a friend or relative of the person with DFD were also present, they were asked to remain silent during the consultation. If they became involved, the session was omitted from data collection and another was used. FIT-related skills during the same session segments were assessed using the Functional Imagery Training -Quality Coding (FIT-QC) 1.0 (40) which includes introduction to imagery, instruction, delivery, use and adaption of imagery, re nement of imagery quality and focus, re ection and promotion of home practice. Each component was scored from 1 (poor) to 5 (high).
FIT skills were assessed by the same authors (TK and DK).
Job satisfaction at baseline and 12-weeks post-training was assessed using the Hoppock Job Satisfaction Scale (41), which has 4 items rated 1 to 7, with higher total scores denoting higher satisfaction. (41) Individual face-to-face semi-structured interviews at 12 weeks post-training were undertaken by the primary author (TK), audio recorded and transcribed verbatim. The interview was designed to gain insight into the podiatry participant's experiences in training and their application of MI skills into clinical practice. Each primary question in Table A1 was augmented by non-speci c questions to obtain additional responses. Have you noticed any differences in your patient's responses since doing the training?
What was your experience of the training like?
What aspects of training seemed most useful for your practice?
What aspects of the training seemed least useful for your practice?
What aspects of the training were easiest to apply into clinical practice?
What particular aspects were more di cult to apply?
What challenges did you face in applying what you learned in the training?
What could we do to help you keep using your skills from the workshop routinely with your patients?

Analysis
Quantitative data were analysed using repeated measures general linear models on IBM SPSS v26, with F values using Pillai's Trace. The primary focus was the overall effect for time, but nonorthogonal contrasts examining changes from baseline to 2-and 12-weeks post-training were also examined. Because this was a small pilot, η 2 (the proportion of total variance accounted for by an effect) was the primary focus rather than statistical signi cance, although both are reported.
Qualitative interviews were analysed using the framework approach (42). After transcription of interviews, participants' responses were analysed using: 1) familiarisation (re-reading transcripts and eld notes), 2) identi cation of the thematic framework (identifying key themes from data), 3) indexing (arranging transcripts by themes), 4) charting (creating a matrix with responses collated), and 5) mapping and interpretation (interpreting the data). Themes were identi ed by one author (TK) and checked and discussed by another (JvN) until consensus was reached.

Results
Of 14 eligible podiatrists who were approached, 11 consented to participate. Of the remaining three, one was on vacation, one did not express interest and one expressed interest but did not consent. Participants were aged between 29-43 years (median 35), nine (82%) were female, and years of DFD-treatment experience ranged from 2-17 years (median 11). Two training sessions were delivered to all participants. Five participants requested peer support following training and three of these received feedback about a clinical consultation session.
Three of the four global score ratings (Change Talk, Partnership and Empathy) and two of the four core behaviour counts (A rm and Persuade) showed statistically signi cant improvements and substantial effect sizes (η 2 = .50-.67) across the three time points (Table A2). Contrasts comparing scores with those at baseline showed signi cant improvements on these six indices 2 weeks but not 12 weeks post-training.
None of the ve other behaviour counts showed statistically signi cant improvements, but Simple Re ections, Complex Re ections and Giving Information each showed moderate effect sizes (η 2 = .43-.48). Only Giving Information showed a statistically signi cant improvement from baseline to 12 weeks. As anticipated, both Giving Information and Questions were frequent throughout the study.
No substantial use of FIT Skills was evident in recordings, with only one podiatrist attempting to use motivational imagery after the training. Accordingly, detailed results on the FIT-QC are not reported.
Three main themes with 10 subthemes were identi ed from the interviews (Table A3). Before training, all podiatrists identi ed struggles with fostering self-care by people with DFD. These struggles included situations where podiatrists tried to achieve commitment to self-care.
"… a bit like that feeling like your hitting your head against a brick wall. That you know they know what they should be doing and they're just not doing it and it's frustrating to be able to have the solution but them not being invested enough in their own care to do it" (Pod B)

Subtheme 1.2 Communication challenges:
Podiatrists also reported challenges with communication, especially when the person had multiple comorbidities or limited education, and when relatives or friends joined consultations.
"The relatives and the family, because it becomes a joint problem a joint concern…some of the carers have sort of pushed their needs onto the patient and you have…to steer the consult back to the patient sometimes." (Pod F).

Subtheme 2.1 Overall training experience:
Most podiatrists reported that MI training was enjoyable, bene cial, interesting and informative. They liked the small group and felt engaged.
"I found the training days very bene cial and informative, and I think that we were very engaged. It was quite good because it was interactive." (Pod G).
"It was good….I de nitely did learn better communication skills to a certain extent" (Pod K).

Some felt:
"There was a lot of information in a short period of time" (Pod C) "Hard to relate it sometimes to a clinical setting" (Pod D) Sub theme 2.2 Role-play: Role-play practice elicited mixed feelings. Some found it useful, despite some negative feelings.
"Nice to see how different clinicians worked." (Pod A).
"As much as I hate role-play, it did help." (Pod I).
Others said: "[Role-play is] tricky as we think differently than the patients." (Pod J). "…having those open-ended questions other than short-ended questions, so you could really just learn to shut up and let the patient speak." (Pod E) "Strategies of getting the patient to commit themselves to something as opposed to us dictating to them what they needed to do." (Pod E) Participants liked the fact that training allowed them to ask people with DFD to re ect on what they had previously done and wanted to do: "I think the open-ended questions and the re ections back were really useful for developing a rapport with the patients." (Pod D) Other useful skills were asking for goals, allowing the patient to talk more, and reducing podiatrists' talk time: "I have allowed the patients to speak more which has given me an insight that I did not have before" (Pod F).
"To be quiet as well, that's a hard one. It's a very practical thing to let them just speak, to ll the silence rather than us." (Pod B).
"Getting the patient to tell you what their plan is and what their goals are, because at the end of the day it's their health care." (Pod A).

Subtheme 3.2 Increased patient engagement:
The new communication skills appeared to result in increased patient engagement: "They've seemed to be more engaged with their self-care. And instead of telling them what to do I feel that they respond better because they are thinking what they can do better. They're taking responsibility more." (Pod A) "Initially getting people to open up and talk about experiences with foot ulcers and how they can change things, its changing their mentality from them thinking that its actually our problem and our responsibility to them actually having to think about it as their wound and their responsibility." (Pod G) Subtheme 3.3 Long-term implementation of MI skills: Participants reported di culty with changing their habitual behaviour in sessions. While anticipating that applying MI over the longer-term would be hard, some were optimistic about achieving this.
"…it was de nitely still challenging. Hard not problem-solving" (Pod D) Participants suggested ways to support them to sustain their use of MI related skills over the longer term: "[We need] more training within motivational interviewing to keep your skills up and keep it fresh. The more you do it, the more it will embed into your practice, but at the beginning it's easy to go back to how you used to do." (Pod A) "We slowly slip back into our old ways because we don't reinforce it and we're really time poor." (Pod F) Subtheme 3.4 Appropriateness of MI: MI was not seen as an appropriate approach for every patient: "Some of them still do the same thing regardless." (Pod J).
"I guess I use it where I see the need." (Pod I).

Discussion
This was the rst study to investigate the impact of delivering MI training to podiatrists who treat people with DFD. The training resulted in some moderate short-term improvements in MI related skills after 2 weeks, but no improvements remained at the 12-week follow-up, and only one podiatrist attempted to use imagery in the assessed sessions.
Improvements in job satisfaction were not observed, although satisfaction was already high at baseline and the training did not encompass broader issues related to the work setting.
The qualitative responses were largely consistent with other studies. Participants saw the interactive nature of workshops as bene cial, although some expressed discomfort about role-play (43). Participants had positive experiences with engaging and empowering patients after training, as found in another study (44), but expressed a need for ongoing supervision and support and noted the risk of reverting to old habits (45).
The short-term improvements in MI related skills of Empathy (30,31,46,47), Change Talk (22,30,48) and Partnership (29) were consistent with changes in previous studies on MI training. Reductions in Persuasion are less often reported explicitly (32); in the current study, this behaviour appeared to return to baseline levels at 12 weeks. The skill Soften Sustain did not change during the study, but statements requiring this type of response by podiatrists were surprisingly rare in the assessed audios. Rises in Simple and Complex Re ections were small in number, and effects for time fell short of statistical signi cance, suggesting that they require further attention in training. While both Questions and Giving Information were very frequent, these were required elements in the podiatry sessions, and it is not clear what would be an ideal frequency. The 43% reduction in Giving Information was encouraging, if it re ected a reduction in the repetition of information within and across consultations, since such repetition runs the risk of undermining rapport and collaborative self-management.
Participants' positive experiences and short-term MI related skill improvements provide a solid basis for further changes in practice. However, the lack of maintained changes in core MI skills and reported di culties with their routine use suggest a need for additional training, including follow-up peer support session observation and feedback. The qualitative responses suggested this would be highly valued. With 8 hours of training provided in the current study, half of the median 16 hours found in a systematic review (34), more training may indeed be needed for lasting positive effects on practice. In particular, further group discussion of di culties experienced in applying MI is recommended. Di culties in maintaining the use of new skills has often been observed in other studies on practitioner training (22), including training on MI. A reliance on workshops alone typically gives limited skill improvement (49) and little sustained bene t (45,50,51). Monitoring accompanied by encouraging and corrective feedback appears critical (44,45,(50)(51)(52), at least until use of the skills becomes habitual. The positive short-term training outcomes and positive evaluation in this pilot study suggest that podiatrists are receptive of more intensive training, can be seen as an encouragement for further research and implementation.
A strength of this study was our incorporation of several principles of implementation science to facilitate training uptake (53,54). The project had strong management support, and one of the authors (TK) was a senior member of the team and modelled early adoption of the approach in her own work. Further, the training highlighted incentives for use of MI, including the potential for MI to reduce podiatrists' guilt or frustration by de ning success in terms of their conduct of sessions rather than clinical or behavioural outcomes. Finally, we minimised change from existing practice by integrating MI with assessment, physical treatment and planning within the existing session length.
Despite these facilitators, podiatrists had di culties integrating MI into clinical sessions if these required substantial clinical assessment and physical treatment. This integration probably needed more training than we were able to deliver, and for some attempting to use a new skill while conducting another task may have been too di cult. Podiatrists were also discouraged when people with DFD did not respond positively, and additional encouragement to persist in the face of this is required. The complexity of people's co-occurring problems also presented signi cant challenges for use of MI, as did addressing the needs of carers.
A limitation of this novel pilot study was the small sample size. However, we recruited su cient participants to provide estimates of potential effect sizes for the planning of future randomised controlled trials, as well as important ndings on training content and implementation. A further potential limitation was involvement of the trainers in the conduct and rating of assessments, which may result in assessment reactivity and rating biases. We minimized this by using validated tools, and having two assessors working independently. Qualitative interviews were short, but derived rich data on participants' experiences. Use of mental imagery may have also been a limitation, as it proved a step too far for the participants. They found it di cult to introduce and use in sessions, which may have blocked additional uptake of MI related skills. In retrospect, improvements in MI related skills should have been consolidated before introducing mental imagery, and more training on introducing and using imagery was needed.

Conclusion
We provided eight hours of valued training in MI to podiatrists who treat people with DFD. This resulted in some uptake of MI related skills, although maintenance of these changes was short lived. Findings of this study provide a solid basis for re nement of the training and support for implementation and future research of its impact on both podiatry practice and outcomes for people with DFD.

Declarations
Ethics Approval and Consent to Participate Approvals were obtained from human ethics committees from the Research Governance Unit The Prince Charles Hospital and Queensland University of Technology (# 45318), and all participating podiatrists and people with DFD provided informed consent before participation began.

Consent for publication
Not applicable

Data availability
Deidenti ed data sets generated/or analysed during this current study are stored at https://github.com/dkavanagh212/DFDPodiatry.git.