Insight in the development of the mutual trust relationship between trainers and trainees in a workplace-based postgraduate medical training programme: a focus group study among trainers and trainees of the Dutch general practice training programme.

Objectives Trust plays an important role in workplace-based postgraduate medical education programmes. Trainers must trust their trainees for granting them greater independence. Trainees must trust their trainer for a safe learning environment. As trainers’ and trainees’ trust in each other plays an important role in trainee learning and development, the authors aimed to explore the development of the mutual trust relationship between trainers and trainees. Setting This study was performed in a general practice training department in the Netherlands. Participants All trainers and trainees of the general practice training department were invited to participate. Fifteen trainers and 34 trainees, voluntarily participated in focus group discussions. Outcome measures The authors aimed to gain insight in the factors involved in the development of the mutual trust relationship between trainers and trainees, in order to be able to create a model for the development of a mutual trust relationship between trainers and trainees. The risk-based view of trust was adopted as leading conceptual framework. Results In the first stage of trust development, trainers and trainees develop basic trust in each other. Basic trust forms the foundation of the trust relationship. In the second stage, trainers develop trust in trainees taking into account trainees’ working and learning performance, and the context in which the work is performed. Trainees trust their trainer based on the trainer’savailability and accessibility and the personal relationship between the trainee and their trainer. Trainee self-confidence modifies the development of a trust relationship. Conclusion The development of a mutual trust relationship between trainers and trainees is a complex process that involves various stages, goals, factors and interactive aspects. As the mutual trust relationship influences the learning environment for trainees, greater emphasis on the mutual trust relationship may improve learning outcomes. Further research may explore the effect of long-term and short-term educational relationships on the trust relationship between trainers and trainees.

1) I didn't see a definition of "trust" in the paper, a term that has psychological, sociological, and philosophical dimensions and definitions. I think a better term than "trust" is that you were exploring self-described "feeling of something I define as trust" rather than actual trust. In its broadest sense, "trust" is a construct in which one person feels he or she can rely on the actions of the other. Given the questions used in the focus groups, it seems the authors let each individual define what he or she meant by "trust", which may be fine. However, it is also problematic because what each person means by "trust", and therefore how it is achieved, is not well defined.
As I step back and consider the results, I come to the conclusion that the model described in Figure 1 can serve as trust definitions, which are different for trainers and trainees.
For trainers, it sounds like their implicit definitions revolved around their feeling that they could rely on both the reports of trainees of their past behavior (e.g., "I saw a patient with. . . ") as well as their belief that the trainee could be relied upon to perform competently in future situations that the trainer could not control (entrustment). For trainees, on the other, hand, it seems the feelings of trust revolved more around their perception of the benevolence of the trainer (i.e., the psychological safety they would create) and the learning and patient care support the trainer would provide to the trainee at the time and in the future. For both groups, there is also the issue of power as it relates to trust that could be explored in the analysis. So, I suggest adding a description or descriptions of "trust" in the introduction and pointing out that you let participants use their own definitions of trust when answering your questions. I would like you to consider that from your focus groups you discovered what were the elements of the relationship that resulted in trainers' and trainees' development of a feeling of what they defined as trust.
2) I have an alternative explanation for trust relationships differing between long (years one and three) and short (year two) traineeships. It seems that an important variable, in addition to duration of the relationship, is that the supervisors in year two are not general practitioners. As a result, there may be power differentials, differences in cultures, and an ethnocentric mismatch between GP trainees and their specialist trainees that hamper the development of both trust and the feelings of trust on both sides (such is the reality, I think, in the United States). I suggest the authors consider whether they want to address this variable in their analysis.
3) I don't see a methodological orientation in the paper. I see your method (thematic analysis) but I don't see a theoretical framework guiding your research. A conceptual framework is necessary to understand the strengths and limitations of your approach (see:

REVIEWER
Malcolm Moore ANU College of Medicine and Health Sciences, Medical School REVIEW RETURNED 04-Feb-2020

GENERAL COMMENTS
Although this study isn't breaking new ground, it contextualises the study of trust to the GP context and successfully aligns its outcomes with the existing literature. The findings about the difference in factors affecting developing trust between trainers and trainees are interesting and provide some useful insights for trainers. I don't think there are any significant revisions required. I note with interest that 82.4% of trainees are female -I suggest that this is mentioned in the results and a comment made about the representativeness of this sample.

VERSION 1 -AUTHOR RESPONSE
Reviewer 1 * 1) I didn't see a definition of "trust" in the paper, a term that has psychological, sociological, and philosophical dimensions and definitions. I think a better term than "trust" is that you were exploring self-described "feeling of something I define as trust" rather than actual trust. In its broadest sense, "trust" is a construct in which one person feels he or she can rely on the actions of the other. Given the questions used in the focus groups, it seems the authors let each individual define what he or she meant by "trust", which may be fine. However, it is also problematic because what each person means by "trust", and therefore how it is achieved, is not well defined.
As I step back and consider the results, I come to the conclusion that the model described in Figure 1 can serve as trust definitions, which are different for trainers and trainees.
For trainers, it sounds like their implicit definitions revolved around their feeling that they could rely on both the reports of trainees of their past behavior (e.g., "I saw a patient with. . . ") as well as their belief that the trainee could be relied upon to perform competently in future situations that the trainer could not control (entrustment). For trainees, on the other, hand, it seems the feelings of trust revolved more around their perception of the benevolence of the trainer (i.e., the psychological safety they would create) and the learning and patient care support the trainer would provide to the trainee at the time and in the future. For both groups, there is also the issue of power as it relates to trust that could be explored in the analysis.
So, I suggest adding a description or descriptions of "trust" in the introduction and pointing out that you let participants use their own definitions of trust when answering your questions. I would like you to consider that from your focus groups you discovered what were the elements of the relationship that resulted in trainers' and trainees' development of a feeling of what they defined as trust.
-Thank you for your thoughts on this part of the manuscript and for your recommendations. We now start the introduction of the manuscript with a definition and a discussion on the definition of "trust".
(Trust (noun): "Firm belief in the reliability, truth, or ability of someone or something".(1) Entrust (verb): "To put one's trust in a person, with regard to a particular task or responsibility."(2)Trust is a complex expectation, not easy to summarize in a simple definition. The amount of trust someone has in another person is very much dependent from personal factors, such as thoughts and motivations, as well as it is dependent of the context in which the trust has to be awarded. (3)(4)(5) Additionally, trust is not an established fact, as it can be gradual and variable.(5) , (P: 7,L: 98-106)) We have also included information in the "Data collection"-section, stating that we encouraged participants to explore their definition of trust and use their own definition of trust during the focus group discussions. ("As a first step, we asked participants to explore their own definition of trust. Participants were encouraged to use their own definition of trust during the focus group discussion." (P: 11, L: 209-212)) We have included information in the results-section to point out that it are elements of the relationship between the trainer and the trainee that contribute to the development of the feeling of trust. ("Based on our data we created a model that is based on the elements of the relationship between trainers and trainees that contribute to the development of a feeling of trust in trainers and trainees. ( Figure  1)." (P: 14, L: 260-262)) At last we have included information in the strengths and limitations-section in the discussion in order to point out that trust-development is a very subjective process, and that the meaning of trust might vary between different persons. Although the feeling of trust is very personal, participants indicate the same factors involved in the trust development process, we therefore think that these results are representative for the trust development process between trainers and trainees.("Trust is a feeling, that might have a different definition for every single person. (4-6)This might cause challenges in investigating trust development. Although the feeling of trust might be very personal, participants in this study indicate the same factors involved in trust as reflected in available literature. (7)(8)(9)(10)(11)(12)(13)(14) We therefore think that the results of this study are representative for the trust development process between trainers and trainees. "(P: 22, L: 459-463)) * 2) I have an alternative explanation for trust relationships differing between long (years one and three) and short (year two) traineeships. It seems that an important variable, in addition to duration of the relationship, is that the supervisors in year two are not general practitioners. As a result, there may be power differentials, differences in cultures, and an ethnocentric mismatch between GP trainees and their specialist trainees that hamper the development of both trust and the feelings of trust on both sides (such is the reality, I think, in the United States). I suggest the authors consider whether they want to address this variable in their analysis.
-Thank you for your considerations. We think that to some extent these power differentials, differences in cultures and the ethnocentric mismatch also hold for the Dutch GP-training program. However, Dutch trainee GPs also have to participate in nursing homes, which in the Netherlands is also a primary care facilitation with a similar educational structure, in which the power differentials, differences in cultures and the ethnocentric mismatch are not as striking as they are in hospital settings. Since trainees also mention these nursing home rotations as rotations in which they find it difficult to develop a personal relationship with their trainer, we believe that the power/cultural/ethnocentric differences between primary care and hospital care settings are not the primary reason for the differences in the development of the feeling of trust. We did however include a consideration in the discussion-section concerning this subject.(Other explanations for the difference between long and short rotations could be the differences in culture between working in hospitals and working in GP-practices, as well as power-differences and ethnocentric mismatches between GPtrainees and their trainers in hospitals. (P: 22, L: 449-452)) We do think that the finding of the differences between short and long traineeship rotations merits further evaluation. Our finding was that trainees indicate a difference in trust development between short and long rotations, future research could benefit from evaluating which factors involved in short and long rotations cause this difference in trust development. We have included a recommendation for future research in the discussion-section.("Future research The trainees provided us with important insights concerning the effect of an educational relationship's duration on the trust relationship between trainers and trainees. However, we do not know which factors involved in short and long educational relationships influence the trust relationship between trainers and trainees. Additionally, in this study we did not explore the vision of trainers involved in short rotations. The distinction between long-term and short-term educational relationships and their effect on the trust relationships between trainers and trainees were therefore not fully explored and merit further study."(P: 23, L: 477-483)) Additionally, literature on the effects of long rotations and longitudinal training relationships is increasing, suggesting that longitudinal training relationships have a positive influence on various aspects of trainee learning, especially due to the fact that trainees really get to engage in everyday working, together with their trainer. (3,15,16) Our hope is that in the future long educational relationships and exchange of trainees between primary and secondary/tertiary care will help us to overcome those differences in power, culture and ethnocentric mismatch. -Thank you very much for this methodological improvement of our manuscript. In our studies on this subject we use the risk-based view of trust (Das and Teng) as a guiding framework for data analysis. We did include some information in the methods-section.(For the leading conceptual framework we chose to adopt the risk-based view of trust (Das and Teng, 2004) (17), where we assume that trust development is an ongoing process. We also assume that developing trust will have certain mutual aspects, because both parties involved in the trust development process have to take a risk in order to be open to each other's actions, and be able to develop trust in each other.(17) (P: 9, L: 162-166)) * 4) Your COREQ checklist doesn't seem to line up with your paper, making it hard to verify.
-Thank you for your attention, we have re-filled the form using the revised version of the manuscript without track changes. It might be possible that page-and line-numbers change with building the PDF-file for submitting the article. We hope that the renewed COREQ checklist now does line up with the paper.
Reviewer 2 * Although this study isn't breaking new ground, it contextualises the study of trust to the GP context and successfully aligns its outcomes with the existing literature. The findings about the difference in factors affecting developing trust between trainers and trainees are interesting and provide some useful insights for trainers. I don't think there are any significant revisions required. I note with interest that 82.4% of trainees are female -I suggest that this is mentioned in the results and a comment made about the representativeness of this sample.
-Thank you very much for your thorough review of the manuscript. Since 77% of the GP-trainees in the Netherlands is female, females are overrepresented in our study, we included some information about the study population in the results-section. Since the study population is relatively small (when compared to large cohort studies), we do not think that this difference affected the results of our study (our study required 2 more male trainees instead of female trainees) and that the results of our study are representative for the Dutch GP-training program. We did include some information in the discussion-section.(Of the participating trainees, 82.4% was female. Since 77% of trainee GPs in the Netherlands are female (18), females were slightly overrepresented in our study.. We do not think that the results of our study are influenced by this variation and feel that the results are representative for Dutch trainee GPs. (P: 22, L: 465-468))

REVIEWER
Allen F. Shaughnessy Tufts University School of Medicine, USA REVIEW RETURNED 20-Mar-2020

GENERAL COMMENTS
I think the addition of the definitions of trust and the use of the conceptual framework make this paper much better. I suggest two minor additions that will help guide the interpretation and to align it better with the existing literature: 1) Devote a paragraph (or three) or two to explaining Das and Teng's model, specifically on how trust develops in the context of risk (trust antecedents, subjective trust, behavioural trust). Your interpretation of their work, as it applies to yours, will be very helpful to readers.
2) In the discussion, return to Das and Teng and explain how trust and risk are intertwined (see Figure 2 of your reference 25) and how faculty members can use your results to engender trust and examine their ability to trust in their residents. This explanation could be somewhat extensive

REVIEWER
Malcolm Moore ANU College of Medicine and Health Sciences, Medical School REVIEW RETURNED 22-Mar-2020

GENERAL COMMENTS
Thanks for answering my query re gender differential.

Reviewer 1
Reviewer Name: Allen F. Shaughnessy Institution and Country: Tufts University School of Medicine, USA Please state any competing interests or state 'None declared': None Please leave your comments for the authors below I think the addition of the definitions of trust and the use of the conceptual framework make this paper much better. I suggest two minor additions that will help guide the interpretation and to align it better with the existing literature: 1) Devote a paragraph (or three) or two to explaining Das and Teng's model, specifically on how trust develops in the context of risk (trust antecedents, subjective trust, behavioural trust). Your interpretation of their work, as it applies to yours, will be very helpful to readers.
interpreted the data. We included additional information about the interpretation of the risk-based view of trust in the "Study design"-section of the paper.
(As leading conceptual framework we chose to adopt a risk-based view of trust (Das and Teng, 2004).(1) By using this framework, we assume that trust and risk are theoretically each other's opposites. This means that when we trust someone, we experience a low risk for us in trusting the other person (perceived risk), and we are willing to accept that risk. On the other hand, when we do not trust someone, the perceived risk in trusting the other person is high, and we are not willing to accept that risk. The feeling of a certain amount of trust in someone is called subjective trust, the perception of trust. Subjective trust is then influenced by personal and situational factors, the socalled trust antecedents in the terminology of Das and Teng. The interplay between subjective trust and trust antecedents leads to certain behaviour in taking risks and trusting someone, also known as behavioural trust. (1) Within the framework of a risk-based view of trust, the outcomes of behavioural trust may again inform and influence a person's subjective trust. Additionally, they believe that trusting somebody, and therefore taking a risk in that person, can boost the development of a mutual trust relationship.(1) We use these assumptions as the basis of our study, as we aim to explore how the mutual trust relationship between trainers and trainees develops, and which factors are involved in this process. (P9-10, L153-168)) Additionally, throughout the "Results"-section of the paper we identified trust antecedents, perceived trust, perceived risk and behavioral trust among our results. (P17-19, L299-425) This creates a better transition to the discussion section. Throughout the "Discussion"-section of the paper we made some adjustments in order to better align the discussion. (P23-28, L434-558) 2) In the discussion, return to Das and Teng and explain how trust and risk are intertwined (see Figure 2 of your reference 25) and how faculty members can use your results to engender trust and examine their ability to trust in their residents. This explanation could be somewhat extensive framework and ads an extra dimension to the discussion. We included additional information in the "Discussion"-section of the paper.
("When returning to the framework of a risk-based view of trust, we see that the opposite of subjective trust is perceived risk. When developing trust in the other person, subjective trust and perceived risk are both influenced by the competence (the ability to fulfil the task) and intentions (integrity and willingness to fulfil the task) of the person we have to trust. This means that estimates of good competence and good intentions lead to high levels of subjective trust and low levels of perceived risk. And the other way around.(1) Competence and intentions also play an important role in the mutual trust development process between trainers and trainees. Trainees evaluate a trainers competences regarding being an accessible trainer and the intentions of the trainer of being accessible and willing to help. Trainers on the other hand evaluate the trainees competence of having insight in their own learning process and the intentions to be open about their working and learning process. Being aware of the role of competence and intentions in the trust development process, and in the differences in competence and intentions between trainers and trainees might help both trainers and trainees in the (more early) recognition of trust and risk. Perceived trust and perceived risk are highly dependent of personal factors (2)(3)(4), leading to different definitions for trust and risk between trainers and trainees. However, insight in the factors and processes of trust development might help both trainers and trainees not only to be able to discuss differences in the understanding of the mutual trust relationship with each other, but also to pay specific attention to these factors when developing a trust relationship or handling a trust-issue." (P23-24, L445-463)) Reviewer 2 Reviewer Name: Malcolm Moore Institution and Country: Australian National University Australia Please state any competing interests or state 'None declared': None declared Please leave your comments for the authors below Thanks for answering my query re gender differential. Literature