Exploring power dynamics and their impact on intraprofessional learning

Abstract Background During postgraduate training, considerable efforts for intraprofessional education are in place to prepare primary care residents (PC residents) and medical specialty residents (MS residents) for intraprofessional collaboration (intraPC). Power dynamics are inherently present in such hierarchical medical contexts. This affects intraPC (learning). Yet little attention has been paid to factors that impact power dynamics. This study aims to explore power dynamics and their impact on intraPC learning between PC residents and MS residents during hospital placements. Methods This study expands on previously published ethnographic research investigating opportunities and barriers for intraPC learning among residents in five Dutch hospitals. We analysed transcripts of observations and in‐depth interviews using template analysis. A critical theory paradigm was employed. Discourse analysis additionally informed the data. Results We defined five interrelated themes that describe characteristics of power dynamics in intraPC learning during hospital placements: beliefs; power distribution; interaction style; subjection; and fearless learning. Power dynamics operate both within and between the themes: power distribution between PC residents, MS residents and MS supervisors seemed to be an attribution affected by underlying beliefs about professional norms or about other professions; beliefs influenced the way PC residents, MS residents and MS supervisors interacted; power distribution based on inequity could lead to subjection of PC residents; power distribution based on equity could lead to fearless learning; and open interactions enabled fearless intraPC learning. Conclusions Power dynamics have an impact on intraPC learning among residents in hospitals. Constructive power dynamics occur when power distribution is based on equity, combined with sincere open interactions, actively inviting each other into discussions and enlisting the support of MS supervisors to foster fearless learning. This can be achieved by creating awareness of implicit beliefs and making them explicit, recognising interaction that encourages intraPC learning and creating policies that support fearless intraPC learning.


| INTRODUCTION
Collaborative practice between primary care (PC) physicians and medical specialists (MSs) is vital and requires mutual trust and respect. [1][2][3][4] In the deep-rooted hierarchical contexts of hospitals, however, it could be a measure of status for MSs to disrespect lower-status professionals with impunity, 5 such as PC physicians. 3,6,7 Power dynamics based on traditional hierarchies are inherently present in (intra)professional interaction and learning processes 5,6,[8][9][10] and could have an adverse effect on collaborative practices 5,8 leading to adverse events in healthcare. 3,11 Often power dynamics are not openly discussed, but referred to implicitly, contributing to the hidden curriculum.
To prepare PC residents (PC residents) and medical specialty residents (MS residents) for collaborative practice, the learning of intraprofessional collaboration (intraPC) through intraprofessional education (intraPE) is an emerging part of postgraduate training. [12][13][14][15][16][17][18] 18 For example, hospital placements, where PC residents and MS residents work together at the same department, provide several opportunities for intraPE. 15 These placements occur worldwide. 15,[19][20][21][22][23] A Dutch study found that PC residents, MS residents and MS supervisors mentioned issues with power dynamics that influenced intraPC learning during hospital placements. 15 Arabic studies have found that the personal attitude of MSs can make PC residents experience inferiority of feel inferior, leading to deficiencies in learning during hospital placements. 24,25 Canadian studies, furthermore, have found that more than one-third of the PC residents experience harassment and intimidation arising from power dominance by MSs and MS residents during hospital placements. 26,27 As such, power dynamics can lead to interpersonal fear. 28 Although considerable efforts are being made to design interprofessional/intraprofessional education (IPE/intraPE), little attention has so far been given to factors that impact hierarchy and power dynamics. 8,29 The vast majority of studies about IPE/intraPE focus on programmes or curricula, but omit to critically investigate the impact of power. 8,30 The same holds true for studies about hospital placements. By not addressing power dynamics, however, an ambiguous and opaque problem remains in place. 30,31 To improve the learning climate for intraPC learning, PC residents, MS residents and their supervisors need to have a better understanding of the impact of power dynamics. 8

| THEORETICAL BACKGROUND
In scientific literature, power and power dynamics seem to be easier to recognise than to define. Dahl (1957) explains power as a form of control: 'A has power over B to the extent that he can get B to do something that B would not otherwise do'. 32 A/B can be a person, team or organisation. King Jr (1968) describes power as the ability to bring about change 33 or as the capacity to act or not to act. Raven (2010) defines power as a form of interpersonal influence which may be based on various sources: expertise, information, (formal) position, being a reference or the ability to exert coercion or reward. 34 Bynum (2021), finally, elaborates that power hierarchies/distribution in medical learning environments are often manifested through knowledge, vulnerability, risk taking and influence. 10 Underlying these definitions are philosophical roots of thinking about power. Arendt (1970) and Foucault (1976) explain that there is not one place or person where power emerges from, but that it is rather constructed between people and continues to exist as long as these people stay together. 35,36 The interaction of power between people can be understood as a dynamic process, 35,37 as an unstable network of practices that spreads throughout society and may exist within workplaces, institutions, or other places where people come together. In this article, we use the term 'power dynamics' to describe the way in which power impacts the interaction of two or more people or groups. Power and power dynamics are essentially neutral, not necessarily negative, 36,38 and its manifestation and impact may be constructive or non-constructive.
Prior research demonstrates that the impact of power dynamics between higher status and lower status individuals may be moderated by psychological safety and perceived connectedness. 8 Edmondson defines psychological safety as the extent to which people view the work/learning environment as being conducive to interpersonal risk-taking, such as expressing themselves or asking for help, without fear of negative consequences. 7,39 It has been shown that an unconstructive manifestation of power dynamics can be overcome with high psychological safety, even in contexts with strong hierarchies. 40,41

| Research aim
The aim of this study is to explore power dynamics and their impact on intraPC learning between PC residents and MS residents during hospital placements. The intention here is to enhance the understanding of the nature and extent of power dynamics on hospital wards and to pave the way for future constructive collaborative learning and practice.

| Context and design
Worldwide, during postgraduate training, PC residents undertake hospital placements in the same departments where MS residents are in training. 15,[19][20][21][22][23] In the Netherlands, this means that PC residents work four days a week on the hospital ward together with MS residents; the fifth day is spent with other PC residents at the PC specialty training institute. This current study expands on previously published research by Looman et al. (2020), which investigated opportunities and barriers to intraPC learning between PC and MS residents during hospital placements. 15

| Data collection
In our previous study, observations and interviews were conducted at three geriatrics departments and three emergency departments of five Dutch hospitals from February to May 2018. During this study, issues of power and power dynamics repeatedly surfaced in interviews, even when power was not initially addressed by the interviewer. After 15 interviews, we decided to incorporate additional questions to explore this issue deeper in the subsequent 27 interviews. Previous studies on psychological (un)safety in healthcare have recommended taking different power status levels into account, and involving the researcher as an observer in the study setting to observe patterns rather than relying on participants' reports only. 42 We finally used all 42 interviews for this study and included 24 fieldnote transcripts for triangulation. More information on data collection can be found in

| Design
We decided that the issue of power dynamics needed another theoretical framework than the prior study on opportunities and barriers to intraPC. Due to the current focus on power dynamics and the sensitivity required for such a topic, we employed a critical theory paradigm. Critical theory is a research paradigm that focuses on the experience of people and seeks to understand how social structures shape these experiences. 43,44 Critical theory is concerned with issues such as power and justice and tries to explain how social systems function by looking into discourses, ideologies and institutions. 43,45 In line with this paradigm, a discourse analysis approach informed our data analyses. 45,46 Discourse analysis focuses on the relation between language, practice and power 46 and assumes that it is important to analyse power relations from the viewpoint of the participant. 44

| Data analysis
Transcripts of the interviews and fieldnotes were analysed employing a template analysis method. 47,48 Template analysis can be accommodated to different paradigms, 49 in this case critical theory and some discourse analysis elements as an additional way of looking at the data. 46 For example, we used mental models and metaphors to analyse the data on a deeper level. 44 Mental models show what people believe about others. 44 Metaphors can reveal beliefs or norms that are normally hidden. We used mental models and metaphors as a discourse analysis approach to explore the power dynamics in our transcripts and to identify implicit forms of power.
Our data analysis started by selecting the relevant material. We combined an inductive and a deductive approach for the operationalization of power dynamics. Two authors (NL and TW) performed a first round of open coding. NL and TW each independently coded three transcripts. We discussed the results together. Combining these with sources in the literature, we made a preliminary template of power dynamics. We used the preliminary template to select relevant parts of the other transcripts. After that, NL and TW coded six transcripts individually and compared the similarities and differences. Due to different professional backgrounds, we had to settle on some definitions. 'Team dynamic', for instance, was coded when it was negative by NL, whereas TW interpreted it as neutral. We agreed to use it as a negative term and to use work-climate as a neutral or positive term. NL and TW made an initial template and discussed this with the extended team: CF, NS and JdG.
In the second round, NL and TW divided and coded the remaining transcripts individually. Six of the transcripts were again coded by both and discussed in weekly meetings, to keep track of differences and similarities. We discussed and settled on differences by meeting with the whole research team and resolved all inconsistencies through consensus. Differences mainly concerned whether a quote was to be interpreted as neutral or negative, or how to choose a slightly different subcode from a larger overarching category (e.g., hegemony or distance). Other differences could be traced back to the different backgrounds of the researchers, in which case we opted for an inclusive approach and kept both codes (e.g., collaboration and workclimate).
Finally, we double coded the fieldnotes and triangulated these with the findings in the coding template. We looked for mentions of power in the fieldnotes and compared these to what the interviewees had said.

| Reflexivity
NL is a psychologist and PhD candidate in intraPC/intraPE. Working as an psychologist, her focus is on the underlying aspects of behaviour, interaction and equity between people in a work environment.
TW has a background in education science and philosophy. She is a teacher trainer and researcher in medical education. She holds an enactivist approach to learning, focusing on the role of affect and environment in learning. DvA is a geriatrician, supervisor and researcher in medical education. She focuses on team behaviour in the hospital ward regarding intraPC learning between residents. NS is a general practitioner, director of PC specialty training and professor general practice in IPC. Her focus is on the role of PC residents with regard to intraPC learning. CF is an MD and educationalist and professor of workplace learning. Her focus is on creating working environments that stimulate learning for both students and professionals, psychological safety and adaptive expertise. JdG is an internist, director of postgraduate medical education and professor of professional performance in PGME. She focuses on hierarchy, psychological safety and policies that affect intraPC learning.

| RESULTS
Based on our analysis, we defined five interrelated themes that describe characteristics of power dynamics in intraPC learning between PC residents and MS residents during hospital placements: (i) beliefs; (ii) power distribution; (iii) interaction style; (iv) subjection; (v) fearless learning (see Table 1).
The themes appeared to be interacting. The observations and interviews indicated that power dynamics (the way power impacts the interaction between people) occurred both within the themes and between the themes. We described the interrelation between the themes as main types of power dynamics.
We found five main types of power dynamics in intraPC learning between PC residents and MS residents in hospitals (see Figure 1): (i) beliefs impact power distribution; (ii) beliefs impact interaction style; (iii) power distribution based on inequity impacts subjection; (iv) power distribution based on equity impacts fearless learning; (v) interaction style impact fearless learning. We will elaborate on these themes and on power dynamics in this section.

| Beliefs impact power distribution
Our interviews revealed that power distribution is influenced by underlying beliefs and vice versa. Supervisors mentioned that professional norms, such as mastery of knowledge, determine the level of hierarchical status assigned to PC residents.

| Beliefs impact interaction style
Our interviews demonstrated that beliefs impact interaction style, and, similarly, that the way PC residents, MS residents and supervisors talk about and with each other (often in metaphors) can create/ maintain beliefs. Participants mentioned that interaction styles have a major effect on generating a constructive or unconstructive manifestation of power dynamics, which subsequently have a conducive or corrosive effect on intraPC learning (see Table 2).  screen. This is followed by a discussion between 3 MS residents. At 12.55 two more supervisors join the session. They recognise the patient on the screen and immediately get involved in the discussion. The atmosphere is still relaxed but the hierarchy feels less flat than before the three supervisors joined the group. The supervisors intervene quickly and often in the discussion and take over the lead and the residents become more and more silent, sharing their perspectives less and less.

| Power distribution based on equity impacts fearless learning
A prevailing view among participants is that a certain degree of hierarchical power distribution in the medical workplace can contribute to a constructive manifestation of power dynamics. As long as collaboration is based on equity, hierarchical power distribution could foster a work climate that contributes to fearless intraPC learning during hospital placements. As the following residents said: There is a hierarchy, but everyone can quite easily contact each other. It's clear who's ultimately responsible.
They're not vague about it because that would actually hinder a good working atmosphere. That [collaboration] just occurs in a very relaxed way.

PC_resident_D40
We stand above PC residents, but not in rank or anything. It's more that you are really above them in terms of knowledge, but not in how you treat each other or whatever… Look, a PC resident may not treat a neurotrauma, that's a difference of course.
It does not make me feel better or higher.

MS_resident_D7
Our observations and interviews suggest that equity can be promoted by sharing a physical space in which everybody literally stands or sits at the same level during patient discussions.
Previously, we were hierarchically separated in the handover room, but we made a conscious decision to have everyone on the same level during the handover, just to be able to discuss everything face-to-face with each other. MS_supervisor1_D5 I think that's also one of the reasons that the day-start is always done standing up, so that everyone is equal.   There should be no threshold for consultation.

MS_supervisor1_D5
Another way to promote fearless intraPC learning in the hospital ward is to start the workday or team meeting with a personal briefing or by registering a smiley face that reflects the person's mood. Participants indicated that sharing thoughts, feelings and learning goals could support the connection between team members and balance power dynamics.
Yes, we consciously chose this [as a start to teammeetings] because studies have shown that employees feel more valued and you also get better team bonding when you first pay attention to whether everyone is fit and if there's anything we need to take into account.

| DISCUSSION
Many calls have been made in previous studies to examine and address the influence of power on intraprofessional learning. 30,31,45 To our knowledge, this is the first study specifically investigating power dynamics and their impact on intraPC learning between PC and MS residents during hospital placements. Our data showed five themes that describe characteristics of power dynamics: (i) beliefs; (ii) power distribution; (iii) interaction style; (iv) subjection; (v) fearless learning. These themes were found to be interrelated, and power dynamics among residents and/or supervisors occur both within and between the themes. We report five main types of power dynamics in intraPC learning between PC and MS residents in hospitals: (i) beliefs impact power distribution; (ii) beliefs impact interaction style; (iii) power distribution based on inequity impact subjection; (iv) power distribution based on equity impact fearless learning; (v) interaction style impact fearless learning.

| Interaction style and fearless learning
This study indicates that a constructive manifestation of power dynamics can occur when hierarchical power distribution is combined with open interactions and collaboration based on equity. This is consistent with prior research revealing an inextricable link between open interactions and psychological safety. 42

| Fearless learning in action
As healthcare and residency training have a strongly hierarchical nature with associated strong professional norms, 5

| Implications for practice
To manage power dynamics and to facilitate fearless intraPC learning between residents in hospitals, the following ideas might be helpful: Representing the residents' and supervisors' perspective is important for understanding the influence of power dynamics on intraPC learning between residents in hospitals, and it becomes crucial when the goal is to balance these power dynamics in order to foster fearless intraPC learning. This study describes a phenomenon that is often more implicit than explicit; however, this study also demonstrates that not all beliefs, biases and practices are 'hidden'; some are perceptible,

| Limitations
We recognise that there may be more types of interactions between the themes, for example between beliefs and fearless learning or between interaction style plus power distribution that may promote subjection, but these did not emerge from our study. Some inter- viewees were very open about power struggles, while others were holding back. As this research was part of a larger project which had a broader scope than power dynamics alone, we may have missed depth or an opportunity to break through interviewees' hesitations.
As the analysis shows data saturation, however, we feel confident about our results.
Triangulation with observations, moreover, helped to gain insight into who were holding back and to get ideas about why this might be the case or what was actually happening in the workplace. Still, it is important to remember that power is a taboo subject, and it may have been difficult for interviewees to really speak up.

| Future research
Further research is needed to determine whether and how the listed implications for practice will help to improve fearless intraPC learning.
Future studies could focus on using a phenomenological approach in the interviews to really understand the interviewees' perspective. As the topic of power dynamics remains a taboo subject, we recommend focusing on trust before the interview and including metaphors to get an idea of actual beliefs. Based on our experience, we recommend triangulation with observations, because this could be helpful in understanding whatever is not mentioned in interviews.

CONFLICTS OF INTEREST
No competing interests.

ETHICAL APPROVAL
We

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.