Developing programme theories of leadership for integrated health and social care teams and systems: a realist synthesis

Research literature

In consultation with information services specialists at both King’s College London (London, UK) and Kingston University London (London, UK), the following search strategy was developed:

“Integrat*” OR “multi-team*” OR “multiteam*” OR “cross-bound*” OR “cross bound*” OR “cross-organisation*” OR “cross organisation*” OR “cross-sector*” OR “cross sector*” AND “leader*” (Limiter: English language only, where available).

This search strategy was run in the following databases: EMBASE, Health Management Information Consortium (HMIC), Social Policy and Practice, Cumulative Index to Nursing and Allied Health Literature (CINAHL), MEDLINE, International Bibliography of Social Sciences, PsycINFO and Education Research Complete. A total of 1446 empirical research papers were identified, of which 532 were duplicates and were removed, leaving a total of 914 papers for review. These papers were divided between two reviewers (SS and SF), who read only the abstract of each paper to determine whether or not it was relevant to the focus of the review. At this stage, 848 papers were deemed not relevant and, therefore, were excluded from the review, leaving a total of 66 papers. These papers were divided between the two reviewers and read in full, as a result of which 43 papers were excluded,25,42–83 leaving a total of 23 papers. 31,84–105

Grey literature

Grey literature relating to policy and organisational-based material was sought by searching Google (Google Inc., Mountain View, CA, USA), Google Scholar, government and other specialist websites [e.g. NHS Leadership Academy, Skills for Care, The King’s Fund, Advance HE (formerly known as the Higher Education Academy), The Institute of Healthcare Management, Social Care Online, NHS England and NHS Improvement]. Key words adapted from the main search strategy were used and included ‘leader’, ‘leadership’, ‘integrated care’ and ‘integrated system’. Enormous numbers of evidence sources were identified in the searches. Most were not relevant. Forty-one pieces of grey literature were identified and read in full by one reviewer (RH). This excluded 27 pieces of grey literature,8,9,106–130 leaving a total of 14. 10,131–143

Stage 1a

A total of 37 papers (23 empirical research, 14 grey literature) were included in the first phase of the stage 1 search. These papers were divided between the three reviewers (SF, SS and RH), who each independently compiled a list of mechanisms or preliminary programme theories based on the papers they had read [to ensure consistency of approach, a small number of papers (n = 4) were read by all the reviewers so that the team could compare their analyses]. The reviewers then met together regularly to discuss their coding of mechanisms and any queries they had until agreement was reached. The following preliminary mechanisms were identified and accepted by all the study team (a short description of these mechanisms can be found in Table 1):

1. supportive relationships and trust

2. team working/collaborative working

3. shared mission/vision/approach/purpose

4. shared responsibility/ownership

5. learning, development and innovation

6. communication

7. providing clarity

8. balancing needs

9. advocacy

10. external liaison/consensus building.

Given the complexity of this review, an important component of the methodology was the inclusion of contributions from key stakeholders throughout the process. This process is recommended in realist methods,144 as learning what stakeholders know about an intervention and its reason for implementation is essential to understanding it. In this study, consultation and discussion with the stakeholder group was a vital part of the review process. Throughout the report we refer to their contribution to developing knowledge about leadership of integrated care teams and systems. Patient/service user and carer representatives were members of this group and therefore patient and public involvement (PPI) was embedded in all stages of the review. To ground the review in the lived experiences of stakeholders, separate consultation meetings were held during the project period with three stakeholder groups: those leading or working within integrated care teams and systems, patients/service users and carers receiving care from integrated services (PPI) and researchers with expertise in integrated care or realist methods. Most members of the stakeholder group were identified by searching on the internet for leaders of integrated care teams and systems. We were keen to find people who were currently involved in leading care delivery and overall system leaders. It was a challenge, as little detail is provided on NHS and social care websites. Some members were known to the research team in their professional capacity or were suggested as relevant by people we invited. Thirty-three people were invited by e-mail; four did not reply to the invitation and follow-up e-mail, seven declined (no longer in the UK, n = 1; retiring, n = 1; too busy, n = 4; unavailable on planned meeting days, n = 1) and three accepted the invitation but were unable to join any meetings and did not contribute to the review of study documents/findings by e-mail. Nineteen geographically dispersed stakeholders agreed to participate in the group and contributed to at least one meeting, consisting of six individuals with expertise in integrated care, eight with direct experience in leading integrated care teams or systems, two researchers with methodological expertise and three patient/service user and carer (PPI) representatives. Several members of the stakeholder group also had experience of both working in integrated care and leading teams or systems. The first stakeholder meeting took place at stage 1a in the synthesis, when the group met in person to discuss the preliminary mechanisms highlighted in Table 1. An independent chairperson led whole-group and small-group discussions, during which stakeholders were asked to comment on the mechanisms, including which they felt were most pertinent, any that did not appear relevant and any important mechanisms that may have been missed.

Overall, stakeholders felt that, although some of the mechanisms or programme theories were valid and relevant to integrated care teams and systems (e.g. ‘providing clarity’), others (such as ‘communication’ and ‘supportive relationships and trust’) were too general, and had already been identified in the generic leadership literature. They felt that the review needed greater interrogation to identify the components of leadership that were specific to integrated care teams and systems. For example, one potential mechanism that stakeholders felt was missing from the synthesis was around the use of power dynamics in teams and the way that leaders negotiate these. They therefore suggested that we include ‘use of power’ as a potential new mechanism and specifically search for any discussion of this in the literature (for more information on how stakeholders contributed to the development of mechanisms see Chapter 3, Stakeholder perspectives). Stakeholders were asked to forward any relevant papers that we may have missed on to us for review. Stakeholders also agreed that we consider exploring literature outside of health and social care, to see whether or not any other fields had identified potential theories around leading integrated teams and systems that may be applicable to health and social care. A subsequent search stage (stage 1b) was therefore undertaken.

Stage 1b

Given the difficulties that we had experienced in identifying mechanisms of leading integrated teams and systems in the health and social care literature reviewed, the scope of the review was expanded to include material outside health and social care (e.g. business and disaster management). We returned to the papers previously identified but excluded because they were based outside health and social care and also identified any possible new papers from their reference lists. Twelve possible new papers were identified at this stage. 47,48,70,145–153 These papers were divided between two reviewers and read in full. Five of these papers were excluded and the remaining seven were included. 47,48,70,145,147,148,150 Five additional papers were also forwarded on to us by members of the stakeholder group who felt that they would be useful inclusions in the review. These were again divided between two reviewers (SS and SF) and read in full. All five papers were included in the review at this stage. 154–158

The additional 12 papers identified in stage 1b were divided between three researchers (SF, SS and RH) for review. Each reviewer independently compiled a list of mechanisms based on the papers they had read and also returned to their papers included in the stage 1a search and re-analysed these to look for any newly identified mechanisms. Thus, a total of 49 papers were included at this stage. The reviewers then met together again to compare their mechanisms and discuss their findings. In these discussions, they explored how their new mechanisms compared with those identified in stage 1a and explored ways in which the different mechanisms could be separated or merged to identify the components of leadership that were specific to integrated care teams and systems. The following 10 preliminary mechanisms were then agreed:

1. inspiring intent to work together (this merged components of the ‘supportive relationships and trust’ mechanism with the ‘communication’, ‘advocacy’ and ‘team working/collaborative working’ mechanisms as well as incorporating newly identified aspects identified in stage 1b)

2. enabling people to work together (this merged the ‘shared mission/vision/approach/purpose’ with newly identified aspects identified in stage 1b)

3. strategic networking/focusing on the bigger picture (this merged components of the ‘supportive relationships and trust’ mechanism with the ‘external liaison/consensus building’ mechanism as well as incorporating newly identified aspects identified in stage 1b)

4. commitment to learning and development (this mechanism remained similar to that identified in stage 1a but also incorporated feedback from the first stakeholder consultation group and searches conducted in stage 1b)

5. clarifying complex processes (this merged aspects of the ‘providing clarity’ mechanism with newly identified aspects identified in stage 1b)

6. creating balance between organisations and individuals/managing conflict (this merged components of the ‘balancing needs’ mechanism as well as incorporating newly identified aspects identified in stage 1b)

7. use of power (this merged components of the ‘shared responsibility/ownership’ mechanism as well as incorporating feedback from the first stakeholder group and searches conducted in stage 1b).

8. leader resilience (this was a newly identified mechanism, based on searches conducted in stage 1b)

9. flexibility of leadership styles (this was a newly identified mechanism, based on searches conducted in stage 1b)

10. use of public narratives (this was a newly identified mechanism, based on searches conducted in stage 1b).

These mechanisms were discussed and agreed by the study team. We then met with our international advisor to the study and expert on multiteam systems (MTS), Professor Stephen Zaccaro. The team discussed the process undertaken and the difficulties experienced by the reviewers in identifying specific mechanisms of leading integrated care teams and systems. Professor Zaccaro advised the team that ICSs could be conceived of as MTSs and suggested that the reviewers review the MTS literature for mechanisms of leadership of complex teams. The team, therefore, ran a search in Google Scholar for ‘leadership of multiteam systems’. Twenty potentially relevant papers were identified and read in full, as a result of which 14 papers were included159–172 and six were excluded. 173–178 These 14 papers were divided between two reviewers, who, again, searched for any context–mechanism–outcome (CMO) configurations and for any mechanisms, contexts or outcomes that were not specifically linked in an explanatory way and added these into the preliminary mechanisms mentioned previously. Thus, a total of 63 papers were included in the stage 1 search (37 papers from stage 1a combined with 26 papers identified in stage 1b). All the included papers (and the mechanisms identified from them) are listed in Table 2.

A list of preliminary mechanisms was sent out by e-mail to the stakeholder group for review and then presented in full at the second in-person stakeholder group meeting. The independent chairperson again led small-group and whole-group discussions around how the mechanisms should be amended or merged. Stakeholders were specifically asked to challenge the mechanisms and were given space for questions to help generate deeper exploration of the key issues. By the end of this meeting stakeholders were content that the following mechanisms (generated as a consequence of literature searching and whole-group discussions) were pertinent to leading integrated care teams and systems (a short description of these mechanisms can be found in Table 3):

1. inspiring intent to work together

2. taking a wider view

3. creating the conditions to work together

4. clarifying complexity

5. planning and co-ordinating

6. balancing multiple perspectives

7. working with power

8. commitment to learning and development

9. fostering resilience

10. adaptability of leadership style.

Although some mechanisms (e.g. ‘inspiring intent to work together’, ‘creating the conditions to work together’ and ‘clarifying complexity’) were felt to be more relevant than others (e.g. ‘planning and co-ordinating’, ‘fostering resilience’ and ‘adaptability of leadership style’), the stakeholders and research team agreed that all 10 mechanisms should be explored in greater detail in stage 2 of the study.

The stage 1a and 1b search processes are shown in the flow chart in Figure 2.

FIGURE 2.

Flow chart of stage 1a and 1b searches.

Overarching ‘inspiring’ leadership qualities

Certain components of the ‘inspiring intent to work together’ mechanism were identified as important at both the systems and team level. This included leaders having a clear vision for collaboration and being able to articulate this vision to others with passion. 22,84,207,208,588 This vision acted as the criterion against which leaders judged the suitability of a proposed course of action, ensuring that they focused on what was central and enduring to integrated working. 208 It was important that leaders lead change according to this vision rather than being ‘swept along’ by external events. 22 Other overriding components of ‘inspiring’ leaders included being visible,213,588 being strongly committed to integration and implementing lasting change,84,197,588 being able to gain the trust and respect of others,198,210,217,588 being a good communicator22,84,208,209,219,588 and being able to develop strong interpersonal relationships with colleagues. 22,198,217 It was considered important for leaders at both levels to have the skills to build and sustain a culture of interdependency, reciprocity and collaboration and to instil key values such as co-operation, openness and fairness among their colleagues. 84,105 As a sense of ‘being in it together’ developed, there was evidence in one study of a reduction in ‘gaming’ between organisations, which made negotiations and collaboration more straightforward. 22

Despite the numerous competing demands on their time and energy, having leaders at all levels who demonstrated good listening skills was identified as important to develop a deep understanding of other organisations and individuals. 208 Being listened to by leaders made people feel as though their input was valued and helped increase motivation and engagement. 208 Similarly, leaders who openly recognised the time, effort and skills that others contributed to integrated working made staff feel respected, appreciated and motivated to contribute more. 208 Conveying genuine respect for the views of all staff, regardless of affiliation or power, reinforced principles of inclusion and elevated members’ respect for leadership. 208 At both systems and team levels, the importance of leader credibility and legitimacy was also identified. 22,105,208,217,588 Credibility was gained through having knowledge of both health and social care through direct experience of working in both fields – for example, a nurse leader who had previously worked in local authority social services;105 through being associated with previously successful developments;588 or through one’s personality, skills and the dynamics of one’s relationships with others. 22 Credibility was lost when leaders appeared to listen to the contribution of others but never incorporated their input substantively. 208

System-level leadership

Important components of the ‘inspiring intent to work together’ mechanism at a system level included having good communication skills, particularly around communicating the vision of an integrated care partnership to the partnering organisations22,207,208 and in maintaining an open, honest and consistent message. 22,220 Successful systems leaders were immersed in developing a clear vision of integrated working and spent considerable time and energy encouraging other executives to share this vision. 22 A study of community health partnerships in the USA208 found that partners valued systems leaders who were forthright and direct in their communications and willing to address issues ‘head on’ rather than trying to minimise or deny them. The authors stated that, because system-level leaders could not rely on formal structures and authorities that facilitated action in organisations, they relied heavily on their own interpersonal skills and effective communication skills to assure a wide and multidirectional diffusion of information. There was also evidence that system-level leaders needed to demonstrate adaptability in their communication style, for example, demonstrating the ability to communicate with individuals across several organisations in multiple directions – both up and down but also horizontally and even diagonally. 208 Traditional leadership styles may have required communicating with a limited number of leaders, but leaders of integrated care systems needed to respect and allow for the diverse needs of a variety of organisations and communities, each of which had different expectations of the timing, extent and channel of communication. Systems leaders were, therefore, required to tailor their style of communication and the language or jargon employed to bridge cultural gaps between organisations and communities. 22,208

A UK-based action research study exploring the aspirations and achievements of a newly formed Mental Health NHS and Social Care Trust220 highlighted the importance of systems leaders in inspiring staff during the early phases of integration. In this study, the formation of a new Care Trust (integrating the local Mental Health NHS Trust and the mental health and learning disability services provided by the local authority social services department) was seen as an innovative move by many of the staff involved, although there were complexities around staff expectations for the ‘culture’ of the new trust. Staff wished to retain several characteristics of their ‘old’ health and social care cultures (e.g. localness, relationships) while simultaneously recognising the need for a new, integrated, culture. This aspiration was acknowledged by the Chief Executive during the consultation process when she stated that the culture of the Care Trust would ‘keep the best of both’ previous organisations in the development of the new one. This key role of the systems leader was, therefore, seen to influence the subsequent culture of the Care Trust and helped to reassure staff that the new integrated systems would contain positive aspects from both the health and social care settings (Box 1). Furthermore, the authors reported that this positive early experience of integration at a system level formed a foundation for the importance of partnership working between health and social services, which became one of the core values of team-level leaders within the Care Trust. 220 Other authors have highlighted the importance of systems leaders creating opportunities for staff to talk about their successes and ensuring that local ‘wins’ are celebrated and communicated to the wider group. 217

A facilitating context for this mechanism was that team members could trust their system-level leaders. Team-level leaders said that their ability to ‘sell’ their message about integration to their teams depended on whether or not they trusted that the arrangements in place at a system level would improve care or allow for the detection of deteriorating quality. 588 However, in an investigation of four international case studies of integration,588 building trust was identified as requiring time, and in some cases, work to strengthen integration was founded on a decade of prior work, through which trusting relationships had slowly grown. Many physicians in these sites said that they were only willing to participate in integrated working because they had trust in their systems leaders and their belief in the mission (Box 2).

Instances were also provided where the ‘inspiring intent to work together’ mechanism was absent in system-level leadership. For example, an Australian case study199 provided two contrasting contexts of health governance for integrated care and found leadership to be significantly lacking at both board and executive levels. The study found that, although interorganisational relationships between boards and chief executive officers were crucial to effective system-level working, they were often absent and instead replaced by an adversarial culture typified by a ‘master–servant relationship’ between organisations. A Scottish study198 described the implementation of a project to develop and test models of integrated working within a health board. Team leaders were asked what helped and hindered them in the delivery of change in this integration project and identified the lack of ‘buy-in’ from senior-level leaders as a key hindrance. Not having a clear steer or support from senior leaders meant that these team leaders felt that they lacked permission to ‘act’ and this resulted in role confusion among the team. A Swedish study comparing two cases of clinical integration efforts following a hospital merger further highlighted the absence of the ‘inspiring intent’ mechanism at a system level. At one case study site, team members described system-level leaders undertaking the process of integration without consulting staff at a team level or including them in decisions (Box 3). Team members felt that these decisions had been made hastily and in secret and so refused to be involved with the merger decision or adapt their ways of working. 216

Team-level leadership

At a team level, leaders were found to have a ‘determinative influence’ in integrating health-care services214 and play a key role in inspiring their team members. 209 The ‘inspiring intent to work together’ mechanism presented itself as a leadership approach characterised by informality, inclusiveness and a sensitivity to the roles and contributions of all team members. 205,206 Other papers reported team leaders managing ‘by influence’221 and of the need for leaders to be enthusiastic; dynamic; ‘on the ball’; willing to take action; and able to motivate, empower and encourage others. 84,207,209,588 In a study of multiagency working involving professionals from education, social services and health settings in the UK,207 a team member highlighted the importance of their team leader going beyond the realms of their own job for the greater good of the team:

. . . I am sure that’s beyond his/her day job, you need that one person who is prepared to keep it going and not let it slip, keep badgering people.

Atkinson et al. 207

Communication skills were also felt to be important for team-level leaders. 22,84,206,209,219,588 A Norwegian study206 highlighted the benefits of team leaders’ having an ‘open style’ and an ‘accessible profile’. This approach was found to foster a positive, collaborative culture and a fertile learning environment. 205,206,209 Furthermore, having team leaders who facilitated open discussions and challenged assumptions that limited thought and action was also associated with a greater degree of team synergy. 219 A Norwegian study of leadership in interprofessional collaboration in residential childcare services206 found that team leaders encouraged interaction processes by influencing and motivating their staff as well as their collaborative partners, young people and their families. For example, team leaders were reported to intentionally facilitate interaction processes to make them cohesive for the benefit of young people, by encouraging all team members to continuously focus on ‘the child’s best interest’, and on running the institution in ‘the best possible way’. Another study210 further highlighted this need for team leaders to encourage ‘professional altruism’ among their staff (i.e. to ensure that the team’s focus was always on the benefit of the patients/clients in their service).

One study found that the personality of the leader, rather than their professional background, was a key factor in their effectiveness and ability to develop positive relationships with their team. 221 Others found that team leaders needed to have accepted their leadership role but also be willing to ask for help when needed. 209 Presenting a unified front with other team leaders or deputy leaders was also seen to be important in inspiring others to participate in integrated working216 (Box 4). Aside from the team leader role, some authors also found it important for there to be additionally identified ‘champions’ in integrated teams, who reminded members of the team’s benefits and provided sustained motivation. 98,197,216 Although anyone could be a ‘champion’, it was felt that senior clinical leaders were particularly successful, because of their ability to engage with and motivate their professional peer group. 22,197,199 Another study also noted the importance of identifying a succession plan for such team champions. 98

Instances were provided where the ‘inspiring intent to work together’ mechanism was absent in team-level leadership. For example, one study highlighted a case where the leader did not identify themself as a strong leader and questioned their own capabilities to fulfil the role. 203 They found that they perpetuated individual rather than collaborative ways of working, which decreased trust and increased dissatisfaction in the team. 203 Furthermore, instead of promoting openness and honesty, this leader’s lack of confidence in their own abilities led them to be complicit in increasing internal conflicts by not exposing or challenging people’s criticisms, complaints or behaviour. This resulted in team members becoming frustrated and the leader becoming even more self-critical and further doubting their legitimacy in the role (Box 5). Others found the ‘inspiring intent to work together’ mechanism was hampered by recalcitrant team members, who slowed progress by working in silos due to a lack of trust. 199

Overarching leadership qualities associated with ‘creating the conditions to work together’

There were components of the ‘creating the conditions’ mechanism that were identified as important at both a systems and team level. For example, both systems- and team-level leaders were required to bring a shared vision or purpose to their team92,105,197,208,215 and to set and drive an agreed strategy. 199 It was also important for leaders at all levels to create mutually beneficial shared goals with accountability for outcomes and to have the ability to bring staff together. 199,211,588 A key enabling context for this mechanism at both a systems and team level was ensuring that leaders had sufficient time to develop working relationships and to build an atmosphere of trust, commitment, professional respect and knowledge. 199

System-level leadership

At a system level, having leaders who were personally committed to and made an effort in engaging others in the overall mission/vision of the system was key. 84,92,197,215,218,220 This included enabling a structure that contained aspects of both health and social care perspectives or encouraging shared beliefs about the benefits of change. 92,197,215,220 For example, an Australian study describing the key enablers of change in reorienting to integrated primary care delivery identified the importance of shared values at an organisational level. In this study, a shared organisational vision and values were developed by the community through grassroots participation. 215 Other authors highlighted the importance of systems leaders creating a framework for action208 or ‘rules of engagement,’218 which they communicated compellingly to mobilise resources and guide action towards long-term aims. 208,218 The redrafting of job descriptions for staff was also identified as necessary in some instances to help facilitate collaborative working. 217

Additional components of the mechanism at a system level involved the need for leaders to put in place effective organisational systems and processes associated with governance, strategic planning, finances, human resources management and IT systems to accommodate service integration. 84 Systems leaders were also required to develop and implement plans and evaluate and update these regularly and systematically to ensure that they met current needs and priorities. 84 Furthermore, while staff at a team level sometimes lacked the authority to make changes to services (described as finding themselves hitting a ‘glass ceiling’), successful systems leaders raised these concerns for them at the board level and helped to instigate the changes that teams wished to make. 213 One study217 highlighted the value of introducing a system-level project co-ordinator specifically focused on improving integration or in developing project co-ordination groups with the aim of developing a clear vision and purpose for the system. The authors concluded that frontline clinicians and practitioners often faced so many time constraints in their daily work that it was difficult to bring about any changes without the assistance of a dedicated project co-ordinator.

The absence of the mechanism was also reported at a system level, with reports of a lack of shared vision, both organisationally and politically, leading to fragmentation. 199 Similarly, progress was reported to be noticeably slower across organisations where a shared sense of vision was not widespread. 197 For example, a Swedish case study of clinical integration efforts following a hospital merger described the consequences when the vision of a systems leader did not match that of a team leader216 (Box 6). In this study, an experienced and established team-level leader believed that the new system-level leader focused too much on a ‘tough business management culture’ and was overly concerned with reducing staff numbers. The team leader felt unable to work in such a culture and so resigned from the position soon after their appointment, much to the disappointment of their team.

Team-level leadership

The creation of one team culture to unify all staff was also highlighted as an important factor in leading an integrated service at a team level. 92 This included helping to develop and negotiate shared purpose, for example, in the form of service models that override narrow organisational, professional or individual interests. 92,105,210 These shared models offered an overarching framework that overruled different models of professional working and could be used to focus effort in times of disagreement or conflict. 105 For example, in one study, authors described the importance of having ‘a common mission so that you’re not always defending your own side’. 92 Ways to achieve such a common approach included providing opportunities for joint meetings for all team leaders and their superiors,84,92,588 establishing specific working groups to enable staff with similar interests across the team to work collaboratively together216 and ensuring that appropriate information and reporting systems were in place. 588 Another way was for leaders to ensure that they involved all their team members in the process of finding effective solutions to problems, regardless of where the problem originated. 92 This was felt to achieve a feeling of solidarity among team members, which was, in turn, a major contributing factor in the provision of sustainable integration of health and social care. 92 It was also important that team leaders were able to successfully navigate and adapt organisational systems and processes to fulfil their role. 84

At a team level, the mechanism also involved team leaders setting priorities and making decisions, balancing the workload throughout the team and constantly monitoring it. 209 In a study describing an evaluation of integrated team management in an English health and social care trust, team leaders described their main responsibility as co-ordinating team members and ensuring that operational management systems functioned effectively so that staff could work together in a ‘joined up’ approach. 221 Furthermore, the ‘creating the conditions’ mechanism also involved team leaders encouraging shared or distributed leadership in their teams or helping the team to perceive leadership as a collective activity. 22,92,205,217 This was based around the belief that integrated care required different skills at different times and these skills did not all need to be embodied by one person. 22 As such, it was important that leaders were open to delegating tasks and responsibilities appropriately. 206,209 Another study217 noted that leadership roles undertaken by team members were often not clearly articulated at the onset but developed in an emergent manner as integrated working advanced in teams where trust was evident. As previously mentioned, some studies also noted the importance of teams having a general ‘co-ordinator’ role, whose duty it was to co-ordinate work between different organisational departments. 209,217 These co-ordinators were reported to play a pivotal role in managing the integration process. 217 For more on shared leadership in teams, see Mechanism 4: working with power (n = 16). An important context for enabling this mechanism was that team leaders received support from their organisational management structure (including Chief Executive Officers), as this provided organisational credibility for any innovative work that they carried out around changing processes and practices. 204

Examples were also highlighted where the mechanism was absent at the system level and this had an impact on team-level leadership. For example, one study212 found that standardised co-ordination processes set in place by integrated systems did not always align with the needs of specific teams, team members or patients. This could result in frustration and conflict among the team and ultimately have an impact on patient care. When this was the case, having integrated team leaders with good personal co-ordination skills (e.g. interpersonal co-ordination abilities such as scheduling meetings with appropriate parties, making personal requests for consultations and demonstrating negotiation skills) helped to resolve conflicts and modify standardised co-ordination procedures.

System-level leadership

Historical power imbalances between health and social care (the sector as a whole and not just local authority services) meant that the latter was often perceived as ‘the poor relation’ and integrating these services did not necessarily mean that social care would be given a higher political priority. 587 Systems leaders, therefore, needed to be mindful of this imbalance to try and prevent this historic pattern from repeating itself. 587 There was some evidence that systems leaders were instrumental to progress with integration through their ability to work across different groups and institutions, acting as ‘diplomats’ where tensions arose. 588 An English study of multiagency working across social services, health and education settings found that a key skill of successful leadership at a system level was the ability to accept and be confident working alongside the tensions and ambiguities raised through cross-sector working. 207

There was also evidence of the absence of the ‘balancing multiple perspectives’ mechanism at a system level. For example, one author198 described a study where a project to develop and test models of integrated working within a Scottish health board was implemented. In this system, there was evidence of cultural differences between the respective organisations and team leaders talked of bureaucracy at an executive level slowing down team processes in both organisations:

. . . it has to go up. . .. to the Clinical Exec[utive], and it has to go to . . . this Board and that Board.

Lunts198

There were also practical problems around the lack of priority given to joint working at executive level. There were reports of team leaders agreeing to undertake tasks at meetings but never following through on their promises. This was because their respective leaders were not checking that they had fulfilled their roles, and ‘. . . if they weren’t going to be made to do it. . . then they’re not going to do it,’. 198 This was explicitly contrasted with more successful single-agency projects. 198 A further example was also provided of the absence of balancing multiple perspectives at a system level in an examination of an integrated care system in a French health-care network. 583 This study described the development of the informal network as being complex, owing to the differing views of the network members [general practitioners (GPs), gerontologists, neurologists, a speech therapist, neuropsychologists] around how to improve the diagnosis of older patients with cognitive disorders (Box 7). Each member had their own interests and vision for the network, but the network leader rated their own profession (neurology) more highly than the others (owing to an unspoken medical hierarchy in the institution). They focused on only their own vision for the network and what they wanted to achieve (i.e. improved diagnosis of cognitive disorders and faster referrals across the network). To achieve this, they relied only on the expertise of the GPs in the network and set aside the knowledge possessed by other members:

He did not want to involve other . . . institutions, out of fear that discussion might lead to designing a network that was different from his own representation of it.

Grenier583

This decision to not equally balance the multiple perspectives in the network had several outcomes, both positive and negative. First, this focused approach enabled the leader to achieve their goal by designing a rapid detection test for cognitive disorders, which was successfully used by the GPs and enabled quicker referrals across the network. However, there were also negative consequences of this decision, including some network members feeling ‘sidelined’ in the decision-making process and unmotivated to participate further, as they felt that the leader’s view of the purpose of the network was limited and did not match their own. At a later stage, the network required funding from the regional authority, which was promoting a more multidisciplinary vision, and so this provided an incentive to the leader to include other perspectives in the network. They therefore incorporated the perspectives of other team members in a controlled way, for example, by incorporating the views of a speech therapist and neuropsychologist into their vision. They continued, however, to exclude the gerontologist (whose views they knew directly contradicted their own) from the decision-making process and so the gerontologist felt marginalised and withdrew from the network. In time, the network members pushed for a new, formal network co-ordinator role that would broaden the focus of the network beyond the limited parameters of the leader [for more information on this, see Mechanism 4: working with power (n = 16)].

Team-level leadership

Key skills of team leaders were identified as the ability to effectively communicate with a wide range of people, including the ability to focus on mutual gains, to encourage team members to appreciate and value the core skills and expertise of others, to bridge diverse cultures and to have the ability to manage difficult conversations and remove any obstacles to change. 105,195,199,200,205,212,217,219 Integrated working could be difficult due to differences in professional language, attitude and values210 but when this mechanism was present, the leader was able to consider the circumstances and ways of thinking of different disciplines and balance them accordingly. 209,213 Team members trusted that they could call on their leaders to help resolve conflict among staff and leaders were able to do so effectively. 84,196,209,210,212

A key context of this mechanism for team leaders was that they were given the necessary time to learn about the different competencies in their team so that they could see these differences as an advantage rather than as a problem. 210 For example, a Swedish study of collaboration between welfare agencies in vocational rehabilitation found that team leaders deliberately involved themselves in the steering committees or working groups of different collaborative projects to reduce territorial behaviour among staff. 210 As a consequence, however, they reported having to wear ‘two different hats’ – both as a leader of their team and as a steering group member for a collaborative project – and said that it could sometimes be difficult to balance loyalties to both groups (Box 8). Another context for this mechanism was whether or not shared leadership was in place in teams. A Swedish study, mentioned previously, comparing case studies of clinical integration efforts following a hospital merger216 found that having shared leadership structures in place (involving the leaders of both pre-existing clinical departments) was an effective way for leaders to cope with the often conflicting logics and goals of multiple stakeholders.

When the mechanism was absent, team leaders were unable to manage conflict between team members appropriately. For example, an English action research project providing an autoethnographic account of a leader of an integrated health and social care working group203 highlighted the case of a team leader who was unable to expose and eliminate the underlying tensions, conflicts and lack of trust within their team (Box 9). By worrying about offending people and trying to keep the peace, the leader did not confront the members causing tension in the team and, therefore, inadvertently perpetuated the culture that was so problematic. This made leading the team ‘exhausting’ and meant that managing interpersonal issues took up much of the leader’s time and energy. The group did not achieve collaborative advantage but collaborative inertia, holding lengthy meetings and discussions with negligible outputs and this made the team leader question their own skills and leadership abilities.

System-level leadership

There was evidence of the fundamental importance of system-level leaders having a requisite level of power (including authority, influence and responsibility) to support the process of integration. 213,218 For example, one study highlighted how the chief executive officer, chief operating officer and chief medical officer were the clear central authorities responsible for all financial and clinical decisions in the integrated system. 218 Having such authority at a system level offered others at team level the permission to work and lead across organisational silos. 213 Systems leaders also required the capacity for change, including leaders who were willing to adjust their role217 and had the ability to adjust their leadership style, driving an agenda forward when required but letting it go when necessary. 207 A US study of community health partnerships found that power sharing was the most important mechanism through which leaders could develop ICSs. 208 This study found that many different individuals could have a significant influence over integrated systems, including both formal and informal leaders and idea generators, and it was important that these influencers were diverse and representative of the community being served. 208 By sharing power to set priorities, allocate resources and evaluate performance, systems leaders fostered a sense of joint ownership and collective responsibility and increased the system’s effectiveness. 208 However, an important context for this was the presence of a strong existing foundation of trust and honesty between system partners and that partners viewed power sharing as an authentic move from leadership. Systems partners with less power could quickly feel disenfranchised or threatened if they did not feel that their views were respected or if they felt that leaders showed preference for the views of one partner over another. 208

Other examples were also highlighted of systems leaders relinquishing their power for the benefit of the system as a whole. For example, a Canadian study of the implementation of a new integrated oncological services network in Quebec102 found that the change process towards integrated working was initially marked by resistance from medical staff across the region. Some of this resistance was attributed to the fact that the network leader was unknown in the region. To counter this, the network allied itself with the regional board and the regional board’s director (who had legitimate power in medical settings and good relationships with hospital staff) to set up a temporary committee responsible for implementing the project. This committee then worked towards winning support from hospitals within the region. This study found that systems leaders using power carefully and appropriately helped the network progress successfully, encouraged trust and respect within local teams at all hospitals and interorganisationally and helped team members develop a shared philosophy and vision (Box 10).

In a previously mentioned study in the French health-care system,583 dissatisfaction from team members around the narrow focus of the network leader meant that the leader eventually agreed to the development of a new network co-ordinator role. This role was undertaken by the neuropsychologist, who used their own professional attributes to broaden the focus of the network. The co-ordinator was championed by the leader as they did not call into question their professional knowledge or institutional position. It was agreed that the leader would continue their focus on the diagnosis and rapid referral of people with suspected cognitive disorders and that the co-ordinator would focus on other issues, such as improving co-ordination and multidisciplinary working. The network members and the value of their input were therefore viewed differently by the two leaders, based on the knowledge and the institutions that they represented and the focus of their roles. The leaders also differed in their opinions of diversity – the network leader viewed diversity as a factor that might disrupt their work and the perception of their hierarchical rank based on their knowledge and they therefore were dismissive of diversity. They chose to rely solely on the opinion of GPs who had medical knowledge. In comparison, the co-ordinator perceived diversity to be enriching for the network and important in making it more compliant with the demands of public funding bodies. They, therefore, broadened the knowledge criterion to include new areas of knowledge, such as social work and psychiatry and other types of organisations, such as care homes and local authorities. By relinquishing some components of their power and legitimising the joint leadership role of the co-ordinator, the leader helped to ensure that all components of the network were developed appropriately, with both the diagnosis of cognitive disorders and the development of multidisciplinary working/co-ordination gaining sufficient attention. In isolation, the narrow focus of the network leader had some negative implications but the joint partnership with the co-ordinator had a more positive and stabilising effect on the team.

There were also examples of the absence of the mechanism at a system level. For instance, a report on the reform of health and social care in Northern Ireland found that interviewees felt that the Department of Health and Social Care held a greater proportion of power than systems leaders. 587 It reported that systems leaders, therefore, looked to the Department of Health and Social Care for solutions to problems rather than coming up with them themselves and this could be detrimental to experimentation and innovation:

In this culture and context, paralysis and lack of decision making becomes the norm. There is no confidence within the system to take ownership of innovation and creativity.

Dayan and Heenan587

Interviewees described an atmosphere of ‘them and us’ between the upper tiers of the hierarchy and those ‘at the coal face’, with senior leaders adopting a forcible manner towards those lower down the hierarchy. There were reports of senior leaders in the department ‘confusing fear with respect’ and making it difficult for systems leaders to drive change. As one senior leader reported, it was best for leaders to: ‘keep their heads down and get on with it’. 587 This culture of mistrust was then passed down to a team level, with GPs voicing suspicions of systems leaders as ‘empire builders’ who were unwilling to work co-operatively. 587 These problems had been recognised by leaders in the system and steps were being taken to address them, for example by taking a co-production approach with the department, leaders, teams and patients to improve partnership working and health and well-being outcomes, although some interviewees reported doubts that this would be successful.

Furthermore, an England-based action research study exploring the process of integrating local mental health and learning disability services into one Mental Health NHS and Social Care Trust highlighted the problems experienced when systems leaders lacked authority or power. 220 In this study, although the Director of Social Care from the local authority was a member of the executive team, it was perceived by staff that they were not sufficiently organisationally embedded and lacked authority or influence in the trust. This lack of influence of social care was believed to have negative implications for the new organisational culture and the director left after a relatively short period of time in the organisation (Box 11). Another study found that systems leaders who made decisions without consulting staff at a team level made staff feel devalued and disrespected, which could lead them to leave their employment. 216

Team-level leadership

Research suggested that authoritative, directive and controlling leadership styles were generally inappropriate, whereas shared and distributive leadership approaches based on shared power relationships were likely to be more productive105,206,213,217 and lead to greater team synergy. 219 A Canadian study of change agency in a primary health-care setting217 found that the ability to influence, initiate and implement change was dispersed, with no single individual or group having full authority, resources or expertise to lead change. They concluded that shared or distributed leadership was highly prevalent in health-care settings with integrated or interorganisational contexts. A UK study exploring occupational therapists’ views of integrating service provision across health and social care213 noted the importance of team leaders giving their staff autonomy and not feeling the need to constantly check up on their work. They described the role of integrated team leaders as needing to:

Set the strategic direction, draw up your action plan and leave it to the people who are doing the job.

Best213

Another study216 also highlighted the importance of giving staff autonomy and found that team members appreciated this management style:

. . . he delegates tasks and trusts that the work will get done. So he rarely issues orders.

Choi et al. 216

A case study of two integrated care services in South Wales105 found that successful team leaders valued inclusiveness and had an awareness of the need ‘to share leadership with others who have more relevant “knowledge” ’. 105 This was also identified in the study of a hospital merger in Sweden,216 mentioned previously, which found that successful team leaders delegated decision-making authority, involved staff in the decision-making process and included their suggestions. This made staff feel involved and valued and thus become motivated to implement ‘their’ plan. In this study, a team leader viewed senior physicians as important and informal ‘leaders’, who should be regularly consulted. As a result, clinical staff reported having trust and confidence in the leadership team.

However, sharing leadership could also have some negative effects for teams, particularly when those who were given the power to lead were unwilling or unable to do so, due to, for example, a lack of time or motivation. 203 When this occurred, tension was created within teams. 203 In addition, on occasion, hierarchical leadership was deemed to be necessary or inevitable, due to prescribed lines of reporting and accountability. Hierarchical leadership was believed to be required when dealing with legal or policy challenges206 or when tackling performance. 105 Some authors also talked of the need for team leaders to balance power across organisations. For example, in a Belgian study of interprofessional collaboration between GPs and emergency department teams,195 emergency department leaders played a significant role in balancing power hierarchies across organisations. They did this by reassuring team members that they did not always need to abide by the requests of GPs for medical examinations but should instead rely on their own clinical assessments. Others talked of the importance of team leaders using their power flexibly by knowing when they should ‘bend the rules’ for the benefit of collaboration. 210

An influencing context of this mechanism was to whom team leaders felt responsible. In one study,216 comparisons were made between two different team leaders: one leader felt that their main responsibility was to the hospital ‘top’ management, whereas the other felt that their loyalties were divided between top management and their team. Both leaders were given the same formal mandate from top management to reduce costs but responded to this request differently (Box 12). The leader whose loyalties lay with top management complied with their top-down directives and tried to integrate their clinical departments in a radical and rapid way. They used coercive control and a one-way management style to communicate between top management and staff members and sought to justify the integration by promoting management’s rationale. Despite similar pressures, the other leader chose a freer interpretation of the formal mandate and consulted with senior physicians to work towards a slower and more gradual integration, in line with the wishes of their clinical staff. They asked for voluntary participation of staff members in the integration activities and respected their need for autonomy. They also used deputy leaders, such as research staff, to highlight the benefits of integration for research excellence as the main merger rationale. Consequently, the first leader faced anger and opposition from their team, who eventually demanded their resignation, whereas the other retained their trust. The repercussions of this early negative leadership experience continued to be far reaching for the team and resulted in longer term problems with integration. For example, a rapid turnover of managers followed, which further increased the teams’ antagonism towards top management and made staff reluctant to participate in any form of integrated working. In comparison, the team with the more inclusive leader had more positive long-term experiences of integration, including low staff and manager turnover, strong group cohesion and a 10% cost saving. 216

There was some evidence of the absence of the ‘working with power’ mechanism, with one study210 acknowledging that it could be very difficult for leaders to relinquish control, because of concerns that this might make them look ‘weak’. Another previously mentioned study216 also identified an example of a team leader refusing to relinquish control to their team members. In this study, the team leader asked their staff to suggest a new structure for their newly merged department but then overrode the team’s proposal with their own, more ambitious, plan. This move upset many team members, particularly the physicians and led to feelings of mistrust and suspicion towards the leader (Box 13). From then on, many physicians no longer attended collaboration meetings for the merger. As one senior physician stated:

He had obviously decided on his own plan that had nothing to do with all our work. That was the first nail in the coffin.

Choi et al. 216

Overarching leadership qualities around ‘learning and development’

An important component of integrated working was having the opportunity to share experiences and learn from others. 22 An overarching context for this mechanism at both system and team levels was the presence of an organisational culture that demonstrated mutual respect and understanding, which was crucial for nurturing learning and innovation within organisations. 199 Interprofessional and interorganisational training and education programmes were also described as key in helping teams to break down misconceptions and support integrated working. 22,197,199,200 A lack of training could lead to staff uncertainty about whether or not they were permitted to undertake particular tasks or feeling unprepared to take on new roles. 197 However, despite its importance, it was unclear whether or not resources and commitment were always readily available to support interprofessional and interorganisational training and education. 199,200

System-level leadership

There was evidence of the importance of the ‘learning and development’ mechanism at a system level, particularly around the willingness and flexibility of integrated systems to evaluate, change and evolve according to need. 197,199,215,218,583 For example, a paper reporting the results of a 3-year, national evaluation of 16 integrated care pilots in England197 found that the delivery of integrated team objectives relied on their overall ability to modify existing systems to create new ones. This was more difficult for organisations that were more accustomed to ‘silo-working’ or where funding models supported care in silos. 197,199 Furthermore, a study of integrated health care delivery in Western Australia found it important that partnerships were observed to be instigating change and overcoming bureaucratic and infrastructural barriers to do so. 215 Regular formal and informal meetings between organisational partnerships, staff and the community provided effective governance of this. 215 A study of community health partnerships in the USA also highlighted the importance of system-level leaders identifying potential team-level leaders and supporting and mentoring them to aid their development. 208

However, examples were also provided of where the ‘learning and development’ mechanism was absent at a system level. For example, an Australian case study contrasting two contexts of health-care governance found it imperative that board-level leaders recognised the need for innovation and supported it as a key strategy for integrated working rather than implementing and imposing existing policies. 199 Furthermore, there were reports of a focus on short-term political gains rather than long-term solutions and examples of ‘centralised bureaucratic control’ that resulted in a culture of risk aversion at the board level and that reduced teams’ ability to innovate, even when individual members were willing and capable of doing so199 (Box 14). Previously, a study noted that innovation could be stifled when systems leaders focused too intently on consensus as an important part of their culture. 220 Innovation required some dissensus among team members but if systems leaders perceived dissent as undesirable then teams would not gain the momentum required to move towards new ways of working.

Team-level leadership

It was important for team leaders to both inspire learning in their teams and show commitment to learning themselves. For example, leaders encouraged learning and development among their team members by placing an emphasis on the importance of information-sharing, reflective practice and innovation to maximise the potential of integration. 105,205,217 They were focused on developing the credibility of their staff and enhancing their performance as well as providing strategic and intellectual leadership. 84 They enabled innovation by creating a culture whereby staff could introduce new ideas with only limited checks and balances. That is:

. . . all we have to do is run it past the manager and do it.

Williams105

Leaders also gave staff time away from their normal duties to attend project meetings and seek other integrative opportunities,217 and they enabled staff to find the answers to their questions themselves, rather than simply providing them with the answers. 105,205 At a personal level, it was important for leaders to be interested in and willing to personally invest time in learning about team members’ roles and responsibilities to successfully manage the ‘bigger picture’. 92

An influencing context of this mechanism was whether or not teams had strong interprofessional relationships, since where networks were strong and social capital high, then learning with and from each other was more likely to occur. 205 An inhibiting context to learning and development was observed when team leaders feared thinking ‘outside the box’ or doing something new and innovative, because of a lack of either time or enthusiasm within teams to support such change. 199 An outcome of an environment of learning and development in integrated teams was the creation of team spirit and ownership, but this could also lead to ‘groupthink’ and intransigence if leaders did not carefully guard against this. 105 It was also noted that the difficulty of measuring the ‘success’ of new integrated arrangements through robust, evidence-based research made it hard to convince sceptics of the merits of adopting new ways of working. 205

Overarching leadership qualities of ‘taking a wider view’

Certain components of ‘taking a wider view’ were identified as important at both a systems and team level. For example, a UK study of multiagency working across health, social services and education settings identified having tenacity and a strategic drive and vision as key leadership skills at any level. 207

System-level leadership

At a system level, it was important that leaders had a deep, intuitive sense of how their community worked and what it needed. 84,208,217 Systems leaders needed to look beyond the interests of individual organisations and even the interests of the integrated system and focus on the needs and priorities of the community as a whole as well as on higher order cause, effect and prevention, rather than on symptoms or quick fixes. 208 To achieve this, systems leaders needed to have strong, pre-existing networks84 and to demonstrate a willingness to work with, listen to and learn from a range of different individuals and organisations from across the community, from both inside the integrated system and out. 208,217,583 This included leaders of government, business, community, and non-profit organisations and so on. Leaders of integrated systems were required to be skilled in navigating through complex and sensitive political issues84 and have the ability to help team leaders from their partnering organisations to see how integrated working benefited them in achieving their specific mission and goals. 208 There was evidence that taking a wider view and seeking out members that represented a wide range of knowledge and expertise could help achieve successful funding requests from public funding bodies, who often expect a multidisciplinary vision. 583

Other authors highlighted the importance of systems leaders knowing who were the ‘right’ people to engage with at a strategic level, which often included like-minded individuals who had previously worked across settings and who sought new ways of working to achieve change. 207 In integrated health and social care systems, the importance of ongoing, planned communications between senior executives in the partner organisations was also noted. 197,208,218 In a study of the clinical merging of two hospital sites in the USA,218 executive leaders at both hospital sites were responsible for designing and subsequently implementing the vision and future of the newly merged entity (Box 15). It was understood from the beginning that for the merger to succeed, these leaders had to be willing to put aside personal interests to convince the rest of the organisations that the integrated health-care system would be a better option for all parties. They achieved this through ‘communication, compromise and time in getting to know one another’218 so that trust was developed between the executive staff and the senior administrators and clinical leaders at both hospital sites. The most effective means of achieving this trust was the provision of multiple and regular common platforms for communication with the whole leadership group while simultaneously keeping local meetings at the sites to a minimum. The group also ensured that no programme would be moved from one site to another without considering the impact that this would have on the sites. 218

There was also evidence of the absence of the ‘taking a wider view’ mechanism at the system level. For example, there were reports of some care trusts having difficulty forging new relationships with wider partners, such as voluntary sector organisations. 220 The authors concluded that it may be difficult to pursue depth and breadth of partnerships simultaneously and the formal mechanisms of the former may be more influential than the informal arrangements that may encourage the latter. 220

Team-level leadership

It was also important for team-level leaders to see beyond their own organisational units and maintain relationships and network with leaders from external and partnering organisations. 105,197,198,206,210,217,221 This was identified as both empowering for leaders themselves and beneficial for co-ordinating the work of teams to provide holistic and integrated services. 198,206,221 For example, a UK study evaluating integrated team management in a health and social care trust221 found that team leaders needed to develop partnerships with a wide range of potential and current services, including those in the voluntary sector, wider local authority services and informal community groups. This enabled them to reshape services and prevented a narrow focus on the health and social care sector.

A Norwegian study of leadership in interprofessional collaboration in residential child-care services, previously mentioned,206 found that, to be able to communicate externally, leaders had to organise regular communication structures that matched their internal ones. This included creating regular meetings with external partners for the purposes of updating information, making decisions and plans, and delegating tasks and responsibilities. Extensive telephone communication between meetings and the production of various written reports were also required. Other authors also highlighted the importance of regular meetings between team leaders of external or partnering organisations to help maintain constructive conversations and prevent working in silos or a culture of blame. 197

A context that restricted this mechanism was identified in a Swedish study exploring collaboration between welfare agencies in the field of vocational rehabilitation. 210 In this study, leaders were evaluated primarily for their contributions to their own agency and not for their capacity to collaborate with external agencies. This restricted wider working and could have negative implications for the careers of those leaders who chose to dedicate their time and attention to working with external partners (Box 16).

Overarching ‘clarifying’ leadership qualities

There were certain components of the ‘clarifying complexity’ mechanism that were identified as important at both a systems and a team level. This included being clear about the management structure, the contributions required from all participants in the system/team and the rules governing how the partnership should work. 92,197 A paper reporting the results of a 3-year, national evaluation of 16 integrated care pilots in England197 found that staff felt unprepared and demotivated when there was an absence of clear and consistent communication from leaders about what work was required and what contribution was needed from all participants. Staff were also reluctant to engage when there was uncertainty about what they did and did not have permission to do. One influencing context on leader clarity is, therefore, whether or not leaders have formal frameworks governing their institution/organisation, such as legal frameworks, policies or guidelines. 206

System-level leadership

There was some evidence of the presence of the ‘clarifying complexity’ mechanism at a systems leadership level. For example, there was evidence that systems leaders needed the ability to navigate through complex and sensitive political issues and possess the relevant skills to oversee and manage complex clinical governance frameworks and practices, workgroup structures and systems and financial structures and systems. 84 This included the ability to introduce changes in a controlled manner, so that teams were not overwhelmed or confused by change. 583 A UK study exploring multiagency working across education, health and social services sectors found that having clear directives at a strategic level from all agencies involved, as well as from government, was key to overcoming conflicting agency and project aims. 207 System-level leaders were also required to keep their attention focused on any changes associated with integration, remove any obstacles to integration and reduce uncertainties associated with change. 217

There was also evidence of the absence of the ‘clarifying complexity’ mechanism at a system level. For example, a study199 found that a lack of clarity at board and executive levels led to a lack of enthusiasm and direction in the system. Another study216 reported that staff felt frustrated by a lack of clarity at a system level, caused by a complex reporting structure following a merger of two clinical departments in Swedish hospitals (Box 17). As one member of staff described:

. . . I don’t just have one boss. I now have many bosses . . . I have to go to the HR managers, division managers, clinical managers . . . I have about five bosses.

Choi et al. 216

Team-level leadership

There was evidence of the absence of the ‘clarifying complexity’ mechanism at a team level, including evidence of a lack of clarity among team leaders around their own role and responsibilities. 200 Other papers reported a team leader’s inability to clearly communicate their reasoning and priorities for the team as being detrimental to integrated team working. 216

Facilitating contexts

Hindering contexts

Inspiring intent to work together

Early discussions of this mechanism during the second meeting reiterated the importance that stakeholders placed on the debate around who the leader is, as opposed to what the leader does. In the third meeting, stakeholders’ initial reaction to the notion of ‘inspiring leaders’ was to question the implication that inspiration alone was objectively positive or valuable. Although one stakeholder felt that being inspirational was the most important aspect of leadership in their area of work (and indeed, the one trait that could not be taught or developed through training), others felt that inspirational leadership alone was inadequate and that leaders needed to be more rounded than the definition of the mechanism implied:

Whilst inspiration is important, other qualities are equally important. To me, the ideal would be someone who is inspiring but equally skilled as a systems-level leader. This suggests that the leader has not only the inspirational skills but has also taken the time to do the work and find out what is involved on the ground. Inspiration is definitely not enough, the person has to be grounded as well.

. . . it isn’t about the personality or the characteristics, it is just about them being able to do their job effectively, efficiently and safely.

Furthermore, although developing a shared vision was identified as key in the literature, stakeholders noted that little emphasis was given to leaders creating this vision around the needs of the individual patient/service user. They also highlighted the need for a leader’s vision to be authentic, credible and ethical if it was to be sustainable. In their experience, a leader’s vision would be shared by some and rejected by others if it were not representative of a credible aim:

History is replete with inspiring leaders who can pull together good teamwork but without the thought about what makes their inspiring vision in some way ethical or ‘good’.

I think there is also something about the authenticity of the vision that’s been put in front of people. In a hierarchical system there is a tendency for people to say and not do or believe because that particular vision doesn’t hold authenticity for them.

Creating the conditions to work together

Stakeholders agreed that inspiration alone was insufficient or even irrelevant in comparison to effectively fulfilling the full remit of requirements of the leadership role. However, it was suggested that the literature focused excessively on the practical attributes of ‘creating the conditions’ and ignored the social skills also required. In their experience, ‘creating the conditions’ required leaders to have the emotional intelligence to create a culture of psychological safety, including developing an environment of transparency, openness and freedom to communicate without fear of repercussion. This required leaders of integrated teams and systems to have self-awareness, self-regulation, motivation and empathy. Consequently, one participant suggested that leaders of integrated care teams and systems were required to be more ‘evolved’ than their business world equivalents:

. . . emotional intelligence is of particular relevance here, as traditional models of business leaders always has intelligence, toughness, determination and vision. However emotional intelligence brings in aspects . . . such as self-awareness, self-regulation, motivation, empathy and social skills.

Balancing multiple perspectives

Stakeholders agreed that conflict and discord would be inevitable across integrated care, but there was a lack of consensus over the way in which leaders should respond to them. Although the research literature generally posited that a leader’s role was to resolve conflict, stakeholders’ experiences were more diverse. Some suggested that conflict should be actively and intelligently resolved by leadership, but others felt that it was more productive for conflict to be ‘held’ by leaders, as they challenged their team to work out problems themselves. These stakeholders felt that a willingness to have difficult conversations with colleagues across different organisations and specialisms was evidence of a leader’s attempts to deal with conflict, without attempting to resolve it directly:

. . . where there has been a holding of two opposing views or difficulties in a system . . . that leads to perpetuating uncertainty, which fractures people’s views, and it can be quite divisive. In order to make progress in system integration, you would need to resolve those dilemmas one way or the other, and then it’s about the skill and art of leadership to take everybody through that.

System leadership is a series of confrontations of dilemmas. The theory goes that dilemmas are to be held rather than resolved, so for me there’s something missing in this, in the context. There’s an assumption in all of this that might be helped if the context is articulated as complex and therefore characterised by dilemmas.

Working with power

As previously stated, the ‘working with power’ mechanism was originally identified by the stakeholder consultation group and continued to evoke discussion in the final meeting. There was comment that power was palpable in integrated systems, particularly at higher levels, and could be both permissive and obstructive. Power was described as working positively and negatively, with stakeholders commenting that even when it was obstructive it was possible to redirect so that it became useful. It was felt that power generally moved through the system, rather than operating from top to bottom and this could have a range of positive implications for leadership and notions of shared ownership and responsibility. The tension that this created with reference to accountability, however, was also discussed, with stakeholders concluding that somebody needed to hold ultimate responsibility. Stakeholders also discussed the notion of ‘borrowed’ power, whereby permission or advocacy (implicit or explicit) of somebody with decision-making power could be productive when distributed evenly throughout an integrated team:

I would very much resonate with the concept of ‘borrowed’ power, so the fact that this somewhat intangible thing exists but if you know you’ve got the implicit permission and support of somebody who does have a clear decision-making power available to them, then you can crack on using that borrowed power to support and enable your colleagues to work.

Others commented on the need for reflexive leadership with regards to power, stating that reflexive leaders used power in ways that offered alternative interpretations. Through these, failure could be rationalised by a reflexive leader, by moving away from personalising shortcomings and exploring the structural and systemic motivations behind things that were perceived to go wrong. This highly nuanced approach to leadership was thought to rely, to some extent, on a paradigmatic shift, which raised questions around its feasibility:

. . . leadership should be reflexive rather than an embodiment of the ‘right way’.

A key context of the mechanism highlighted by stakeholders was the differential perceptions which surrounded integration, noting that some may perceive integration to be fair and equitable whereas others may perceive it as an overwhelming, authoritarian takeover. Leaders of integrated care should therefore be aware of the way in which power could be enacted through subconscious, endemic perspectives of the inequalities between health and social care and between other professional and sectoral boundaries:

. . . just because we say that something is integrated doesn’t mean that it actually is, integration can feel like a takeover rather than integration . . . the way that was experienced may be very different depending on the power that people have felt they are bringing from their respective organisational cultures . . . the power probably never feels equal, however integrated it may be. Experiences of people going through that may be very different, a good leader would need to be able to work with that.

Trust was felt to be another key context of the mechanism. That is, trust triggered the effective operationalisation of power, whereas its absence enabled more problematic power relations. Consequently, it was recommended that leaders be clear when discussing how power is operationalised, as instinctive reactions could position power more negatively:

. . . what is power? . . . It’s about getting people to do things without using coercion. It seems quite a negative thing, which can be highly disruptive, and that’s because it’s finely balanced with its opposite, which is trust. Power is getting and telling people to do things and trust is about enabling people and being assured that people will go off and do the right thing . . . I think there’s a nexus between those two, they’re very closely related and power can easily destroy team relations if it’s abused.

Stakeholders also referred to the absence of this mechanism and its implications for integrated care. They made connections between the historic power imbalances between health and social care highlighted in the previous mechanism and their associations with power inequalities. Stakeholders used the topical example of the COVID-19 pandemic to illustrate the inequalities of power between health and social care and the negative impact that this subsequently had on care homes:

. . . that’s a really great example of the absence of power being wielded, and the absence of decision-making and people not using their power in a way which was timely and effective led to the disasters we’ve seen over the first part of COVID. The absence of power sometimes can be hugely negative, and not using it in a proactive way can lead to as much disaster as obstructing people.

Stakeholders also referred to the absence of the service user voice throughout the research literature and the lack of power patients/service users held within integrated care teams and systems:

Exploring professional hierarchies tend to focus on the more powerful professions and the service user voice is largely absent.

There is a culture of deference in this country, as a leader comes to a situation, they are already endowed with acceptance . . . It presents so much difficulty as a user of services.

Commitment to learning and development

Stakeholder discussions on this mechanism tended to focus on the role of innovation and creativity in integrated care, exploring how leaders could support their team to work in innovative ways. It was felt that mid-level leaders only had scope to innovate themselves when they were given the flexibility to work in different ways and were provided with the ‘professional safety’ to do so by their leaders:

. . . to let them know that its ok to work in a different way, and that they need to provide provision around the risk of having a flexible job and changing roles, so provide some reassurance there.

However, the tension between the need for leaders to innovate and the need to take risks in an environment in which they were heavily accountable and constrained by a raft of regulations was felt to represent an impasse for the development of ideas. Mismatched roles and expectations of senior and mid-level managers also served to distort relationships and lead them to become dysfunctional.

To encourage others to work flexibly and innovatively, mid-level managers also needed to offer their team an environment of psychological safety, which included developing a culture where staff had the ‘freedom to fail’. This freedom to fail was elaborated by the contention that huge knowledge gaps existed in many aspects of health and social care and this would always necessitate a leap of faith by those both in practice and at leadership levels. There was consensus that current environments would benefit from a cultural shift, in which individuals felt comfortable to both feed their curiosity and develop in a way that accepted inevitable mistakes. The tension between innovation and the practical constraints of working in integrated care was also highlighted. There was suggestion that effective leaders needed to understand the gap between their own visions and the reality of working in integrated care to understand the lived experiences of their team:

. . . if you don’t feel like you can disclose what your real world is like, then they won’t hear what that real world is like. So it is tied in with the cultural implications of a psychologically safe environment and the ability of the leader to just sit and simply listen, so the staff don’t fear the repercussions of actually telling them things without punishment or sanctions. So that builds that concept of work as imagined versus work as done versus work as prescribed versus work as disclosed.

Taking a wider view

No research evidence was found to support the requirement of leaders to be ‘politically astute’, yet stakeholders felt that leaders could not operate without a sense of political leadership. One stakeholder described leadership in the NHS as being effective when ‘political capital’ was relinquished, to allow other leaders to bring in their knowledge, contacts and networks. It was also felt that people often intended to work together but were unable to and that leaders needed to look beyond the parameters of practice to provide non-conventional opportunities for integrated working. The research literature highlighted the importance of leaders knowing who were ‘the “right” people’ to engage with at a strategic level, but stakeholders felt that this could result in leaders seeking connections with ‘like-minded’ individuals, creating a bias in outlook and undermining the diverse perspectives that leaders were described to adopt in the initial programme theory. It was suggested that associating only with like-minded individuals was reflective of anxiety, which may have been caused by (or contributed to) the disruptive and unsettling capacity of integrated care:

. . . there’s something for me here about going to like-minded people when I’m feeling anxious. So there’s something about being aware of how anxiety plays out in individuals, groups and systems.

It was commented that working only with like-minded professionals encouraged the continuity of conventional assumptions around leadership and raised questions about the diversity of integrated care:

. . . senior leadership in the NHS in particular is framed around talking to like-minded people and that feels embedded in the language I’m hearing here from these mechanisms.

Stakeholders therefore suggested that there was a danger of these findings reinforcing existing assumptions around leadership. Noting that the research evidence for this mechanism was scant, stakeholders questioned whether or not the lack of literature could be reflective of how hard it is to both practically lead integrated teams or systems and make sense of this.

Clarifying complexity

Although there was little research evidence for this mechanism, stakeholders maintained the important responsibility of leaders to clarify complexity in integrated care. It was commented that policy documents, such as the NHS Long Term Plan,8 did little to address the day-to-day complexities of working in an integrated system and that there was no leadership blueprint in these settings. Discussions were held around the tendency of leaders to confuse the notion of ‘complexity’ with that of ‘complicated’, highlighting that many of the decisions to be made by integrated care leaders were more nuanced and challenging than the term ‘complicated’ implied (i.e. about more than simply sorting out complicated practical circumstances):

From my experience of talking to senior leaders there’s a tremendous pull to just reframe complexity as just being complicated . . . and some of the people I talk to perceive things as complexity but then judge themselves, saying ‘why can’t I make decisions around how we’re going to work together?’ . . . I seem to spend a lot of time around people who are getting evaluated very harshly as if they’re failing to come to decisions, whereas the situations that they describe are ones of complexity.

There was an implication that the way in which a leader accurately judged the context of a situation and responded to complexity and how this had an impact on their decision-making could be indicative of their perceived effectiveness. However, stakeholders felt that the way in which a leader talked to people during the decision-making process, particularly where there were multiple and sometimes incommensurate perspectives and interests, was the critical criterion, not necessarily the result of those decisions. This, again, spoke to ideas of ‘holding’ or ‘containing’ challenges, rather than actively attempting to overcome them; a concept missing in the research evidence:

. . . the psychological therapeutic terms of holding and containment . . . has been absent in some of this.

Fostering resilience

Although no research evidence was identified for this mechanism, stakeholders did have comments to make. Early in the study, stakeholders identified this mechanism as being a more generic aspect of leadership and less specific to leading integrated teams, although they suggested that we search for research evidence for anything more specific. They were unsurprised that no research evidence emerged to support the initial programme theory.

Stakeholders commented that no amount of resilience could help leaders of integrated systems with complex issues. Some felt that the term ‘resilience’ was code for ignoring or managing anxiety, without understanding why that anxiety had been evoked. They, therefore, felt that successful leaders were those with an understanding of the necessity, practicality, political nature and intense psychological impact of providing leadership. There was also suggestion of a psychological dynamic below the surface of integrated working, which became overt in bullying and uncivil behaviours. Stakeholders felt that it would be useful to explore the cultures that leaders set around this and if they were complicit, resistant or able to explain it. The longitudinal nature of developing effective ICSs was also highlighted, with stakeholders noting that it was a non-linear process and that resilience building was a long-term endeavour.

General discussion

Stakeholders provided comments around leading integrated systems that were unrelated to the identified mechanisms. For example, the influence of professional anxiety on patients, service users, and families was raised, which reinforced the need for leaders to address the psychological effects of system change. Another participant queried how this study aligned with current contexts, such as the Black Lives Matter movement. This was significant, as research evidence had implicitly suggested conventional assumptions and actions surrounding leadership. There is a pressing need for leadership of integrated care to address this through an awareness of the necessity of anxiety, disturbance and disruption that comes with challenging the status quo.

Paucity of empirical evidence

There was little evidence that specifically addressed leadership of integrated care teams and systems despite the widespread policy rhetoric in support of, and partial implementation of, this model of organising services. This was surprising. We assessed many papers and information sources, but most referred to older leadership theory, which was not tailored to understanding how leaders contributed to complex multisector organisations. There are several potential reasons for this. These include:

• Attention on and development of integrated care teams and systems is relatively recent, with the result that little research has been undertaken or completed.

• Research on integrated care teams and systems has focused on the implementation and outcome of service innovation rather than some of its structural underpinnings.

• The lack of literature could reflect how hard it is to research this area and a reliance on existing framings of leadership when teams and organisations are less complex.

• There are assumptions about leadership being a homogeneous activity that is transferable across many different settings and types of service and, therefore, it is not a research priority.

It is also important to note that there was very little evidence of the contexts that influenced how leaders work and even less evidence on the outcomes. Identifying CMO configurations was challenging. This limited the degree to which we could draw definitive conclusions about what works, for whom and in what circumstances. However, making explicit some of the assumptions about how leaders lead integrated care teams and systems has provided new perspectives offering fresh theoretical grounding that can be built on, developed and tested further.

There is an emphasis on the individual/personal qualities of the leader

The strongest evidence found in the review was around how leaders inspire people’s intent to work together in integrated care. This evidence focused on who the leader is rather than what the leader does. It referred to their personality, characteristics and ability to inspire other people. This is quite challenging to consider within aspirations for greater diversity and inclusion. One of our expert stakeholders asked: ‘If we just rely on people blowing us away with their personalities because they’re born like that, is that what it is?’. This seems counterintuitive. Why have extensive leadership programmes to develop, prepare and support leaders if the main qualities they need to lead well are innate? The stakeholders were clear that this was not enough; a leader must be knowledgeable, skilled and spend time on finding out what is involved on the ground. Within the ‘creating the conditions’ mechanism the focus was on the practical attributes of leadership, but stakeholders felt that it ignored the social skills also required. For example, stakeholders believed that ‘creating the conditions’ also involved creating a culture of psychological safety and an environment of transparency, openness and freedom to communicate without fear of repercussions.

The stakeholders referred to aviation safety, stating, ‘you don’t really care who the pilot is and how inspirational the pilot is when you get on the aeroplane, you just want them to fly the plane safely’. This overemphasis on the influence of individuals is recognised in leadership theory. The culture of individual leaders as ‘heroes’, romanticising their individual abilities and dispositions, is thought to prioritise the importance of being in control and having the power to decide, steer and influence others. This overestimates their contribution and influence598,599 and can obscure, in part, the tensions and complexities inherent in leadership. 600

An alternative perspective is processual leadership,601 which views leadership as an ongoing process of social interaction and negotiation with all members and stakeholders of an organisation who participate in and influence the organisation’s activity. Processual leadership acknowledges the complexities of social reality. It focuses on how leadership is socially constructed, deconstructed, and reconstructed in a fluid and dynamic way as relationships and circumstances change. 602 Through these ongoing processes, power and agency are negotiated between all actors in the organisation (e.g. leaders, staff and patients/service users) and no one falls outside these complex organisational processes. 599 This is attuned to the stakeholders’ views of the importance of social skills and emotional intelligence. Developing processual leadership practices that are attuned to a complex, changing, organisational context is thought to require recognition of the value of disagreement, tensions and dissent. This is not to decry consensus but acknowledges that it is not always achievable where stakeholders inevitably have different views and professional values. Leadership in organisations that are characterised by complexity and ambiguity like integrated care systems require a ‘viable sense of self-as-a-leader’ where leaders accept that social reality is constantly changing, and that control is an illusion. 599 However, maintaining a processual approach of leadership in an organisation or system where the dominant conception of the ‘hero’ leader is deeply embedded and where leaders perceive high levels of ambiguity and insecurity is extremely difficult. 599 The theory of ontological security595,596 provides a useful perspective to our understanding that leaders need to feel comfortable and secure in a climate of uncertainty to provide this security to the systems and team members that they lead.

We propose that perpetuating the importance of individual characteristics of leaders reduces the scope for integrated care teams and systems to develop leadership practices that are attuned to the complexity of multi-sectoral, multiorganisational and multiprofessional working. Furthermore, in line with The NHS Long Term Plan,8 42 integrated care systems have been created to cover the whole of England from April 2021, growing out of Sustainability and Transformation Partnerships. It is intended that they will become commissioning bodies that have a partnership governance structure and performance framework and need to be held accountable, committing the NHS and partners to increased complexity of health and social care services. There is a timely need to ensure that the skills of integrated care teams and systems leaders are intentionally developed to keep pace with service reorganisations.

Absence of evidence of patient/service user perspective

A central policy driver for the introduction of ICSs was the need for services to be ‘integrated around the patient’,7 to provide the best patient/service user experience and the best value for money. The ambition was to support patients/service users to be actively involved in their own care, give them more power to manage their own health and make informed decisions about their care and treatment. Furthermore, NHS England’s Patient and Public Participation Policy603 states that:

One of the great strengths of this country is that we have an NHS that – at its best – is ‘of the people, by the people and for the people’.

NHS England603

It also adds that:

. . . we need to engage with communities and citizens in new ways, involving them directly in decisions about the future of health and care services.

NHS England603

It is a stark finding that we found no evidence of the patient/service user’s perspective of leadership or their involvement in leadership of integrated care teams and systems.

The importance of power

This synthesis found the exercise of the power and influence of leaders to be very important. Power was held by many people to varying degrees, by government, across sectors, in organisations and by individuals at all levels of activity. As discussed previously, there was no evidence of power and influence being held by patients/service users in this context. Historic imbalances of power between health and social care were clear and irrefutably demonstrated by the response to managing surge capacity during the first wave of the COVID-19 pandemic. The NHS Long Term Plan,8 which sets out plans for the integration of health and social care throughout England is NHS-centric and potentially reinforces significant existing tensions between health and social care. New legislation to create a ‘legal form’ for integrated care systems is expected in 2021. 604 The NHS Long Term Plan8 did not call for new statutory bodies, but we believe that it will be important to provide clear power-sharing requirements to protect social care from being disadvantaged. This synthesis also demonstrated that higher levels of influence and power were held by members of the medical profession. Honesty and reciprocal trust were important in how power was used, which links to the individual characteristics of the leader and being known by those being led. The nature of power is far more complex and nuanced than the evidence suggests, and questions remain about how leaders of integrated care teams and systems see their power and reason how to use it. With the large size of integrated care systems and the numerous members of staff potentially involved, how realistic is it that leaders will be known sufficiently to be credible and trusted and to know the nature of the services that they are leading to make decisions?

The benefits of and barriers to pre-existing networks

The importance of system leaders looking beyond the interests of their own organisations and the integrated system to focus on the needs of the community was discussed in the research evidence. However, to facilitate this, leaders drew upon their pre-existing networks of like-minded people and those considered to be the ‘right people’ to engage with strategically. This would appear to be in contradiction to taking a wider view of the system. The benefits of pre-existing networks include that leaders are known and trusted, and progress in advancing innovation is likely to be faster, which is important given the political investment in the success of ICSs. It may also help leaders experiencing anxiety or uncertainty about the challenge of leading large, complex systems. However, a tendency to drift towards organisational, cultural and professional familiarities is likely to narrow the focus of innovation and this may inadvertently be a barrier to diversity in leadership. This lack of diversity could include sector, professional group, leadership style, service delivery priorities and protected characteristics, such as age, disability, ethnicity and sex. This reduction in diversity could reduce the potential for innovation, reach and equality of opportunity as demonstrated in the ‘Snowy White Peaks’ of the NHS report605 and hinder the ambition to ‘understand, encourage and celebrate diversity in all its forms’. 606 If leaders have a background in one organisation (e.g. as a senior leader in the NHS), they will be aware of the needs of that organisation and address them more easily. Furthermore, will integrated care disadvantage the perspective or involvement of smaller, less resourced, less experienced sectors who produce fewer senior systems leaders, such as charities?

Lack of practical guidance about how to lead in integrated care teams and systems

Throughout the evidence, there were general statements about the important activities that leaders undertook in leading integrated care teams and systems. For example, they managed conflict and mitigated tension among team members, they looked beyond the interests of their own individual organisations and they encouraged team members to develop and innovate. However, there was very little explanation about how leaders undertook these activities, their reasoning for what the best approach would be, the trade-offs they may have made, and the challenges that they encountered. This limits our understanding of what aspects of leadership work, for whom and in what circumstances. In the same way that clarifying complexity was thought by our stakeholders to be an important responsibility of leaders (albeit with little empirical evidence), leaders also required clear directives at a strategic level from all agencies involved, as well as from government, about leading challenging processes. However, policy documents were thought to do little to address the day-to-day complexities of working in an integrated system and our stakeholders felt that there was no blueprint for leadership in integrated care.

Implications for governance structures

New legislation in 2022 to create a ‘legal form’ of ICSs is expected at the time of writing. 604 The findings of this review suggest that it would be very important to ensure that legislation provides clear power-sharing requirements to protect social care and non-NHS organisations from being disadvantaged. Clarifying complexity and designing governance and incentives structures to support ICSs, rather than individual organisations within the systems, is likely to support leaders to take a wider view. This study will inform leadership structures for these new governance structures.

Implications for education preparation of leaders of integrated care teams and systems

The review did not find much evidence on specific education preparation for leaders of integrated care teams and systems. However, some important considerations were highlighted. These include:

• the importance of understanding the whole system, which suggests that leaders need a wider understanding of organisations

• the highly complex, dynamic nature of leading integrated care teams and systems and the imperative to adapt to varied circumstances demonstrates that leaders need to develop a viable sense of self-as-a-leader and be comfortable with uncertainty and ambiguity rather than the command and control approach that is common in the NHS

• acknowledgement of the importance of emotional intelligence, communication and social skills development in leadership preparation.

In view of the unique, highly situated context of leading integrated care teams and systems, we propose that leadership training needs to encompass not a generalised training programme or course, but instead bespoke, individualised mentoring/coaching programmes. Approaches that increase exposure and understanding of other sectors may be useful, such as work placements, coaching or secondments, where individuals have opportunities to shadow leaders from another part of the system, work with coaches from other sectors, or rotate between the range of organisations across the system. Innovation approaches to design leadership education and training to build networks across systems would be beneficial and would contribute to the plans outlined in the NHS People Plan. 606

Implications for individual leaders and integrated care teams and systems

This review carefully unpacked the available evidence on the assumptions of what leadership is and the unique challenges of leading large, complex, inter-sector systems. To our knowledge, this is the first realist review in this area and offers leaders insights about their actions that potentially affect care delivery and outcomes, and team and system working. We hope that this understanding supports leaders to reflect on their practice and the factors that may support them in their work.