Examining Academic Leader’s work in implementing Competency-based Medical Education using Organizational Learning Theory

Context Competency-based medical education (CBME) implementation is being carried out in many medical schools worldwide. Academic Leadership is a strategy where selected Faculty act to influence peers to adopt change. The Université de Montréal medical school, has adopted this strategy to implement CBME. Purpose This paper aims to describe the work of Academic Leaders in the process of CBME implementation and to explore relevance of the Nonaka and Toyama organizational learning theory to map implementation progress. Method Because knowledge creation model focuses on the relationships between leaders and social structures, embedded case study was selected. Diverse sampling method was used to select three departments: internal medicine, surgery and psychiatry, based on the number of CBME training activities. Data collection was at two intervals, two years apart. Semi-structured interviews (individual and group) were conducted with Department Heads and Academic Leaders. Thematic analysis was conducted on the 15 interview transcriptions. Results As implementation begins, Leaders critically revisit accepted teaching routines and develop a common conception of CBME. This enables leaders to communicate with a wider audience and work within existing committees and working groups where they “break down” CBME into practical concepts. This practical understanding, disseminated through Entrustable Professional Activities, enables observable change. Conclusion Leaders’ roles evolved from an “expert” that disseminates knowledge about CBME through lectures, to a responsive and pragmatic supporting role by developing and writing practical tools in collaboration with peers and program directors.


Introduction
One the difficulties medical schools have encountered in implementing Competency-based medical education (CBME) stems from the fact that CBME frameworks describe wide scoping curricular outcomes with little guidance to instructors on the ground on how to teach and assess competencies (1)(2)(3)(4)(5)(6). This lack of practical guidance generates much resistance towards CBME implementation (7) and little attention has been paid in the scientific literature to organisational strategies employed to attenuate the resistance to change. This study describes how Academic Leadership can bridge the gap between CBME concepts and practical tools that instructors can use on the ground.

Leadership in Higher Education
Studies in the field of leadership in higher education suggest that leadership in academic organizations is shifting towards a cultural perspective, based predominantly on relationships (8): 'Without doubt, the key to being an effective leader is an ability to connect with others in the organisation and gain their cooperation in working collaboratively towards the organisational goals and objectives' (p. 29). Another perspective, that of transformational leadership, suggests that leaders communicate a vision, provide support, empower others to innovate and lead by example (9).
Among the strategies relied upon to support innovation in Higher Education (10) and CBME implementation in medical schools is one that rests on Grassroots Leaders (11) or Academic Leaders (12,13). This approach consists of training a select group of faculty, generally interested and motivated by the vision of change promoted within the institution. The training can take the form of formal lectures on topics, group mentoring with an Educational Expert, or even graduate studies in medical education. Although Academic Leaders are generally not in a position of authority over their peers (14), they are expected to influence, motivate and inspire colleagues in adopting and successfully implementing change within a university department (15)(16)(17). This study proposes a theory guided exploration of the work of such a group of Academic Leaders as they work towards CBME implementation in a large medical school.

Organizational change as knowledge creation
In our study, CBME implementation is construed as organizational change (18) within the medical school. We turned to the organizational development field that seeks to uncover levers for change (19,20) provide a systematic perspective on the process. Some organizational change frameworks focus on power relationships and structures within the organization (21), others on the organization's identity and interactions with the surrounding environment (22,23) and still others on communities of practice (24,25) and developing new knowledge (26). In contrast to the former, this latter perspective construes the individuals in an organization as knowledge creators.
A medical school's decision to implement CBME does not come with clear definitions of how teachers are to conduct their teaching nor how they are to assess students' competencies. New knowledge, generated from testing new conceptions of teaching and assessing, is required to solve the problem. As medical instructors experiment, exchange and discuss their insights with their colleagues, they embark on the process of organizational knowledge creation.
Tacit knowledge (27,28), which hinges on the idea that an individual knows more than they can tell, is a key concept. Individuals develop personal insights into unique ways to conducting their work, which is reflected in the dual consciousness that characterizes humans at work: A practical consciousness, embedded in tacit knowledge, which informs what they do and a discursive consciousness that informs what they say (21) that is shared with others. Knowledge creation is conceptualized as the process of narrowing the gap between practical and discursive consciousness, or making tacit knowledge explicit. Unless there are overt efforts to facilitate this process, individuals' tacit knowledge remains an untapped resource.
Hence, the challenge in implementing CBME is making the tacit knowledge held by instructors explicit and available to colleagues within the organization. This process can be referred to as organizational knowledge creation which rests on the idea 'that knowledge is created and expanded through social interaction' (29)(p. 61). This process, also referred to by Nonaka and colleagues as knowledge conversion, cannot be understood as a linear process, but as a spiral that expands organizational knowledge and distributes from individual, to group, to organisation and interorganisational levels (30).

The knowledge creation process
It is implied that knowledge creation is not an individual activity and that it occurs within a given space and time, what Nonaka calls ba (30). Ba is a shared mental space and time where information is shared and interpreted collectively. The knowledge creation process is comprised of four stages that unfold within the ba: socialization, externalization, combination and internalization (SECI) which reflect a cumulative upward spiralling process.
The socialization stage emerges through day-to-day experiences in teaching medical students in a clinical setting. For example, medical educators become aware of contradictions in the way teaching is carried out and talked about. Talking to peers brings these contradictions to the fore.
The socialization stage ends when individuals embrace actions to 'resolve these contradictions' During the externalization stage, individuals use their discursive consciousness to rationalize and articulate the contradictions they have encountered. Nonaka and Toyama insist that at this stage, individuals 'seek to detach themselves from routines' (p.4) to eventually put new ones in their place. Practical knowledge is actively shared with peers in an effort to form new concepts. The externalization stage ends when these new concepts becomes clear enough to be practical.
In the combination stage concepts are tested and disseminated to other members of the organization. This stage is where leaders 'break down' concepts, such as CBME, so that peers can develop their understanding of them in their discursive consciousness and make sense of them. The combination stage ends when peers take ownership of the new concepts and introduce them in their practice.
In the internalization stage newly created concepts generate new practical knowledge. The change in the organization becomes visible as a greater number of members adopt change and new routines take hold. This stage ends when individuals begin to critically question theses routines and a new cycle of knowledge creation begins. Table 1 summarizes the four stages of the knowledge creation model. Articulating practical knowledge through dialogue and reflection (discursive consciousness).

Internalization
3. Combination Embodying discursive knowledge through action and practice (practical consciousness).
To summarize, an organization creates knowledge through the knowledge conversion process (SECI) which occurs in a specific time and space (ba) in the interactions between individuals and groups.

Study aim
Some institutions of Higher Education make use of Academic Leaders to catalyse progress towards organisational change (32,33). Yet, we know little about the ways in which Academic Leaders in universities develop and sustain the conditions and processes that generate change.
In our study, Academic Leaders, drawn from medical school faculty, were trained in CBME and were expected to guide and support their peers. How do they articulate their ideas of change and ultimately persuade peers to adhere to them?
Our study aimed to describe Academic Leaders' work through the organisational knowledge creation process : socialization, externalization, combination and internalization. We were interested in identifying factors that enabled or hindered the transition from one stage of the process to the next. This would yield deeper insight into how Academic Leaders contribute to enacting change in higher education institutions.

Methods
The largest francophone medical school in North America has undertaken CBME implementation and has chosen a strategy based on Academic Leaders. Academic Leaders' mandate is to facilitate CBME implementation by supporting peers, developing and delivering training activities and advising Program Directors. Faculty of the medical school is made up of approximately 3000 instructors (roughly 20% tenured, 80% part-time or non-tenured), working in over 100 different teaching sites (hospitals, family medicine and community clinics, etc.) and offering 73 programs, from the undergraduate MD program to 72 graduate (residency and specialty) programs. The Faculty comprises 16 academic departments as well as two health sciences schools. Leaders who step-up from the ranks were trained in CBME and offered remuneration for their work.
In order to achieve the aims of the study, a case study approach with embedded levels (34, 35) was selected. The unit of enquiry is thus not the individual leaders, but the academic department to which they belong. This approach allowed observation of the interactions between the leaders and the social structures in which they work, which is a crucial dimension to the organizational knowledge creation model. We focused on leaders' interactions with Department Chairs and with colleagues.
Departments were selected for intense study using a purposive sampling method (35)(p. 88). The small number of departments (N=16) makes random sampling unreliable because there is no way to ensure that these cases are representative of a larger population. In order to meet the goals of reproducing the relevant causal features assumed to be in the larger population and providing sufficient variation along the dimension of interest (CBME implementation), the diverse case sampling method was used (35)(p. 97). Hence, the departments were selected on the basis of a range of levels of CBME training activities in 2012 roughly equivalent to high, medium and low The three Department Chairs were senior physicians (cardiologist, urologist and psychiatrist). All three had been involved in medical education, as instructors, program directors and Department Chairs for more than 10 years. When CBME implementation was announced in 2010, all three were already Department Chairs.
The 12 Academic Leaders interviewed were mid-career medical educators who demonstrated an interest for pedagogy. Two of the Academic Leaders had a graduate degree in medical education, the rest had gone through all the relevant CBME training sessions. One of them had been program director for many years and regularly facilitated training sessions on education for first year residents.

Data collection and analysis
In order to apprehend the evolution of the work of Academic Leaders in CBME implementation over the two-year period of the study, two waves of data collection were carried-out. All three All interviews where held on the main campus of the university. In the internal medicine department, two of the three leaders were the same at each data collection wave. Hence, 4 different academic leaders were interviewed in this department.
Thematic analysis was selected to code and identify common themes embedded in the transcripts of all 15 interviews (36,37). Analysis was performed in three steps. In step one (initial review), one member of the research team (NF) identified and described themes that emerged from the data set. To limit the effect of our own biases, in step two, all themes were examined independently by two other members of the research team (NL, DN) to label and categorize each extraction until theme saturation was achieved. To complete step two, the research team reviewed the independently created themes and discussed their interpretation according to the research questions. Disagreements were resolved through discussions until a consensus was achieved. In step three, all codes were reviewed for accuracy and consistency with regards to the themes that emerged.
Ethical approval was sought and granted from the University Research Ethics Board (12-079-CERES-D).

Results
In the following section we present the results of our analysis of the interview data of 2012 and 2014, revealing a more in-depth view of the work of Academic Leaders within the four knowledge conversion stages.

CBME Implementation in 2012
Since 2005, school administrators had been announcing that CBME would be implemented and that funds were being set aside for the Academic Leaders in order to achieve this. In 2010, some Academic Leaders signed contracts with their Department Chair and the Vice-Dean for Continuing Medical Education (CME) that stipulated that they were to assist in Faculty development by supporting CBME implementation. work. They recognized that a training session given by them could attract many Faculty members, but the participation rates observed at the time weren't high enough to fully justify the investment.
Leaders themselves realized that the same people, interested in CBME, were recurring participants in their lectures. There still was a substantial number of Faculty who conveniently avoided lectures on the topic.
This rather pragmatic outlook contrasted sharply with the dynamic view held by Academic Leaders in 2012. We have to underline that the three internal medicine Leaders interviewed in 2012 were already convinced of the relevance of CBME. Their engagement in the program reflected a desire to improve teaching practice and manifested itself in the assumptions they challenged and the contradictions they singled out.
This initial interest was seen as crucial, but other elements emerged that facilitated or hindered view of how Academic Leaders were going to achieve CBME implementation.

CBME implementation two years later
In 2014, Academic Leaders were active in all three departments. In the following section, we present the results of the second set of interviews, conducted between March and April of 2014.

The Internal Medicine Department
As presented in the precious section, the socialization stage in the Internal Medicine Department was well underway in 2012. Active Academic Leaders had identified the reasons they wanted to get involved and they had started to identify a common vocabulary and goal. Another indication of the combination stage is the recognition that Academic Leaders received from within their Department. They became known as people who know about CBME 'people willing to help, willing to motivate people.' (IM-FG/2014). Even residents (graduate students) come to Academic Leaders when they have a problem with teaching. They come to see them because they know they've been working on CBME but not because they are Academic Leaders. They were also involved in larger initiatives for CBME implementation such as defining the competency framework and participated in meetings with colleagues from other departments. They had also introduced themselves to the academic committees at the psychiatry teaching hospitals: 'It was really to train people on competencies in all programs so that there could be people who piloted the project somewhat….but I think that at that moment there wasn't a Here Academic Leaders signal a change in the way they work, based on listening to what colleagues have to say far more than trying to exhort them to change. Leaders expressed that the key was being present in committees, whether they be university committees or hospital committees. 'The aim is to get people thinking of pedagogy.' (PS-02/2014) As CBME implementation progressed, Academic Leaders felt that they were making a difference.
At the very least they were under the impression that CBME is getting simpler and concrete. They The first issue brought up by the Academic Leaders is the need to bring solutions to the teaching sites. This was seen as central to their work; their role is to help Faculty find new and alternative solutions to problems they encounter.
'everybody sees that there are gaps here and there, but they don't necessarily put in any requests [for assistance] because everyone does things the way they have been done traditionally since…you know, the same way…'(SU-01/2014) Leaders in the Surgery Department also equated CBME implementation with an opportunity to improve teaching practice. Surgery Department Academic Leaders were aware from the outset they needed to work collectively on issues that didn't coincide with their personal interests, but rather emanated from above, and were part of the general 'top-down' initiative to implement CBME.
They said that other colleagues had refused the invitation to become Academic Leaders because they didn't adhere to the project 'it's not true that they'll make me talk about [pedagogy], only because they want me to talk about that.' (SU-02/2014) This shift to a collective approach to CBME implementation marks the passage from the socialization stage to the externalization stage.
The externalization stage in surgery is also marked by resolution of issues of legitimacy. Some Leaders in surgery clearly had doubts about their ability to be Academic Leaders; 'it's not easy to be convinced that we can do it, that we are the right person that can necessarily do it because we took the week-long training.' (SU-FG/2014) 'What we need is that leaders be recognized in the different settings, especially in the programs, so that precisely, people come to us.' (SU-FG/2014) At issue was the Academic Leader role, not in a position of formal authority, but a sort of expert consultant whose advice you can ignore if you wish: 'So, the leader is something new, it's not a program director, it's not a division chief, it's someone who will tell us that we have to teach more, in another way; who is going to give us work to do but we don't know where he's coming from and what title he has really. It wasn't easy to deal with this.' (SU-FG/2014).
In 2014, recently hired Academic Leaders realized that they needn't bother explaining the CBME concepts to their colleagues, but they had to make them workable. information about pedagogy could be shared to all instructors in the Department.

The internalization stage in the three Departments
By 2014, Academic Leaders and Department Chairs in all three departments recognized that few people questioned the need to implement CBME anymore. Indeed, discourse on CBME was rarely cast in a negative light. Pedagogy became a recurring topic in most Department Continuing Medical Education meetings and yearly Pedagogy Days are offered by the departments explicitly focused on writing competency assessment tools. However, EPA implementation was just beginning and was far from becoming a widespread tool for teaching and assessment at the medical school.

Discussion
This study presents a dynamic portrait of how the work of Academic Leaders for CBME implementation evolved over time. From an intuitive approach focused on 'training' colleagues on CBME, they adopt a collaborative approach, working within local academic committees and colleagues in positions of formal authority. This portrait affords unique insights into how Academic Leaders contributed to change.
At the socialization stage, CBME was an abstract concept and Academic Leaders simply relied on their intuition to 'inform' colleagues about it. As time passed, Academic Leaders realized that this approach was not the most effective. The realization that their target audience would not be reached this way, was key in their changing their approach. In order to gain traction among a greater audience, they reached out to create alliances. The sharing of their tacit knowledge, stemming from their questioning of assumptions, became crucial at this stage.
During the externalization stage, Academic Leaders came together with others who had received the same CBME training. This allowed them to acquire a common vocabulary, to share teaching materials and insights. CBME was becoming much more familiar to them and they felt more confident when explaining it to peers. Consequently, they were recognized as CBME experts within their Departments, and this consolidated their positions within the community and strengthened their resolve to persevere.
The transition from the externalization stage to the combination stage appears as a crucial step.
Beside institutional supports (e.g. CAPHS logistical support, time off and explicit vision from Chairs expressed in Department meetings), Leaders' work was clearly facilitated by the introduction of EPAs. Academic Leaders saw the value of this tool for CBME implementation and became key contributors for their development with colleagues. Another crucial step ensued: Academic Leaders worked together with Program Directors or Pedagogic Committees to develop Department specific EPAs with peers. At this stage, Leaders perceived that CBME was becoming the norm and that resistance to it was diminishing.
It is undeniable that CBME implementation was facilitated in large part by clear institutional support, the appearance of EPAs and the fact that CBME was required by the national accreditation agency. What our study focused on is how Academic Leaders carried this process within their communities. Their work in this context can best be summarized by : collaborating among each other to strengthen their mastery of abstract notions inherent in collective project, securing recognition as experts from peers, transitioning from a role of 'preacher' to one of 'advisor', and actively participate in the writing of tools and teaching materials within departmental committees.
In sum, by questioning current teaching practice and engaging with peers at first and then with Faculty members in positions of authority and committees, Academic Leaders figured out how to persuade their peers to adopt CBME. These results provide evidence that the work of Academic Leaders is to translate vision statements and goals adopted by institutional authorities into workable pieces of information that make practical sense to their peers.

Study limitations
In order to examine the evolution of the Leader's role and to test the use of the organizational knowledge creation model we selected the case study approach with embedded levels. This meant that the individual differences amongst individual leaders themselves were not taken into account.
The cost is that factors such as personality, medical specialties, hospital settings, etc. that could play a role were overlooked. These factors will have to be the focus of further research on the topic.
However, focusing expressly on the relationships between leaders and the social structures in which they act is of sufficient value to offset the cost. The impact of these relationships on organizational change cannot be underestimated and their examination can provide valuable insight on how Academic Leaders contribute to it. Also, our selection of the diverse sampling method allows us to be confident that the results constitute a valid and credible portrayal of the evolution of Academic Leaders' work in the implementation of competency-based education in institutions of higher learning.

Conclusion
Implementing CBME is considered an important organizational change for medical schools.
Getting Faculty to envision such change is a daunting challenge. Given the nature of academia, in which Faculty enjoy academic freedom, the challenge appears unachievable, even if improving teaching practice becomes a requirement imposed by accreditation agencies.
Academic Leadership is a concept that has been introduced as a strategy to overcome these challenges. This study illustrates one such example where Academic Leaders were trained and supported to convince their peers that change was necessary and possible. In order to map these changes, the organizational knowledge creation framework yielded descriptions of stages in the evolution of the work over two years that allowed us to identify specific factors that facilitated change. We hope that Higher Education Institutions authorities and researchers engaged in reforming their organizations will find it useful to use this framework to track progress towards change.