Chances for learning intraprofessional collaboration between residents in hospitals

Abstract Context Intraprofessional collaboration (intraPC) between primary care (PC) doctors and medical specialists (MSs) is becoming increasingly important. Patient safety issues are often related to intraPC. In order to equip doctors well for their task of providing good quality and continuity of care, intraPC needs explicit attention, starting in postgraduate training. Worldwide, PC residents undertake a hospital placement during their postgraduate training, where they work in proximity with MS residents. This placement offers the opportunity to learn intraPC. It is yet unknown whether and how residents learn intraPC and what barriers to and opportunities for exist in learning intraPC during hospital placements. Methods We performed an ethnographic non‐participatory observational study in three emergency departments and three geriatric departments of five hospitals in the Netherlands. This was followed by 42 in‐depth interviews with the observed residents and supervisors. The observations were used to feed the questions for the in‐depth interviews. We analysed the interviews iteratively following the data collection using template analysis. Results Hospital wards are rich in opportunities for learning intraPC. These opportunities, however, are seldom exploited for various reasons: intraPC receives limited attention when formulating placement goals, so purposeful learning of intraPC hardly takes place; residents lack awareness of the learning of intraPC; MS residents are not accustomed to searching for expertise from PC residents; PC residents adapt to the MS role and they contribute very little of their PC knowledge, and power dynamics in the hospital department negatively influence the learning of intraPC. Therefore, improvements in mindset, professional identity and power dynamics are crucial to facilitate and promote intraPC. Conclusions Intraprofessional collaboration is not learned spontaneously during hospital placements. To benefit from the abundant opportunities to learn intraPC, adjustments to the set‐up of these placements are necessary. Learning intraPC is promoted when there is a collaborative culture, hierarchy is limited, and there is dedicated time for intraPC and support from the supervisor.


| INTRODUC TI ON
Adverse events resulting from human error are reported frequently in health care. 1,2 A common contributing factor to these events is an ineffective collaboration between primary care (PC) doctors and medical specialists (MSs). 3,4 Frequently reported issues are deficient communication and information transfer. 4,5 These problems could increase in many health care systems because of the current tendency to translocate part of health care service provision from hospital to PC settings. 6 This involves transitions of both patients and knowledge, leading to an increased risk of error. 3,4,7 Therefore, in addition to being proficient in their professional work, PC doctors in the PC setting and MSs in the hospital should be aware of each other's context, expertise and roles, and how to communicate and collaborate intraprofessionally. 8,9 In order for doctors to be well equipped for their task of providing continuity of care between PC and the hospital setting, intraprofessional collaboration (intraPC) needs to receive special attention during postgraduate training. 10 This can be realised by intraprofessional education (intraPE). 11 However, the distance, both physical and conceptual, between PC and MS workplace and teaching environments seems to be a deeply rooted obstacle to this strategy. 8 During postgraduate training, PC residents and MS residents do collaborate around referral to and discharge from the hospital.
Their training programmes, however, take place in isolation from each other and focus on their own specialties. 8 In the Netherlands, learning during PC and MS postgraduate training is predominantly workplace based. Both curricula and clinical commitments limit the time PC residents and MS residents can work together. 10,11 As a result, the opportunity to build on and learn from and about the strengths of each other is limited. Because the proximity of different professions in shared educational and clinical spaces and sufficient time allocation can help to build mutual rapport, 12 it is precisely the proximity that requires specific attention when organising intraPE.
We explored whether and how intraPE could be organised during hospital placements. In many countries, PC residents, such as general practitioner residents and elderly care physician residents (see Box 1), undertake a hospital placement during their postgraduate training. 13,14 This hospital ward, where PC residents and MS residents work in proximity, offers the opportunity to learn intraPC through intraPE. Currently, formalised intraPE is limited; therefore, if learning of intraPC occurs, it will be predominantly unintentional. 8 To the best of our knowledge, there has not been investigation of whether and how PC residents and MS residents learn intraPC during these placements.
This study aims to gain insight into the potential of hospital placements for learning intraPC, by answering the following questions: (a) When and how do PC residents and MS residents learn intraPC during hospital placements?, and (b) What are opportunities for and barriers to learning intraPC during these placements?

Box 1 Definitions of professionals and settings within the Dutch health care system
General practitioner (GP) Doctor 'working in the frontline of a healthcare system, taking the initial steps to provide care for any health problem(s) that patients may have […] including prevention, diagnosis, cure, care, and palliation' 15 Elderly care physician (ECP) Doctor working in long-term care for elderly people and chronic patients, mostly in a nursing home. In the Netherlands this is a PC specialty 16,17 Primary care (PC)-setting

| ME THODS
We carried out a constructivist ethnographic study. A constructivist approach acknowledges that researchers' background assumptions, disciplinary perspectives and programmatic efforts along a line of study shape their research processes and conceptual emphases.
Therefore, in our study, particular time and attention were paid to reflexivity throughout the research process on how our assumptions and perspectives have shaped our data collection and interpretation.
The research group consisted of general practitioners, educational scientists, a psychologist, an internist, a geriatrician and a medical student. All group members were experienced in providing intraPE and/or conducting research into intraPE in different contexts. This multidisciplinary research group functioned as a form of triangulation as it brought together disciplines whose profession or training calls on highly different assumptions and knowledge areas. 19,20 An experienced psychologist (NL) and a medical student (MvW) performed the observations and interviews. Both researchers were trained in qualitative methods and analysis. For the ethnographic research, these researchers were trained during this study by an anthropologist and an educational science researcher.

| Rapid ethnography
We used a non-participatory rapid ethnographic research approach. 21

| Study setting and inclusion
Using purposeful sampling techniques, we sampled emergency departments and geriatric departments of both academic and regional hospitals in the Netherlands. After inclusion, we announced our visit with posters and emailed an information letter explaining the purpose of our study, including an invitation for the interview to all residents and supervisors. For the interviews, we applied purposive sampling, including snowballing. We sampled younger and older residents and supervisors and we talked about the results with participants. This allowed us to gather broad and deep information on learning intraPC during hospital placements. We excluded residents and supervisors who worked in the hospital department for less than 1 month.

| Data collection
Data collection through observations and in-depth interviews was

| Data analysis
Transcripts of the interviews were analysed using a template analysis method. 27 We chose template analysis as in this way we could handle the large dataset more comfortably than some other methods of qualitative data. 28 The use of a priori themes within template analysis helps focusing on themes that need to be incorporated into the analysis. A first template was developed by NL and MvW. The codes of this preliminary template were derived from the main questions from our interview guide but also arose from inspection of the data. 27 After each day of observations and interviews, the researchers (NL and MvW) discussed their findings. The first three interviews were coded by two researchers (NL and MvW), leading to an initial coding template. After rereading our data and discussing our template we decided to use this template as this would represent the data as fully as possible.
It contained higher level codes (representing major themes) and low to lower-level codes, representing more specific topics. The

| Ethics
This study was reviewed and approved by the NVMO (Netherlands Association for Medical Education) Ethical Review Board (NERB dossier number 983). Written informed consent of all participants was obtained before participation. In some cases, nurses, (para)medical professionals and medical students were visible during the observations and therefore they were asked for informed consent to be observed, after receiving an information letter on the day of our observations.

| RE SULTS
We conducted 45 hours of observations (10-360 minutes per observation) and 42 interviews (

| Theme 1. Incidental and purposeful learning
Our data showed that learning intraPC on the hospital ward occurs by two routes: (a) incidental (implicit learning activities), and (b) purposeful (explicit learning activities).

| Learning implicitly and incidentally
The majority of intraprofessional learning activities occurred im-

| Learning explicitly and purposefully
We observed that intraPE is purposeful and planned in some departments, especially in departments with a collaborative culture, dedicated time for intraPE and intraPE mindset of the supervisor (see Box 2).

| Role of supervisor in purposeful learning
The PC residents indicated that some supervisors consciously stimu-

| Placement goals
Both PC residents and MS residents indicated that learning intraPC is essential, but they are not always aware of opportunities to learn

| Theme 3. Work environment
A prevailing view amongst participants was that learning intraPC between residents is only possible when a safe work-learning climate and significant practicalities are secured.

| Work-learning climate
We observed that the placement of residents within the room during team meetings can reflect (in)equality. Within some departments, everybody was seated equally in the room. In other departments, PC residents were not sitting around the (PC-resident3_H1) Participants mentioned that supervisors are in the position to steer power dynamics, and within some departments supervisors showed an active policy against unconstructive power dynamics.

| Practicalities
Intraprofessional education can hardly take place when PC and MS residents are working in different shifts or having different offices.
Supervisors and residents indicated that the opportunity to meet each other is necessary for intraPE to take place. This is possible by sharing physical space together.
We are in a set-up in which we sit in a circle (behind computers) and where you easily pick up things from each other. And then an interesting (intraprofessional) discussion, a case-based discussion arises spontaneously. (Supervisor2_H5)

| Residents' and supervisors' perceived needs
In relation to the above themes, residents and supervisors men-

| D ISCUSS I ON
All participants found intraPC essential for good health care and consider hospital wards to be rich in opportunities for learning in-traPC. However, we also report that these opportunities are seldom exploited for various reasons. First, intraPC learning goals are often not apparent and both residents and supervisors lack awareness of the intraPC learning opportunities. When learning intraPC occurs, it is predominantly implicit. Second, PC residents often adapt to the role of MS resident and hardly ever share their PC expertise. The MS residents often neglect to search for PC expertise. Third, too much hierarchy led to inequity, which had a hindering effect on building relationships and formed a not-safe-enough work-learning climate in which residents did not feel free to speak up. Therefore, improvements in mindset, professional identity and power dynamics are crucial to facilitate and promote intraPC.

| Mindset
When learning of intraPC occurs between PC residents and MS residents, this is mostly random through informal mechanisms: the learning occurs implicitly, spontaneously and with little conscious reflection, which is in line with the description by Watkins and Marsick of informal and incidental learning. 31 To our knowledge, our study is the first to investigate intraPE during hospital placements.
Our findings are consistent with previous studies in other contexts, which also showed that learning of collaborative competences lacks structured implementation and is generally not in the mindset of medical professionals. 11,32,33 Residents are expected to learn during their postgraduate training and, therefore, it could be expected that they are always on the lookout for learning opportunities. However, with regard to intraPC, this happens only to a limited extent.
Frequently, mindset is associated with the growth mindset theory from Dweck. 34 However, in social psychology and organisational leadership, mindset is seen as a cognitive filter through which one looks at the world, a predefined reference frame, 'used throughout the totality of an individual or organization's cognition.' 34 Johnston clearly recasts a long-standing idea when she states that 'excellent medical education occurs in secondary care settings' 7 and elaborates that primary care has an 'inferior status' 7 and is considered to be much less advanced. Consequently, MS residents teach PC residents, but they are not accustomed to asking for PC expertise from PC residents, maybe not realising or appreciating their PC expertise. The MS residents rarely have placements in PC settings. These historical patterns can lead to a mindset for predominantly unidirectional learning at the workplace. Uhlig et al described that, in order to successfully realise interprofessional collaboration, many deeply rooted patterns, role cultures and assumptions must be carefully adjusted. 35 Our results underscore that MS supervisors and PC teachers have an important role in creating a mindset for learning intraPC. They can do this by formulating placement goals for both PC residents and MS residents and by stimulating two-way learning and conscious reflection. 36,37 The above indicates that intraPE is the responsibility of all parties involved: PC residents; MS residents; supervisors; teachers, and programme directors.

| Professional identity
In the Netherlands, the purpose of hospital placements for PC residents is to gain expertise in emergency care and diseases that are not very prevalent in a PC setting and to learn intraPC with medical specialists. We found that PC residents often adapt to the role of MS resident. This is useful for learning medical skills and fitting into the hospital team. However, the majority of PC residents hardly ever share their PC expertise. This is counterproductive for learning intraPC. At first glance, the PC resident appears to have little influence on the dynamics of an expert team within the hospital ward.
However, our results show that also temporary team members can bring a fresh eye to common practices. We found that PC residents who expressed their professional PC identity and easily alternated between the MS role and PC role, created intraPC discussions and bidirectional learning. Previous literature shows that pre-existing teams are more receptive to the influence of newcomers when the newcomers are more assertive. 38 Proactive PC and MS residents would also rapidly take charge of their intraPC learning process once they are included in the learning cycle. 31 This stresses the importance of empowering PC residents to express their professional identity and to proactively share their PC knowledge, and empowering MS residents to proactively ask for PC knowledge.

| Power dynamics
The participants mentioned that hierarchy is useful for clarifying roles and responsibilities within the hospital, but too much hierar-

| Strengths and limitations
A strength of this study is the use of four types of triangulation: (a) method; (b) data source; (c) investigator, and (d) research group triangulation. 45 An interprofessional research group brought together disciplines with highly diverse assumptions and different knowledge bases, 19,20 and triangulation allowed researchers to examine different data sources to confirm and contrast findings. 45 The psychologist, for example, had a keen eye for the possible effects of adjusting hierarchy and the general practitioner focused on elaborating the importance of sharing PC expertise.
We consider the variability in the nature of the observations as a strength. The short observations consisted of five meetings lasting less than 15 minutes. These were meetings to start the day in an interprofessional way. Although short, these meetings provided us with very rich observations with respect to (opportunities for) intraPE. Because our observers were familiar with the context of hospital placements they could easily recognise relevant activities.
Another strength is the cooperative attitude of residents and supervisors in participating in this study; we had to cancel some hospitals, which had applied to participate, after conceptual depth was reached.
Because of this cooperative attitude, we could get a rich conception of the potential of hospital placements for learning intraPC.
We acknowledge several limitations. Our presence during observations may have had an impact on the participant reactivity, which is defined by Paradis and Sutkin as: 'a form of participant effect that comes from participants' active engagement with the research and its aims, leading to behavioral adaptation that aligns with perceived social norms.' 46 We think we minimised participants' reactivity by being embedded in the environment and checking our observations during the in-depth interviews with the participants. 23,46 Observers were dressed in a hospital uniform and we undertook at least four observations in every hospital department. We noticed that people did interact with us as if we were new colleagues and continued their actions seemingly uninterrupted, especially when we revisited departments. Another limitation is that we only performed observations in locations where no patients were involved. Therefore, a part of informal learning intraPC remained outside the scope of our study. By practising reflexivity in an interprofessional research group, we think this limitation was reduced as much as possible.

| Implications for practice and future research
When organising the learning of intraPC through placements for residents from different medical backgrounds, we think the following should be kept in mind. First, informal learning can be planned or unplanned, but it involves at least some conscious reflection. 31 It is necessary to implement intraPE within workplace-based learning, to make the learning of intraPC purposeful.
Second, the hierarchy must be taken into account; for example, by sharing a room and sitting equally around the table, asking for different perspectives, and letting PC and MS residents speak first during discussions and then letting supervisors add their information. Third, supervisors need extra training to be aware of and create learning opportunities and to create a mindset for learning intraPC. Finally, residents need some level of professional identity to be able to show their expertise and for supervisors to steer intraPE. A professional role identity is developed from a combination of personal factors, the working environment and role modelling. [47][48][49] However, PC role models are absent during hospital placements. Therefore, peer-to-peer meetings during placements could be a valuable alternative. 49 We recommend release days, where PC residents learn about having a dialogue with their peers about intraPC. Future research is needed to investigate how the development of professional role identity can be supported, and how power dynamics can be managed in a constructive way.

| CON CLUS IONS
All residents and supervisors indicated that learning intraPC is essential and requires more explicit attention. IntraPC is not learned spontaneously during hospital placements. Even in a promising setting where PC residents and MS residents work together in the same department, intraPC receives at best limited attention as a competency to be learned. The MS residents are not accustomed to asking for PC expertise and PC residents often adapt to the role of MS resident and they hardly ever contribute their PC knowledge. Hierarchy and a lack of psychological safety in the hospital department negatively influence the learning of intraPC. We conclude that in order to benefit from the opportunities to learn intraPC during hospital placements, attention to mindset, professional identity and power dynamics is needed. Learning intraPC is promoted when there is a collaborative culture (with not too much hierarchy), dedicated time and goal setting for intraPC and support from the MS supervisor on the ward and PC teachers during release days.

AUTH O R CO NTR I B UTI O N S
All authors (NL, CF, MvW, EdG, PD, DvA, JdG and NS-dH) met all of the following conditions: substantial contributions to the conception and design of the work; the acquisition, analysis and interpretation of data for the work; drafting the work and revising it critically for important intellectual content; final approval of the submitted paper and the version to be published; and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy and integrity of any part of the work are appropriately investigated and resolved.

ACK N OWLED G EM ENTS
We wish to thank Dr Marieke de Visser-Oomen (Department for Research in Learning and Education, Radboudumc Health Academy, Nijmegen, the Netherlands) for advising us during the coding and analysing process.

CO N FLI C T S O F I NTE R E S T
There are no competing interests.