Conceptual and practical challenges for implementing the communities of practice model on a national scale - a Canadian cancer control initiative

Western health care systems are challenged to provide accessible, relevant, and integrated services. To confront these challenges, publicly-funded health systems must find effective ways to increase inter-professional and inter-regional collaboration so that expertise can be enhanced, disseminated, and more effectively translated into practice.

Recently funded by the federal government, the Canadian Partnership Against Cancer (CPAC) is an innovative response to the need for forward nation-wide strategic planning to help the Canadian population meet the challenges that will accompany the growing burden of cancer. Because cancer program delivery, like the rest of health care in Canada, is the responsibility of different jurisdictions, such as provincial governments, national solutions rely on fostering inter-provincial collaborations. The Partnership has thus embarked on an ambitious agenda that recognizes the need to develop both technological and social platforms to facilitate and nurture sustainable pan-Canadian collaborative efforts. In this context, the Partnership has been exploring the utility of Etienne Wenger's "communities of practice" (CoP) concept. The Partnership's CoP Project was conceived to identify and evaluate the effectiveness of communities of practice to enhance collaboration and knowledge translation in cancer control across Canada. Here we describe what we have learned from preliminary stages of the CoP Project, and offer further insights into this promising model of collaboration.

Since the early 1990s, CoP-type structures have been implemented primarily in the business and development sectors; however, their application to health care is a more recent consideration. In this paper, we identify the potential application of Wenger's concept to health care settings, based on a literature review and our experience with the CoP Project. Our aim is to gain a more precise understanding of communities of practice, including their capacity as a form of social infrastructure to further knowledge translation and sustain collaborative efforts across traditional health care jurisdictions and disciplines. The terms collaboratives and groups are used generically throughout the paper to denote interactive work groups.

"Communities of practice are groups of people who share a concern, a set of problems, or a passion about a topic, and who deepen their knowledge and expertise in this area by interacting on an ongoing basis"[1].

This definition of communities of practice captures the three structural elements of a community of practice, namely, its domain, practice, and community. It is worth elaborating on these three elements in the context of Wenger's other writings to understand this concept more fully.

The domain of a community of practice refers to the common "concern, set of problems, or passion about a topic" that all members share and around which they organize. It is the area of members' overlapping expertise, and it provides the initial centripetal pull uniting individuals who might otherwise work in separate organizational units, disciplines, or locations. Members have an abiding concern about their common topic, and collaborate to solve problems relating to it. They are also "passionate" about it. Wenger observed that, "If members don't feel personally connected to the group's area of expertise and interest...they won't fully commit themselves to the work of the community" [2].

According to Wenger, communities of practice form serendipitously when members recognize their common domain. Communities of practice are thus "fundamentally informal and self-organizing" entities, and can "benefit from cultivation" [2], which is the agricultural metaphor employed extensively by Wenger. The largely organic and non-prescriptive origins of these groups provide the conditions for members to participate creatively in practice [1].

Practice is the second structural element of the communities of practice concept, and represents the basic body of knowledge the group shares and builds. Members of a community of practice aim to "deepen their knowledge and expertise" on a topic by learning from each other. More specifically, members learn from peers and through practice, which is why members of a community of practice are practitioners. Here Wenger draws from earlier work with Jean Lave on guild systems and apprenticeship models, where learning and its application take place in the same setting [3]. "Through its practice--its concepts, symbols, and analytic methods--the community operates as a living curriculum," according to Wenger et al [1].

Communities of practice value both tacit and explicit forms of knowledge. They recognize the need to complement individuals' intuitions, perceptions, and vernacular knowledge with the more explicit concepts of an evidence-based paradigm. "The knowledge of experts is an accumulation of experience--a kind of 'residue' of their actions, thinking, and conversations that remains a dynamic part of their ongoing experience" [1]. In this respect, communities of practice are social learning structures: they are open venues of exploration, "where it is safe to ask hard questions and speak the truth" and where members "develop the habit of consulting each other for help" [1].

Wenger et al argued that when the three structural elements function well together, they "make a community of practice an ideal knowledge structure--a social structure that can assume responsibility for developing and sharing knowledge" (emphasis in original) [1]. Here, knowledge and sociability are not just linked, but are mutually informing, and the dynamic between them generates the specific know-how the group builds, shares, and manages.

Wenger's concept has gained its share of champions among practitioners and scholars in the business, development, and more recently, health sectors. A recognized strength is its insistence on collaboration as a practice-based learning process. Literature on communities of practice has almost unanimously supported the view that learning by doing is more effective than didactic instruction for adults working in today's knowledge economy [4‐9]. In the health sector, authors of several studies proposed that health workers and organizations interrelate learning, practice, and peer input to enhance individual and institutional performance [5, 10‐13]. Parboosingh, for example, argued that such interrelationships provided continuing education for physicians through ready access to a wealth of data, experiences, and mentors; he also suggested that through the dynamics of shared learning, health professionals become more reflective about their practice when they subject it to peer-based critical scrutiny [5]. Parboosingh proposed that physicians resemble Donald Schön's "reflective practitioner" when their knowledge is embedded in practice, informed by a body of evidence-based research, and shaped by the process of rethinking tacit assumptions [5]. CoP-type structures thus acknowledge the need for knowledge workers to be skilled at appraising and integrating various types of knowledge into practice and, ideally, to be forthcoming about it among peers [14, 15]. Reflection on practice is also suggestive of the need for workers to contend with the medical, economic, political, and socio-cultural understandings that inform the work they do [5, 10‐12, 16].

Because Wenger's concept does not privilege research evidence over experience-rooted knowledge, communities of practice can be venues for bridging traditional rifts in the health sector between research and practice, and among disciplines [12, 17‐19]. This is because a community of practice is an interpretive community. Wenger has drawn explicitly from Stanley Fish's theory of interpretive communities in the practice of literary and legal hermeneutics. Fish proposed that trained professionals determine the constraints on interpretation in their respective fields [20]. For Wenger, communities of practice are interpretive communities because they create their own understandings of the practice-world through the ongoing process of interpreting that world, thus producing the very knowledge that is held in common. The production of common knowledge through professional practice is echoed by Kerfoot: "When physicians, nurses, and clinicians are willing to come together in best practice teams, they are a learning community that has developed the relations that will make them effective" [18].

Notwithstanding the theoretical predictions, whether or not communities of practice are effective forms of collaboration remains an open question. Authors of a 2008 Canadian study found that extant literature on Wenger's concept gives little evidence of its usefulness beyond the level of description [21]; this observation is echoed throughout the communities of practice literature and lays the groundwork for future research in the field. Yet the Canadian study, like other studies, also recognized within the literature a burgeoning account of the inherent value of communities of practice for both the individual and the organization [11, 12, 19, 21, 22]. Among managers and employees alike, communities of practice are a compelling form of infrastructure because collaboration is organized around knowledge domains, not products or markets; this appeals to an increasingly educated workforce operating in settings where innovation is valued.

A further appeal of communities of practice is that they foster an egalitarian ethos by placing knowledge management in the hands of practitioners, not managers [23]. Collaborative learning has been shown to increase members' sense of ownership and autonomy when they take responsibility for setting work agendas, with ownership and autonomy identified as important factors for success in collaborative work [7, 13, 24‐26]. Enhanced collaborative learning has also been shown to improve job satisfaction [4, 5, 13, 25], which in turn can lead to potential increases in recruitment and retention of employees as high-value assets in learning organizations [18, 27]. Conversely, Pereles et al. observed that when small work groups are unable to take charge of learning, membership and personal growth wane [4]. Even groups that collaborate solely through email networks value the communal nature of knowledge development and sharing [28, 29]. Moreover, while evidence is only emerging on the success of heath sector communities of practice to implement the know-how they develop [30], in 2003 members of the Surgical Oncology Program (SOP) at Cancer Care Ontario endorsed Wenger's model as a means to facilitate quality improvements in cancer care [31]. Subsequent pilot tests of the model in Ontario regional cancer surgery settings revealed above-average ratings of physician satisfaction and levels of compliance with guidelines in clinical environments [13, 26].

Wenger's model has drawn its share of detractors. Critics have focused mainly on the internal dynamics of his concept of communities of practice to argue that the complexity of social interaction within them is inadequately recognized. The application of social science theory in particular to Wenger's concept has been a fruitful avenue of inquiry, because it has helped illuminate barriers to knowledge development in CoP-type structures, as well as identify key strategies to overcoming such barriers [13, 32]. Specifically, it has helped researchers to 1) gain a critical perspective on how power relations may operate in professional collaboration, and 2) better understand where Wenger's concept lacks clarity, particularly around the notions of community and the so-called natural formation of CoPs.

From Wenger's definition of a community of practice, it is clear that the main purpose of a CoP is to build and share knowledge related to practice. If this is the case, what, then, do we make of groups faced with the task of coordinating human activity or implementing change, in addition to sharing practice-based know-how? Strictly speaking, is it a misnomer to call them communities of practice, because they go beyond enhancing collective know-how to also include collective action? Consider two scenarios from studies involving collaborative groups in health care. The first is D'Amour et al.'s 2008 study on the effectiveness of collaboration among professionals in the Canadian province of Quebec. Collaboration involved linking hospital care with primary care services around the early discharge of newborns from hospital [43]. Members of these groups were committed to their practice of perinatal care, and shared information in order to improve their practice. D'Amour et al. developed indicators to measure the groups' intensity of collaboration and link it to clinical outcomes. They argued that it is important to understand how groups manage the responsibility of incorporating novelty into action, so as to learn how to leverage it. D'Amour et al.'s interest in the processes of decision making around new practices augments Wenger's model, and the study's groups are called collaboratives, not communities of practice.

The second scenario is CHAIN (Contact, Help, Advice and Information Network for Effective Health Care), which is a large informal online community of practice. In the UK, CHAIN's function is to share information and experience among professionals interested in evidence-based health care, and it has been highly effective in giving participants access to professionals from different organizations and disciplines, and in linking novices with experienced practitioners [28]. However, unlike members of D'Amour et al.s collaboratives, CHAIN's membership is not required to make group decisions, coordinate compliance with guidelines or objectives, nor to take collective action: for members of the UK CHAIN, collaboration does not entail joint deeds. Is CHAIN a better test case for Wenger's concept than D'Amour et al.'s groups? Does it matter if enhanced know-how--or improved practice--occurs at the individual, group, or system levels? The answers to these questions may determine the extent to which Wenger's model of collaboration is sufficient for moving thought into action that is associated with professional learning and knowledge translation. They may also determine how the effectiveness of Wenger's groups is to be identified and measured.

Under the auspices of the Canadian Partnership Against Cancer (CPAC), the CoP Project is currently addressing questions regarding the effectiveness of Wenger's model in facilitating practice-based learning. The initiation of the CoP Project was prompted by the need for an accountability tool for a strategy aimed primarily at facilitating the development of social collaborative structures to enhance knowledge translation in cancer control. These collaborative structures span traditional jurisdictions that ordinarily do not interact at a national level in Canada. The CoP Project's main objective is to design and test a tool for evaluating the performance of communities of practice in achieving their own stated objectives and advancing the agendas for which they implicitly exist. The secondary objective is to create a workbook similar to standard operating procedures, which communities of practice can use to guide data collection in order to assess their own performance over time. The products from this project will therefore be an evaluation tool and a guide for documenting activities that can contribute to evaluation. The evaluation tool will include modules for qualitative assessment based on semi-structured interviews of CoP participants; a survey tool of key indicators of CoP performance within several domains and that can be scored quantitatively; and a more formal quantitative social network analysis. To develop these tools, the project is prospectively observing and documenting two newly formed entities that we believe will evolve into more sustainable communities of practice: one of these is engaged in practice guideline development activities from a national perspective using an adaptation model[44], and the other is a newly formed national community of academically-based and community-based surgeons engaged in developing synoptic reporting templates for cancer surgery[45]. In addition, the project is collaborating with the Partnership's knowledge management team to design a tool for monitoring virtual communities of practice created through access to a common web-based collaborative site. The online collaborative site may be a potential vehicle for facilitating collaborative efforts among different types of communities of practice. Finally, the project is working with an already established community of practice of oncology nurses for developing guidelines in supportive cancer care.

The CoP Project should help test in the Canadian health care context Wenger concept and the claim that communities of practice provide a concrete organizational infrastructure for realizing the dream of a learning organization [1]. However, in the course of generating and scaling our list of indicators in preparation for field testing the project toolkit, it became apparent that adhering to the integrity of Wenger's concept did not go very far in diagnosing what makes collaborative arrangements work, or not work. Missing from the concept are statements about how group activities are negotiated. The decision was thus made to collect a wider range of data on matters relating to group conflict and how differences of opinion are resolved. In so doing, we hope to generate a more robust account of what constitutes successful collaboration, particularly in the context of groups of health professionals who must find adept and credible means to interpret the wealth of research evidence in their respective fields and generate consensus on recommendations for practice. Results from the CoP Project will be published in 2010.

Notwithstanding some limitations to realizing Wenger's concept, communities of practice models have been recently and successfully adopted in Canadian health care [6, 13, 26, 29]. One instance is in the province of Alberta, where a communities of practice model was used to optimize inter-professional communication and patient safety across three large health care regions in the province [6]. Each community of practice was allowed to choose a topic on which to focus, and to determine what was important for that topic. This autonomy created a sense of ownership and buy-in among participants, thus leading White et al. to conjecture that the changes implemented by the groups--such as streamlined admission and discharge processes for patients, and improved inter-professional communication--are likely to persist. A second example involves the current Surgical Oncology Program at Cancer Care Ontario, which adapted Wenger's concept to facilitate quality improvements in cancer surgery in Ontario. Adaptation included integrating the concept with other theories, such as health care improvement methodology and Nonaka's knowledge creation theory, as well as providing considerable management support to the groups and tying professional accreditation incentives to practitioners' involvement in improvement initiatives [13]. Preliminary results are encouraging, and show high levels of satisfaction with clinical leadership and knowledge translation among participating physicians. Results from this latter study suggest that Wenger's concept may be a supple starting point to effective collaboration in Canadian health care.