Leadership as boundary work in healthcare teams

Leadership as contextualized practice

Recent advances in leadership scholarship have called attention to the practices of leaders. Several authors have pointed to the limitations of traditional theories of leadership which focus on competencies and traits, and have argued for the need to focus as well on micro level practices of leadership (Alvesson and Sveningsson, 2003a, 2003b; Carroll et al., 2008; Denis et al., 2010; Larsson and Lundholm, 2010). Crevani et al. (2010: 84) state that ‘In terms of theories of leadership, there is a clear need for a deeper empirical understanding of everyday leadership practices and interactions (in contrast to the current preoccupation with individual leader competences and grandiose deeds)’.

A focus on leadership practices and roles calls for consideration of both interactions that occur in local sites and authority structures that influence how relationships are shaped. This requires attending to – as well as extending the view beyond – individuals in formal power positions. Bolden et al. (2008) point out that theories of leadership tend to center on leaders, treating followers as passive or subservient. They call for theorizing that attends to relational elements in leadership that would be viewed as a social influence process. Similarly Carroll et al. (2008: 374) call for a broadening of the concept of leadership and a redefinition of who is engaged in its practice, which would allow the study of how leadership is constituted ‘from different organizational positions (or ‘non-positions’)’. Consistent with Larsson and Lundholm (2010: 165), we view leadership as an aspect of regular work interactions in which influence is exerted. In line with a practice perspective, these authors conceptualize influence as ‘a joint accomplishment’ or ‘an interpersonal process, rather than one of transmission from leader to follower’. Thus, in this study, we attend to influence in the process of interaction, but we also remain mindful that the ability to influence varies among members. As our data will show, the practice of influencing how boundaries are drawn and managed is not limited to actors in formal leadership positions; however, neither is it evenly distributed.

The study of leadership as a practice also calls for attention to context (Bolden et al., 2008; Bryman et al., 1996; Gronn, 2002). Alvesson and Sveningsson (2003a) emphasize the importance of the micro-level context (or the specific local situation), arguing that understanding leadership requires attention to how ‘natives’ or participants in local sites view leadership in its different expressions. Thus, our main focus here is on leadership practices in such micro contexts. Nevertheless, we view these practices as embedded in a broader macro environment that also impacts how influence is viewed and authority is exercised locally. Bolden et al. (2008) argue that the context includes the social, cultural and political environment. We agree with researchers who have argued that restricting attention only to micro processes and practices (McGivern and Dopson, 2010) or only to macro processes and structures (Bechky, 2003) provide deficient accounts of how boundaries are drawn and managed.

In brief, we view leadership as involving the practices of influencing how boundaries are drawn and managed. These practices are dispersed and are not only the domain of individuals in formal authority positions. The practices are embedded in local contexts and influenced by the macro environment. We now examine in more detail how boundaries and boundary work have been considered in the literature, and relate this to the study of leadership practices.

Boundaries and boundary work

The notion of boundaries has received much attention in sociological (Burri, 2008; Gieryn, 1983; Lamont and Molnár, 2002) and organizational (Santos and Eisenhardt, 2005; Zietsma and Lawrence, 2010) studies; but so far there has been little research on the role of boundaries in leadership. The concept of boundaries refers to physical, cognitive, relational, structural, knowledge-based or other delimitations that separate one entity from another (Ashforth et al., 2000; Carlile, 2004; Kreiner et al., 2009; Lamont and Molnár, 2002). Studies in workplace contexts have focused on a variety of boundaries such as role boundaries (Ashforth et al., 2000; Kreiner et al., 2009), professional boundaries (Brown et al., 2000), hierarchical boundaries (Kahn and Kram, 1994), knowledge and task boundaries (Bechky, 2003; Carlile, 2004; Orlikowski, 2002) and boundaries that separate an organization from its environment (Santos and Eisenhardt, 2005). Boundaries are manifested at different levels in work contexts. The boundaries between roles occupied by a single person – such as the role of manager and the role of clinician – tend to be of an intra-individual nature. As Ashforth et al. (2000) state, transitioning between roles held by an individual is a boundary crossing activity involving a psychological movement for the individual. Other boundaries are of an interpersonal nature, such as the boundaries between members of the formal leadership group or ‘executive constellation’ (Denis et al., 2001; Hodgson et al., 1965). These members may engage in negotiation of the boundaries of their respective leadership roles (Stewart, 1991). While some authors have emphasized the need for ‘differentiation’ of roles in such circumstances suggesting clearly established boundaries (Hodgson et al., 1965), others have noted that effective co-leader pairs may in some conditions avoid boundary-setting, creating a ‘shared role space’ (Gronn and Hamilton, 2004). Still other boundaries are of an intergroup nature, such as professional group boundaries, which come into play when nurses, psychiatrists and social workers must interact in the performance of patient service tasks (Hall, 2005).

At the individual level, boundaries are interpretive entities that individuals perceive to delimit different domains, such as the different roles in which they engage. At the group or organizational levels, boundaries are seen as socially constructed by people involved in interactions. Lamont and Molnár (2002) note that boundaries are distinctions that social actors make to categorize objects, people or practices. It is important for studies on boundaries to attend to micro contextual elements, given that boundaries result from interpretations and social constructions in the sites in which they occur. Understanding how boundaries are manifested and managed in local sites is further enriched by attention to institutional factors within which boundary practices take place (Allen, 2000). For example, professional practices and boundaries in workplace contexts are highly influenced by professional templates at the institutional level (Chreim et al., 2007).

Some research attention has been given to boundary work – the tactics used to establish, obscure or dissolve boundaries (Gieryn, 1983). Establishing boundaries or engaging in boundary maintenance helps prevent infusion across domains and to cordon off one element from another (Ashforth et al, 2000; Gieryn, 1983); the demarcated domains will remain segmented and the attributes and identity of each domain will be preserved (Lamont and Molnár, 2002). Dissolving or opening boundaries allows for exchanges and diffusion across domains. Encouraging the sharing of knowledge across departments or communities is an example of dissolving or opening boundaries that might allow for integration, convergence or enrichment of points of view (Carlile, 2004; Orlikowski, 2002).

In the area of healthcare specifically, a variety of studies have addressed the notions of boundaries and boundary work. Studies have attended to how groups with different interests appropriate, interpret and attempt to shape ‘boundary objects,’ which may be abstract or concrete objects (Allen, 2009; McGivern and Dopson, 2010) that ‘inhabit several intersecting social worlds’ (Star and Griesemer, 1989: 393). For example, in Allen’s study, the boundary object was the design of a care pathway that allowed clinical, management and service user interests to be aligned. Boundary work related to demarcations between clinical and managerial roles has also been the subject of attention (Hotho, 2008). Still others have considered boundary spanning activities across different organizations, such as Fennell and Alexander (1987) who examined differences in boundary spanning strategies engaged in by freestanding hospitals and hospitals that belonged to multi-organizational systems. An extensive area of research relates to professional boundaries at the macro (e.g. Freidson, 1993; Beardwood, 1999) and micro levels (e.g. Brown et al., 2000). Notable among micro level studies is the emphasis on boundaries – and how they are managed – between physicians and other professionals (e.g. Allen, 1997; Svensson, 1996). Within this research stream, Brown et al. (2000) noted the need for adequate authority and structured management of multidisciplinary teams for the proper management of professional, role and team boundaries, and Reeves et al. (2010) argued that clear leadership roles are essential for interprofessional collaboration to occur. However, neither of these studies conceives of leadership as a set of boundary work practices.

In brief, the study of boundaries and boundary work has received much attention in a variety of literature streams, and the critical nature of boundaries in the healthcare field has been clearly established. Yet absent from the literature is attention to boundary work as leadership practice in healthcare teams. This leads to our research question: How are leadership practices exercised across and within boundaries in interprofessional teams?

Leadership practices as boundary work

In this section, we elaborate on each of six types of boundary practices identified. For each type, we describe the nature of the boundary and variations in leadership practices across the different teams. In the following discussion section, we draw together these findings by relating them to the literature and identifying three generic forms of practice and their potential consequences.

The findings indicated that the four teams we studied had a formal leadership constellation with distributed roles, i.e. a group of individuals – typically two people – who were seen as collectively playing hierarchical leadership roles. Leadership practices thus involved managing the boundaries between the roles that these individuals assumed as leaders. Further, because some of these leaders assumed frontline work seeing clients in addition to the formal managerial role, they also managed the boundaries between their role as team leader and their role as service provider coordinating their work with that of other members of the team. Boundary work was also evident in demarcations that were made between the formal leaders and other members of the team. While such demarcations may occur naturally in other environments, this boundary work activity warrants special attention in the context of interdisciplinary healthcare teams that are intended to function with flat hierarchies, where the negotiation of authority may be delicate. Additionally, leaders worked on helping team members manage the boundaries between the personal life experiences and the professional work involved in working with clients. Another boundary work activity for leaders involved managing the roles of different professionals who brought different skills, perspectives and areas of expertise to the client service; professional boundary work was part of leadership practices in these interdisciplinary teams. Finally, leaders engaged in boundary spanning – managing the boundaries between the team and the organizational and wider environmental context within which the teams operated. Figure 1 illustrates these boundaries, and Table 2 provides a listing of the boundaries with representative quotations from the various teams. We elaborate on each of the six boundaries next. OPEN IN VIEWER

OPEN IN VIEWER

Table 2.

Representative quotes.

Boundary work practices	Representative quotes
Managing boundaries within the leadership constellation	The bulk of the work is with the manager (who is) an Applied Nurse Practitioner and has her Master’s in Health Administration and she is outstanding. So the runnings of the team in terms of the budget and the overall management fall under her portfolio. Mine as Clinical Director is more looking at the overall scheduling, are we meeting the needs; and that’s done really in a collaborative way with our team as well. And that if there was a clinical problem, if we were short of physician coverage, or when we’re looking at how are we performing as a group and servicing the populations; that I see falls under my purview as clinical director. (Clinical director, Shared Care)
Managing boundaries between the managerial role and the clinical role	I think it makes a huge impact when managers are involved in actually providing clinical services, that you never lose sight of some of those challenges for other frontline staff, that you experience them yourself and then you bring them forth often as issues thinking, did you guys experience this? These are the dilemmas that I have when I’m out there as a nurse doing a home visit … Sometimes I feel in conflict, like decisions as a team member versus a manager. And when do I step in as a manager versus a team member? I found that an interesting transition. I still struggle with it sometimes. (Program manager, ACT)
Managing boundaries between formal leadership and other team members	In order for our team to work, everybody has to be respected equally and you can’t create tiers. There’s naturally going to be tiers. I’ve been on the team the longest. I’ve got a big grey beard. I’m a big guy. So I carry a certain weight … I think there is this but it’s more undertone. We really don’t encourage that. (Clinical director, AMHP) Technically the formal leaders are the clinical leader and the program manager. But ultimately there are some strong people on this team that provide tremendous leadership and the clinical leader and program manager will sit back and let that go. (Team member, ACT) I tend to take on a lot of special projects … I’ve done things such as creating the chart audit tool that we use, I’ve created the procedures and guidelines manual. We’re really lucky here in that our program manager is very open-minded at this point to having the team make decisions around functioning and programming. (Team member, ACT)
Managing professional boundaries	One of our agenda items was – there is no hierarchy … A nurse gets to make the final decision, or the counsellor makes the final decision – it’s you make a decision as a team and you make a decision based on your scope of practice … We have a lot of newer nurses coming in and so they get that right from their learning experience … So they aren’t necessarily coming in thinking, I’m a nurse so I’m the most important team member. (Program manager, AMHP) Each team member has their domain of expertise but over time so much of what we do is grey. I don’t mind debates on even medications and medication choice although I (psychiatrist) am the authority figure in that area. (Clinical director, AMHP) We do not call for another professional based on our impotency by saying I’m not capable, take this case. No. It is based on the specificity. The client needs one specific treatment, could you see him because you are the expert for that treatment. (Program manager, External Care) Everybody is better in his specificity, and everybody knows more about what his colleague does because we have torn down the silo culture. (Clinical director, External Care)
Managing boundaries between personal experiences and professional roles	That’s what I mean about the leadership helping the team use the group dynamics to sort out what isn’t applicable to the client and what is, and how to process that data to make the best decisions in terms of clinical orientation. All the time, it’s like a radar … I think it’s my responsibility to have leadership in keeping an observing gaze on what goes on and trying to put that back into a clinical framework, and if the discussion reflects the ambivalence in the conflicts of the client and how can we best use all those ways of looking at the situation in a way that helps the client the most … It’s like water in a turbine, if you don’t harness it and help (team members) process it in a way that’s useful it can be quite destructive. And if there isn’t leadership then you can have internal dynamics that kind of, people personalize views, like the politics of the team. That’s one of the criticisms of ACT, so much energy gets used up in the politics of the team as opposed to just helping the client. If there’s poor leadership that could definitely happen. (Clinical director, ACT)
Managing boundaries between team and environment	It is not always easy, but people realize that we are better if we try to work together. I am not only talking about the interprofessional level. Everything is linked. The culture in the entire environment supports the organization. I am also talking about the inter-institutional level, so with the Youth center, the Southern part of the region as well as the Northern part, the Readaptation Center and the schools. We realize that together, we are not so bad. (Clinical director, External Care)

Boundary work within the leadership constellation

In all the teams, two or more individuals were designated as formal leaders, typically with one person being designated program manager and another designated clinical director. Program managers were members of different professions, while the clinical directors were typically psychiatrists. Given that two or more individuals were involved in formal leadership at the level of the team, it was interesting to note if and how they set boundaries around their roles. This is thus the first boundary we identify as implicated in leadership practice (see Figure 1). What we found is that the roles of these individuals were more or less delineated, indicating varying degrees of clarity in terms of boundary demarcations. Clearly defined roles are illustrated below in the case of Shared Care.

The situation is similar in other teams, such as the External Care team, where administrative leadership is mostly assumed by the program manager and clinical leadership by the medical chief (a psychiatrist). As the program manager describes it:

We have joint management; we call it clinical-administrative. The clinical part is theoretically the psychiatrist and the administrative part is me. We function almost like twins.

In the quote above the roles of the program manager and the clinical director seem clearly delineated. However, the expression of ‘twins’ is telling for there is some replication between these two leadership positions. During meetings, we observed that both the program manager and the clinical director made efforts to provide a vision for the program. Both share information with the professionals about the strategic goals and the situation of the clinic in the regional health system. So while there are some clear boundaries between the daily work of the two leaders, their practices intersect. The administrative and clinical sides influence each other.

In the ACT team there was some delineation of the role of the two leaders. The program manager held responsibilities for operational decisions and strategic goals of the team and also ensured that they meshed with those of the hosting organization – a community health center. In addition, the program manager dealt with issues such as conflict between two members requiring formal resolution. The clinical director’s role focused on the clinical direction of the team and encouraging team members to enhance their professional development. The view was that in unison, both these leaders paid attention to the ‘milieu of the team’. We noted the importance of practice based on a ‘partnership’ between the two individuals. The program manager stated:

I can't do this job without the clinical director … You must have a working partnership with the psychiatrist on the team … It's pivotal that the psychiatrist and the manager are the two bodies that help guide leadership.

The ACT team clinical director described her role:

All the ACT teams are different but the role that I’ve identified for myself is taking clinical leadership in terms of psychiatric treatment and responsibility for clients, and taking a consultative role to the other professionals on the team, which means using whatever influence I have to help them maximize their professional development and contribution. The third aspect is responsibility for team functioning. So in partnership with the team manager, that aspect of my role is to observe and help influence team functioning so that the team is most effective.

Such terms as ‘joint management’ and ‘partnership’ indicate the importance of good working relationships for the practice of distributed leadership. We see that on this and the other teams we studied, the program manager and clinical director distribute the leadership role, clearly demarcating the boundaries in terms of some responsibilities while opening the boundaries in terms of others. There was no unique way of doing this, but all the formal leaders navigated between elements of differentiation and commonality in enacting their leadership practices.

Managing boundaries between the managerial role and the clinical role

The second type of boundary involved in leadership practice concerns that between managerial and clinical roles (Figure 1). Participants mentioned that there are major benefits to having a program manager who is a clinician – namely benefits derived from the ability to understand the issues that front-line practitioners face as well as having credibility because of this role. Here, we use the example of the ACT team:

There are many key ingredients to have a good ACT team; one key ingredient is having the manager do clinical work. Very important because they maintain credibility, they keep their feet on the ground, they know the reality of what the team is doing, they share … It’s huge for the manager and for the team.

However, there were also difficulties associated with managing the boundaries between the managerial practices and the clinician practices on the team as the program manager indicated:

I talked about it to the staff here and I said, ‘I struggle with my role some days about when I need to be a manager and manage something and when I'm the team member’. They said, ‘It's a struggle for us too, because you have a different authority than we do, and sometimes when we're listening, it's like, is she saying that we have to do this? Or is she just introducing a different idea - so not as a manager saying that that has to be done.’ I'm still trying to work it out as I go along and letting them know that they have to tell me when they think I've either gone over that boundary or they're not sure what I'm asking.

Thus, the managerial and the clinical work were experienced by the program manager as a duality. It is possible that in less egalitarian teams, the clinician-manager would step into the managerial role and make decisions without experiencing the boundaries between the two components of the work. In the quote provided above, however, we see that the boundaries between the two practice components are not well defined by the manager, and that she seeks input from the team members to help clarify the demarcations. Team members can play a role in helping define the boundaries between managerial and clinical roles for those in leadership positions as is also illustrated in the case of the External care team. Although the clinical and managerial roles are not exercised by the same person, the program manager lives a duality similar to the one expressed by the program manager of the ACT team. As a former mental health counsellor, the program manager expressed the desire to cross the managerial boundaries and intervene in the clinical work of the professionals.

Here I don’t have clinical tasks, even if that is what I am most interested in. I participate sometimes in the discussion about clinical cases, I like that. I talk to the professionals about their practice. I know the clinical practice sufficiently to follow what they are doing. But now, I am not a mental health counsellor anymore.

The clinical team members reminded him of his leadership role and the existence of the aforementioned boundaries. For them the program manager was exclusively a reference for managerial-administrative issues.

Peter, [program manager] is there for all administrative matters. For clinical matters, it’s not him. If I have questions regarding my clinical work, I go see my colleagues … Peter is there to ensure the functioning of the team … Those are two different things … , and it is easy to differentiate them.

For those in program management positions, the boundaries between the administrative and clinical practices may not be clearly delineated. They may seek or simply receive confirmation of the boundaries of their roles from other members of the team, who through their practices confirm the limits of administrative leadership, and help prevent encroachment of administrative leadership onto the clinical domain.

In summary, the managers of the interprofessional teams not only had to negotiate boundaries mutually with other co-leaders, but they also found themselves engaged in working out role boundaries between leading clinical work and actually practising it, a phenomenon that seems particularly likely to occur in the context of knowledge-based organizations in general, and in healthcare in particular. The paradox is that engaging in clinical roles may help maintain expertise and boost a manager’s legitimacy with other professionals, but it can also lead to considerable ambiguity as focal managers grope to understand how to interact appropriately with other professionals in clinical situations.

Managing boundaries between formal leadership and other team members

A third boundary that was negotiated, drawn and occasionally contested in interactions is the boundary between leaders occupying a formal managerial position and other members of the team (see Figure 1). We analyzed the extent to which boundaries delineating formal hierarchical positions were maintained or erased. One indication of hierarchical boundary erasure is the extent to which members of the team are invited to participate in important decisions. The input from team members into what might be classified as managerial decisions was evident in AMHP. Softening of the formal manager/ team member boundary was achieved when the manager admitted that she did not always know what was best for the team, and actively sought input from team members.

The erasure of the hierarchical boundary can also be prompted by the members of the team, who themselves take the lead to accelerate change in the work place. Continuing with the case of AMHP, we report on an exchange.

Interviewer: Are there people on the team who help bring out the quieter voices?

Respondent: There are … we’ve got a nurse that’s a 35 year veteran and if you present something to her she will help you hash it through and decide what’s the rationale? Is it worth going forward with? And if you don’t want to present it, she’ll say - ‘I’ll do it’. And so you can work with her and do it. And that’s just one. There’s several.

Consistently with our social constructionist stance, we were interested in what members in local sites considered to be leadership. We found this to involve practices that were engaged in by members of the team who were not in formal managerial positions. In the ACT team, for example, a common theme was ‘we take turns with leadership’ and several team members indicated that they lead a variety of client programs, such as the Voice Studio (where clients gather to ‘find their voice’ and sing) and the Healthy Lifestyles Program (where clients engage in outings and physical exercise activities). The following extracts from interviews with members of the ACT team who were not in formal leadership positions demonstrate this issue.

Interviewer: So when you think about leadership, who do you think of? What is their role and what do they do as leaders?

Respondent 1: I would actually say that leadership, it sort of rotates depending on the situation, but it comes more from everybody.

Respondent 2: To give you an example, when we do business and function meetings we rotate Chair (across all members of the team). The Chair is responsible for running the meeting, keeping people in check, on time, navigating the conversation … So we’ve made a conscious team decision that we want everybody to have the opportunity to build those skills.

Respondent 3: You never feel micromanaged … You know that there’s a couple of leaders but … they give a lot of room …

There was confirmation of the view of leadership held by team members in the words of the program manager who indicated that ‘I'm not sure that I'm always the leader of everything. Sometimes team members come up as leads on many things.’ The members of this team spoke about a time when the program manager made decisions and managed more autonomously and when the team was experiencing difficulties emanating from unresolved conflicts among members. It was at this time that four members of the team attended an ACT conference during which a speaker addressed issues related to team functioning. These four members brought ideas on steps to use to improve the team functioning to the manager and the rest of the team saying that ‘We need to figure out how to communicate better’. This ushered the start of a process that members qualified as ‘difficult’ when team procedures (such as chairing meetings) and functioning (setting commitments for conflict resolution) were renegotiated among the members.

The above examples illustrate that leadership is conceived as being inclusive of practices by members who are not in formal management positions, that the boundaries of leadership can be contested by team members and that interactions and negotiations help shape how leadership is constructed and practised in local sites.

Managing professional boundaries

The fourth boundary examined (see Figure 1) is perhaps the most obvious one in inter-professional teams. One might expect professional boundaries and boundary encroachment to be problematic in interdisciplinary teams such as the teams we studied, and that leaders might need to invest in managing the tensions between professionals. We did find this to be the case in the External care team where the practice of placement (placing a child in a Center) evokes some tensions between the professionals. Different professionals play a role in the implementation of this treatment, especially the psychiatrists and the social workers. The clinical director practiced boundary management by reminding the professionals about their specific roles. For instance, during one meeting, he supported the claims of one social worker who, as a representative of her professional group colleagues, informed the team that the practice of placement should result from a social worker’s assessment rather than a medical prescription. The clinical director recognized the distinction between the role of the physician and the social workers.

The children (who need to be placed in a Center) need adaptation strategies. The physicians are not the professionals who are responsible for deciding that. We must not interfere in the process.

In another vein, we found that team leaders may also take actions that reduce professional boundaries by engaging in practices that help enhance the scope of practice of certain professions. In Shared Care, for example, the team manager illustrated an attempt at erasure of some professional boundaries:

Similarly, managers in AMHP were instrumental in softening the boundary that scope of practice created between professions. In an interview, the manager indicated that in AMHP prior to the collaborative model, nursing had been focused on medical aspects and counsellors on ‘psycho-social stuff’. The vision of this manager at that time was that specific scope of practice activities had to be performed by the designated professionals, but that other activities could be shared such as the psycho-social assessment. In an interview a year later the manager described the functioning of professionals in AMHP:

I have nurses there that are really part of the team and they don’t necessarily wear their nurse role as being who they are. Like it could be, I’m the nurse on the team but I also play games and do counselling … The fact that I’m the only one that can give the injection and the medication is really about my scope of practice.

The degree to which the traditional clinical boundary between physician and non-physician team members had been erased in AMHP was illustrated by the willingness of the psychiatrist to have his diagnoses and medication prescriptions debated.

We do the history … then I usually go back and we discuss people’s thoughts. What’s the diagnosis? What do we think we need to do? And everybody kind of throws in their two bits and we’ll even have debates on whether we’re going to push for medicines or whether we’re going to go conservative. Although that’s my domain. Each team member has their domain of expertise but over time, so much of what we do is grey. I don’t mind debates on even medications and medication choice although I am the authority figure in that area. So push comes to shove, I’m the one that has to prescribe it, so we’ll debate all those things.

The traditionally strong professional boundaries would likely have prohibited much of this open discussion prior to the formation of the interprofessional collaboration. Indeed, the clinical leader indicated that new team members were included in these diagnostic and medication discussions immediately upon joining the team, so as to show them in practice how the boundaries between members of the team were being softened.

Another indication of professional boundary demarcation is addressing the clinical director (or psychiatrist) as ‘doctor’. We noticed that this boundary was demarcated in some of the sites, and erased in others. In Shared Care, for example, members referred to each other on a first name basis, and on the basis of observation of team meetings, it was difficult to tell what position members occupied. In Shared Care, the clinical director (psychiatrist) stated the following:

We’re very respectful of one another’s viewpoints … I think in some ways we started with maybe team members being a bit more deferential to the psychiatrists … but the psychiatrists on the team have sort of resisted that.

In general, we found the professional boundaries in all these teams to be carefully attended to and managed by the team leaders, sometimes with contributions from leaders of particular professional groups. Team leaders tended to close the boundaries when they felt that scope of practice – which is institutionally determined – and unique expertise had to be maintained and protected, and sought opportunities to open the boundaries and reduce professional demarcations when interprofessional participation might enhance client care.

Helping members manage the boundaries between personal experiences and professional roles

The work involved in mental health can be heavy psychological work, where it is possible for healthcare workers’ personal psychological issues, as well as their personal feelings towards each other or towards the client, to influence clinical decisions. For example, in the ACT team, members need to work very closely with each other: there is a very high level of integration of tasks performed by different members, which creates the possibility for interpersonal tensions to dominate clinical decisions. A member of ACT indicated that ‘You’re trying to separate your professional and your personal feelings and that’s hard, so the clinical director (the psychiatrist) is a really good place to do that.’ In this team, leaders spoke about the importance of helping members understand the boundaries between the personal and the professional in their dealings with each other and with clients. This is well illustrated in the quote in Table 2. We identified this issue as a fifth critical boundary subject to leadership practice (see Figure 1).

The management of the boundaries between personal experience and the professional role manifested itself in another way in the case of External Care. Here the clinical director emphasized the demarcation between clients and the professionals. This leader emphasized the importance of taking a step back during the process of diagnosis and treatment, to question oneself on the real impact of the therapeutic relationship. In a meeting, he explained that it is important for a mental health professional to recognize the limitations of their intervention, indicating that ‘The challenge is to deal with the heroic commitment, the will and the dedication, without inducing iatrogenic treatment.’ He highlighted the demarcation between professionals and clients, cautioning against over-dedication of professionals that might create dysfunctional dependencies. As he expressed in an interview:

Sometimes, by following our clients too long or too closely, we nourish their problems. Our goal should not be to maintain a clinical relationship. We hope that they will learn to deal with their problems and grow.

In a meeting of Shared Care, we observed that a team member brought up the long waiting list, and her worry about clients who needed treatment but had to wait to receive it. The clinical director intervened, reminding the team that three months was an acceptable time to wait, and that if the client had an urgent need in the meantime, there were alternative means of receiving help.

The above examples are all instances of leadership practice aimed at helping team members demarcate the boundaries between personal emotions and professional work. The situations described here are remarkably similar across cases, suggesting that this boundary is a common preoccupation in mental health teams, likely also to manifest itself in other situations where emotional involvement with clients risks invading personal lives.

Managing boundaries between team and environment

We found that an important component of leadership practice was managing the boundaries between the team and the wider environment (the sixth type of boundary identified in Figure 1). The environment consists of the organization within which the team operates, as well as the macro environment that goes beyond the organization. While boundary work at the periphery of the team generally involves spanning and making linkages between the team and the environment, it might also entail closing the boundaries in an attempt to insulate the team from influences that are counterproductive. For example, in the ACT team, the program manager indicated that budget allocations to different programs in the community health center led to lengthy discussions from which she shielded the team.

My job as the leader is not to let that seep down into the rest of my service and to keep that separate … not to let those things impact on the team too much, because then your focus gets lost on what you're supposed to be doing with clients. I'm trying to just protect the team from any of that political stuff going on.

While in some instances, leadership involves buffering staff from what is happening elsewhere and closing the team’s boundaries to information coming from the environment, in other instances, it involves opening these boundaries. Opening the boundaries of the team to the external environment entails choosing a frame by which the team’s mission and objectives can be defined. Whether in Shared Care, ACT, External Care or AMHP, there are choices to be made between alternative frames that the team can adopt. For example, there are different models of ACT that vary in the extent to which the team focuses on client stabilization versus rehabilitation. In the case of the ACT team we studied, the leaders gravitated towards the rehabilitation focus, which was strongly promoted by an influential peer specialist that had served on the team since its inception. Similarly, there are different models of Shared Care that can be adopted (for example, primary or acute service). Leaders play an important role in influencing what frames are chosen. The following relates to Shared Care:

Where we go and how our model incorporates different aspects of things is where the leadership comes from, the clinical director. She may have more influence than others over that issue, has been part of the Shared Care milieu for longer than some of the rest of us, and knows more of the players nationally, and attends many of the conferences, around where we’re heading, how our model is going to evolve and change.

The leaders of External Care also exerted an influence on the choice of the frame of the clinic. Boundary spanning was manifested in the comparison of the ideological and practical models of the clinic with those that exist in the mental health system. The leaders attempted to distinguish their clinic from the traditional mental health model.

Our system is organized in a different way from elsewhere: here, the physicians gravitate around their teams rather than the opposite. Moreover, the physician is not the only person responsible for the case; the other [psychosocial] professionals inform the physician and co-sign the case.

Boundary spanning with the environment in External Care was also practiced by highlighting the distinctions between the mandate of the team and the responsibilities of other organizations that the team had to deal with (for instance the Youth Center, the Recovery Center, etc.). During meetings, the leaders explained to the professionals how they can open or close their boundaries as a team vis-à-vis other organizations. Opening boundaries could result in collaborative work with other organizations.

Leadership practices involving boundary spanning with the external environment thus revolved around choosing a frame that helped define the team in its wider context. It also revolved around opening the team boundaries for collaboration with other organizations or closing the team boundaries to influences considered to be negative.

Opening and closing boundaries within leadership constellations

First, as we showed, leadership in healthcare often involves constellations where several individuals share or distribute leadership roles. The data from our sites indicate that the boundaries between the roles of the members in the leadership constellation were opened or closed depending on the issue requiring leadership attention. While budget and operational decisions made by the program managers seemed to be separated from clinical direction decisions made by clinical directors, as in the case of Shared Care, the boundaries were erased when it came to decisions about the milieu of the team. In the case of the ACT team, both leaders engaged in practices related to improving team functioning. References to joint management and partnership between the leaders are indicative of both separation and sharing – or of demarcation and erasure of the boundaries between the roles.

Gronn (2002) indicates that one of the properties of distributed leadership is interdependence, which can be manifested in responsibilities that are overlapping or complementary. We see overlapping and complementary practices as defined by open and closed boundaries respectively. The literature on leadership role constellations (Denis et al., 2001; Hodgson et al., 1965) suggests that a lack of differentiation (or open boundaries) between the roles of constellation members is problematic because it can lead to conflict, duplication and rivalry. We did not observe such phenomena in our data even when there were instances of overlap, because the constellation of members in all cases appeared to have developed a fluid mutually supportive back and forth style that seemed to constitute a form of heedful interrelating (Weick and Roberts, 1995). Gronn and Hamilton (2004) argued that this rather fluid kind of relationship defined as an undifferentiated ‘shared role space’ may be more common where members of the constellation had similar kinds of expertise as for co-principalships for schools. Here however, in all cases, the leaders had different professional backgrounds, suggesting that even in such cases, greater proximity and fluidity in roles across boundaries is conceivable.

To summarize, within the constellations we observed, opening boundaries around areas such as improving the functioning of the team promoted knowledge sharing, collaboration and mutual support. Closing boundaries around specialized tasks such as budget management protected co-leaders from being overwhelmed, avoided unnecessary duplication and provided clarity for others. We observed in our data little mutual boundary contestation. However, this does not mean that such a phenomenon might not occur especially when one or more constellation members change and boundaries need to be renegotiated and re-communicated to others.

Closing boundaries at the manager/ clinician interface

Leaders in healthcare often occupy administrative and clinical roles, and segmenting these roles may not be an easy task, as our findings show. While some studies indicate that clinician-managers may resolve the tension emanating from dual roles by subsuming one role under the other (Hotho, 2008), our data indicate that the issue is not so easily resolved. The weaker the boundary between these roles, the more there is potential for confusion regarding which role to enact (Ashforth et al., 2000). The potential for confusion also exists for members of the role set who may be unclear as to which role the individual is enacting. In our sites, the leaders sought or obtained help from team members who contributed to the drawing of the boundaries. In the case of External Care for example, members helped demarcate (close) the boundaries by ensuring that the program manager did not surpass the administrative role or encroach into the clinical domain.

The boundary issues created here indicate that boundaries between roles can blur, creating ambiguity for leaders and other members of the team regarding where the lines should be drawn. This points to the importance of recognizing these ambiguities and engaging in tactics that can clarify demarcations and delineate different practice domains.

Opening, contesting and renegotiating boundaries between managers and others

It became apparent that some boundaries were influenced by individuals who were not in formal managerial positions but who engaged in practices leading to the reshaping of boundaries. In AMHP and in External Care, team members participated in decisions about the team’s direction at the invitation of management. In ACT, members took leadership on a number of defining projects (e.g. the Voice Studio). Opening boundaries at this level contributed to bringing more viewpoints to bear on decisions and to empowering team members to take initiatives in their areas of expertise.

Members sometimes also actively contested the boundaries of formal management. In the ACT team, members contested prevalent practices and took the lead in bringing about changes to the way that the team functioned. They saw that leadership was much more widespread than the formal management, and succeeded in renegotiating roles towards greater openness.

These findings are consistent with Bolden et al.’s (2008) observation of extensive influence by individuals lacking formal leadership titles. Examples from our sites indicate that ‘leadership interactions and practices … include possibly diverging processes … and debates … ’; they also remind us that ‘we must be open to local leadership constructions that involve all participants, and that may result in multiple local constructions and ways of relating’ (Crevani et al., 2010: 81). Our findings clearly suggest that leadership and boundary practices are negotiated among participants and socially constructed in local sites. In healthcare organizations, the contesting and renegotiation of boundaries between management and others may be particularly common as issues of professional autonomy, managerial control, and the mobilization of expertise intersect.

Opening, contesting/negotiating and closing boundaries between professions

Of the boundaries we identified in healthcare teams, it is likely that none are as ubiquitous as those separating one professional domain from another. Extensive research at the macro level (e.g. Scott and Backman, 1990; Freidson, 2001) and the micro level (e.g. Ferlie et al., 2005; Hall, 2005) have addressed the erection of boundaries between professions in healthcare. Other studies, however, have documented opening of professional boundaries in interdisciplinary teams (e.g. Belling et al., 2011; Nugus et al., 2010; Suter et al., 2009). We found instances of a variety of practices in our teams.

The leaders opened boundaries when they believed that members of different occupational groups could play a role in providing assessments and suggestions about client treatment options. For example, in AMHP and in Shared Care, the psychiatrists sought input from other members of the team. In AMHP different professionals provided suggestions regarding medication, and in Shared Care, having nurses titrate medication was being considered in an organizational and institutional context which set down medication management as a reserved act for physicians. This indicates that in some contexts, the focus on improving patient care through interdisciplinary collaboration trumps the rigidity of interprofessional boundaries. Opening the boundaries is desirable when knowledge from different domains needs to be integrated and is believed to produce richer decisions (Carlile, 2004; Orlikowski, 2002).

Yet this should not detract from the fact that physicians had authority that was not shared equally with other members of the team. Note that in AMHP, final decisions about medication remained the domain of the psychiatrist (‘So push comes to shove, I’m the one that has to prescribe it’), and in Shared Care, the nurses were allowed to titrate medication, but not to prescribe. As Nugus et al. (2010: 908) indicate, ‘there is still - at this snapshot in the history of interprofessional relations - a ubiquity in the actualized and potential domination by doctors across various care settings, with cultural and institutional currency’. These are longstanding boundaries perpetuated at a macro level by legal and political structures, and as Allen (2000: 329) states, ‘while they may not determine work boundaries in a straightforward way, they certainly help fashion their contours’.

Interprofessional boundaries may also become a locus of contestation and renegotiation in interprofessional teams. An interesting instance is in External Care, where some of the psychiatrists were seen as encroaching on what the social workers perceived to be their domain. The leaders of the social work group contested this boundary encroachment. In response, the team’s clinical leader asked the medical professionals to hold the boundaries around social workers’ roles. In this case, social workers’ practices were protected from encroachment by the physicians. It is interesting to see that it was the clinical leader (a psychiatrist) who was instrumental in closing and reinforcing these boundaries against members of his own professional group thus implicitly separating himself from them in enacting an authoritative leadership role. The institutionally-based hierarchy among professions seems to mean that clinical leaders are usually those whose leadership and boundary work practices contribute most strongly to the empowerment of other professional groups.

Here we see that leading an interprofessional team may not be a one-sided push to open boundaries but a delicate operation combining opening in areas where collaboration and knowledge sharing will improve services, while holding boundaries in other areas to ensure that each group’s skills and expertise are best utilized. We also see how boundary work may involve continued situated renegotiation as leaders and professionals experiment with, stretch and contest domains of action and expertise.

Closing boundaries at the personal/professional interface

As we showed, the leadership practices we observed to be associated with the boundaries between professionals’ personal and professional lives all illustrate attempts to better establish these boundaries rather than to eliminate them. There is a common perception that mental health professionals naturally tended to overinvest rather than underinvest in client relationships, requiring leaders to act to counterbalance this tendency. The more strongly interventionist forms of care (such as ACT) pose particular risks in this regard, and leaders were particularly attentive to this concern. Clearly here, boundary work was intended to protect the professional, and in some cases the client from the risks of over-involvement, and to re-segregate work and personal roles (Ashforth et al., 2000; Kreiner et al., 2009). In the professional work literature, attention has been paid to the issues associated with professional/ personal boundaries and the risks of dual roles (Kagle and Giebelhausen, 1994; Lambert and Davidson, 1999). However, there has been little attention to leadership practices in relation to these dilemmas.

Opening and closing boundaries at the environmental interface

As we noted, the opening and closing of boundaries with the external environment was also a key preoccupation for team leaders, who were quite selective in the way they positioned their teams with respect to others. It is perhaps no accident that more fluid and open interprofessional relationships within the team were sometimes associated with a need to buffer and protect the team from a broader macro context characterized by institutionalized professional distinctions.

Leaders of these teams tended to pay particular attention to ideological boundaries differentiating them from other groups of providers. In Shared Care, leadership work involved drawing the boundaries of team work around primary care to the exclusion of acute care, and this drawing of the boundaries was done with a strong understanding of models prevalent nationally. In the case of ACT, formal and informal leaders chose a rehabilitation frame – from possible alternatives prevalent in the environment – as a defining element of team practices. Leadership work involving setting the ideological boundaries of the team entailed engaging with institutional themes, and choosing a frame to define the team in its wider environment.

There is a broad literature on boundary crossing among different organizational contexts in healthcare, much of it examining ways to increase fluidity and continuity of care (e.g. Denis et al., 1999; Chreim et al., 2007; Currie et al., 2011; Rodriguez et al., 2007). Yet, there has been limited focus on how leaders within healthcare units negotiate boundary relations to combine the potential benefits of openness associated with knowledge flows and collaboration with the integrity of the work group. When intragroup boundaries are fluid and open, there may be a greater need to carefully manage intergroup boundaries to protect and preserve the gains achieved.

To summarize the discussion, we note that the three patterns of leaders’ boundary work are differentially observed in our data, with some boundaries more subject to practices associated with opening boundaries (professional boundaries; within constellation boundaries), some more associated with closure (professional/personal boundaries), and some more frequently the site of contestation and renegotiation (hierarchical and professional boundaries). Different combinations of patterns might obviously arise in other settings. However, the key points are that leadership practices are fundamentally tied to boundary work, and that this work requires tradeoffs between opening up and closing down lines of demarcation in ways that balance the need for enhanced knowledge flows and sharing with the needs for clarity and security.