1 Introduction

In many countries, increases in life expectancy entail a growing demand for long-term care and a rising number of patients with multiple chronic conditions. To meet these patients’ needs, scholars and policy makers recommend that health- and social care providers coordinate activities in inter-organisational networks to enhance care experiences and outcomes for patients and their families (Glasby et al. 2011; Goodwin et al. 2014; Minkman 2017). Despite these considerations, network-based service integration remains challenging, and progress has been limited. As Goodwin (2019) notes, “we have yet to make any significant breakthrough to understand the implementation and sustainability of complex service innovations that so characterise the development of integrated care programs” (p. 1). One of the major challenges in forming and developing integrated care networks concerns the design and use of effective governance arrangements. Struijs et al. (2015) point out that, “In all initiatives, multiple actors are involved, with their own organisational interests, leading to varying governance arrangements. How to best arrange these new governance arrangements … is still widely discussed and yet to be resolved” (p. 1). Against this background, research has called for more innovative thinking about the governance of integrated care networks (Minkman 2017).

In this chapter, we contribute to this debate by synthesising knowledge on the governance of integrated care networks. To provide a more meaningful portrait of this growing field of scholarship, we draw together three perspectives. The first perspective, “governance-as-structure”, addresses how the effectiveness of integrated care networks is determined by contingent network governance structures. The second perspective, “governance-as-process”, explores how effective network governance results from individual actors’ activities, skills and competencies. The third perspective, “governance-as-practice”, combines the two previous perspectives and studies governance as a situated practice in the context of evolving network structures. For each perspective, we identify its theoretical origins, empirical focus, illustrative empirical findings, critical reflections and avenues for future research. Table 11.1 illustrates the key dimensions of the three perspectives that will be discussed in the following sections. Before going into depth, we define the key terms used in this chapter.

Table 11.1 Perspectives on the governance of integrated care networks
Full size table

2 Conceptual Background

The three key terms used in this chapter—integrated care, inter-organisational network and network governance—are all “polymorphous” concepts that have been defined from various theoretical and disciplinary angles and with multiple objectives (Goodwin et al. 2017, p. 5). To establish common ground, this section defines the three terms and specifies the conceptual boundaries guiding this chapter.

The term “integrated care” has been defined in various ways and our understanding of what integrated care “is” and what it comprises still evolves (Goodwin 2016). For the purpose of this chapter, we follow Kodner and Spreeuwenberg, who define it as “a coherent set of methods and models on the funding, administrative, organisational, service delivery and clinical levels designed to create connectivity, alignment and collaboration within and between the cure and care sectors” (adapted from Kodner and Spreeuwenberg 2002, p. 3). Designing and using integrated care methods and models, actors aim to overcome inefficient care fragmentations and improve people’s care experiences and outcomes through the coordination of their service activities (Goodwin 2016). Although the definition explicitely refers to multiple levels of analysis, this chapter focuses on the inter-organisational level, while agreeing that a multi-level perspective is required for understanding and dealing with the complexity of integrated care methods and models.

Similar to integrated care, research on inter-organisational networks is highly fragmented. Huxham (2003) notes that “even the most basic terminology is subject to varied interpretations and there seems to be little agreement over usage of terms such as ‘partnership’, ‘alliance’, ‘collaboration’, ‘network’, or ‘inter-organizational relations’” (p. 402). This chapter draws on Müller-Seitz and Sydow (2012), who define a network as “a social system in which the activities of at least three formally independent legal entities are coordinated in time–space, i.e. there is some reflexively agreed upon inter-firm division of labour and cooperation among the network members” (p. 108). This definition excludes dyadic relationships, recognising that third actors give such relationships a distinct social quality, e.g. one actor’s option to play two or more others against each other for his or her own benefit (Simmel 1950; Sydow et al. 2016). Moreover, it is open to several types of integrated care networks, like cancer or diabetes networks, and includes multiple directions, i.e. vertical, horizontal, cross-sectoral or population networks (Goodwin et al. 2017).

Over the past three decades, a considerable literature has developed around the topic of network governance. A first research stream has discussed inter-organisational networks as a distinct mode of governance situated between markets and hierarchies (Powell 1990), analysing networks as a means to address complex (“wicked”) public policy problems like migration, global warming or health care (Rittel and Webber 1973; Rhodes 1997; Emerson et al. 2012). In health care, governance has been understood as policy tools and processes needed to steer a system towards population health, which entail a move from hierarchical models of service delivery towards network-based collaboration among a range of independent organisations across different sectors (see the contribution of Mallinson and Suter in this book). A second research stream, which is the focus of this chapter, has studied the governance of inter-organisational networks per se, identifying structural and processual determinants of effective network governance. In this stream, network governance entails the design and use of structures and processes enabling actors to direct, coordinate and allocate resources for the network as a whole and to account for its activities (Vangen et al. 2015). This chapter sheds light on this second stream and synthesises current knowledge into three perspectives on the governance of integrated care networks.

3 Three Perspectives on Governing Integrated Care Networks

This section identifies three perspectives on the governance of integrated care networks. For each perspective, it describes its theoretical origins, empirical focus and selected findings. The findings are selected for illustrative purposes without any claim to a systematic review. What is more, the three perspectives are presented as ideal types, which researchers sometimes combine in their actual empirical or theoretical work. The subsequent section will discuss the contributions and limitations of each perspective and suggest avenues for future research (see Table 11.1 for an overview).

Governance-as-Structure

The governance-as-structure perspective explores how the effectiveness of integrated care networks depends on governance structures, which involve the member organisations of a network and the formal relations between them (Vangen et al. 2015). Rooted in relational sociology (Simmel 1950) and social network analysis (White 1963), this perspective explains social behaviour with reference to patterns of relationships among actors like organisations. It thereby constitutes a theoretical alternative to methodological individualism, which explains social behaviour in terms of the activities and properties of individual actors. This perspective has gained significant momentum from the 1980s, when networks among organisations were considered as a promising way of dealing with “wicked” policy problems and a globalising business environment (Powell 1990). Noting a surge of inter-organisational networks in the private and public sector, researchers began to explore how network-level outcomes depend on various governance structures under different contingencies.

In a pioneering study, Provan and Milward (1995) proposed a theory of network effectiveness, conducting a comparative case study in four US mental health delivery networks. They explain network effectiveness by various structural and contextual factors including network integration, external control, system stability and environmental resource munificence. Their findings indicate that networks are more effective if they are tightly integrated and led by a central core agency. Centralised network governance facilitates the coordination, monitoring and control of activities. With respect to other structural and contextual factors, the study suggests that network effectiveness is enhanced if networks are fiscally directly controlled by the state (rather than by regional agencies) and under conditions of general system stability and resource abundance.

Continuing this line of research, Provan and Kenis (2008) delineated three forms of network governance (shared governance, lead organisation governance and governance by a network administrative organisation/NAO) and developed propositions about the relationship between these forms and network effectiveness under various contingencies. For instance, shared network governance is most effective for advancing network-level outcomes when there are only few network members, trust and goal consensus are high, and the need for network-level competencies is low (Provan and Kenis 2008). By contrast, NAO governance is more effective when there are a moderate number of network members, trust and goal consensus which are moderately high, and the need for network-level competencies is high.

Further advancing this perspective, Cristofoli and Markovic (2016) examined how the effectiveness of twelve home and social care networks is determined by various combinations of resource munificience, formalised coordination mechanisms (e.g. formal agreements of defined procedures) and forms of network governance. They find that high network performance is influenced by different combinations of these factors, suggesting that there might be several paths to success. In addition, they highlight the role of individual actors’ interventions, finding that in a resource-munificient context network, effectiveness depends not only on centralised network governance structures, but also the presence of managers promoting interaction, relieving tensions and supporting identification among network members. The observation of the importance of individual actors leads to the second perspective on governing integrated care networks, governance-as-process.

Governance-as-Process

This perspective examines how networks are governed through individual actors’ skills and activities, which include ways of communicating, sharing responsibility and taking decisions (Vangen et al. 2015). It is grounded in methodological individualism, which assumes that the elementary unit of social life is the individual human action (Lukes 2006). To explain integrated care networks and their change, it shows how they result from the activities of individuals rather than structural patterns of relationships, explaining action by referring to individual purposes, intentions and skills.

An example is the study of Vendetti et al. (2017), who examined the barriers and facilitators associated with the formation of alcohol and other substance abuse networks in the USA. They find that the successful formation of these networks depends on committed charismatic individuals driving progress by delivering strong and consistent messages regarding programme importance, and by encouraging communication among key stakeholders. In addition to the activities of these higher-level strategic network “champions”, network formation is supported by committed individuals at the operational level like nurses, who mobilise engagement for the network within different network member organisations.

These findings are corroborated by the work of Lyngso et al. (2016), who analysed the barriers and facilitators of network-based activity coordination for COPD patients in Denmark. They find that the effective coordination of activities among service provider organisations depends on managers who share a vision of integration with employees, acknowledge the tasks involved with inter-organisational activity coordination and allocate sufficient time to complete these tasks. In addition, managers support activity coordination by regularly arranging social events and informal network meetings to build up and strengthen personal relationships among involved health professionals.

Given the pivotal role of individual actors for building integrated care networks, research has addressed their required personal skills, abilities, competencies and experience. In a study including UK health promotion networks, Williams (2002) examined the skills and behaviours of “boundary spanners”. He finds that effective individuals building inter-organisational relationships have strong abilities in communicating and listening, understanding and resolving conflict, managing through influencing and negotiating, managing complexity and interdependencies and managing roles, accountabilities and motivations. He concludes that an understanding of these skills is important to inform the training, development and education of current and future integrated care practitioners (Williams 2002).

Governance-as-Practice

The third perspective, governance-as-practice, studies situated governance practices in the context of evolving network structures. Inspired by the practice turn in the social sciences (Schatzki et al. 2001) and theorists like Bourdieu (2013) or Giddens (1984), it explains social order by referring to practices rather than structural relationships or actors’ skills and purposeful activities (Reckwitz 2002). Explaining social order by practices means analysing routinised activities that are informed but not determined by structures and reproduce and change these structures over time (Schatzki 2005). Unlike the two previous perspectives, governance-as-practice studies the recursive interplay between social structure and action, suggesting that social action cannot be explained without considering the structural context in which it is embedded (Sydow et al. 2013). Vice versa, social structures have no “existence” independently of actors referring to them in their practices. In their practices, actors refer to previously established structures and thereby reproduce and modify them over time. Reproduce and modify means that actors do not simply reiterate practices of the past, but have a capacity to reflexively generate alternative trajectories in response to continuously emerging demands, tensions and contradictions in evolving situations (Emirbayer and Mische 1998).

A recent example illustrating this perspective is the study of Embuldeniya et al. (2018), who describe how care activities across organisational boundaries are coordinated by generating connectivity and consensus. The study finds that generating connectivity and consensus are not isolated individual activities, but social practices that are contextually embedded in histories of the existing cultures of clinician engagement and established partnerships. By showing how these practices are “contextually and temporally contingent, with the capacity to produce new contexts, which in turn generate new sets of mechanisms” (p. 783), it highlights a recursive relationship between social structure (cultures of clinician engagment, established partnership) and situated practice (generating connectivity and consensus). It thereby offers an analysis of how network-based activity coordination is enabled and constrained by the interplay of social structure and action in local contexts and histories.

In a related vein, the study of Mitterlechner (2018) explored how new trajectories of network governance evolve, analysing governance and activity coordination practices in a Swiss integrated care network over time. It finds that network governance and activity coordination evolve through repetitive sequences of collaborative inquiry, a practice through which involved network members jointly identify and address recurring contradictions in creative and experimental ways. It contributes to the governance-as-practice perspective by drawing attention to the pivotal role of meaning making, creativity and experimentation for understanding governance dynamics in integrated care networks.

A further example is the study of Martin et al. (2008), who showed how actors are not only enabled, but also constrained by evolving network structures as well as interfering market and hierarchical structures, observing the development of cancer networks in the UK. They examine the degree of convergence between the introduction of centrally mandated network structures and the possibilities of individual action for introducing structured cancer care pathways. In theory, they suggest that network structures, which are looser than hierarchical structures, should enable actors to implement change and coordinate service activities more effectively. In practice, however, they find that this is not always the case. While the structure of a network might indeed create an opportunity space for actors to coordinate activities, actors are simultaneously embedded in overlapping hierarchical (e.g. central performance management) and market structures (e.g. competition among network members), which constrain their ability to coordinate activities in the network without complimentary action at these two additional levels. The authors emphasise a recursive relationship between social structure and action, concluding that network structures have no causal force without embedded actors’ agency.

4 Discussion

This section discusses key contributions and limitations of the three perspectives and outlines possible avenues for future research. The governance-as-structure perspective advances the field by conceptualising integrated care networks and their governance as social structures aiming at coordinating care activities across network member organisations. It thereby establishes a separate level of analysis that is different from the level of the participating network member organisations and the institutional field in which the network and its members are embedded. The creation of a separate level of analysis makes it possible to direct the view from individual actors and their categorical attributes to the systemic level and explain social phenomena by means of patterns of relationships among network members. This allows researchers to provide precise representations of the governance structures of integrated care networks and evaluate the impact of these structures on network outcomes under various contingencies. These phenomena and causalities are operationalised and measured using modern instruments from social network analysis. Not least thanks to rapid methodological progress in this area, the governance-as-structure perspective has become an important and valuable domain in integrated care research.

While the governance-as-structure perspective captures the systemic level, the governance-as-process perspective draws attention to individual actors’ purposeful actions and competencies. It thereby responds to calls to bring individual actors back into the picture and avoid an overly structural view on integrated care networks and their governance. Williams (2002), for instance, notes that “comparatively little attention is accorded to the pivotal role of individual actors in the management of inter-organisational relationships” (p. 103). Similarly, Provan and Kenis (2008) demand that the role of network managers should be discussed in much more depth. The governance-as-process perspective redresses this imbalance, shedding light on crucial activities enabling the formation and development of integrated care networks. In addition, it draws attention to important personal competencies and thereby provides valuable insights for the training of current and future network managers advancing integrated care.

While both perspectives have enriched the debate, they are not without limitations. On one hand, the governance-as-structure perspective tends to lack an explicit concept of human agency, emphasising how actors are constrained by structure. Some authors in this tradition regard actors not as sources of action, but as vehicles for structurally induced action (e.g. Burt 1992), thereby downplaying how social structures and human agency presuppose each other. Parkhe et al. (2006) note that this perspective “risks understating the role of the very actors composing the network” (p. 561). On the other hand, the governance-as-process perspective tends to overestimate the possibilities of individual agency, ignoring how actors are situated in social structures simultaneously enabling and constraining (although not determining) their actions. In addition, the insights generated by both perspectives tend to be relatively static. Although the governance-as-structure perspective proposes optimal structures under various contingencies, it provides little explanation of how these structures change. Vice versa, reducing actors’ interventions and their consequences to sets of interrelated variables, the governance-as-process perspective tends to be similarly limited in its capacity to grasp the temporal experience of acting in integrated care networks (Denis et al. 2010). This limitation is unfortunate because dynamism and instability have been shown to be a central characteristics of successful networks (Majchrzak et al. 2015). To deal with it, Provan and Kenis (2008) call for more research on the evolution of networks, research «focusing … on how the governance of public networks emerges … and how it changes over time» (p. 248).

The third perspective identified in this chapter, governance-as-practice, is able to cope with some of these limitations. It considers social structure and action not as a dualism, but as a duality. As Giddens (1984) writes, “The constitution of agents and structures are not two independently given sets of phenomena, a dualism, but represent a duality” (p. 25). This theoretical approach enables researchers to study the evolution of network governance over time. It implies that social phenomena are always “in the making”—network governance is an ongoing accomplishment reproduced and possibly adjusted in every instance of practice (Feldman and Orlikowski 2011). At the same time, the governance-as-practice perspective is not without problems, either. It requires deep engagement in the field to study the dynamics of network governance over long periods of time, often many years. In addition, practice-theoretical accounts are not designed for statistical generalisations, which may limit their acceptance in certain journals and research communities. Instead of universal variation, they produce theoretical generalisations and thick descriptions of situated dynamics that can be useful in understanding governance dynamics in other contexts (Feldman and Orlikowski 2011). However, to the extent that there is a need to open the “black box” of network governance and its underlying social practices (Goodwin 2019), this perspective can serve as a valuable complement to the other two perspectives. Future research adopting this perspective could explore the role of tensions and contractions as potential sources of change, connecting the network, organisational and industry level of analysis (Berends and Sydow 2019).

5 Conclusion

Countries around the world adjust the way they deliver health and social care services, responding to the changing needs of an ageing population and people living with one or more chronic conditions. In many cases, service provider organisations break new ground and start coordinating activities in inter-organisational networks. However, despite best intentions, progress has remained limited, not least due to the challenge of governing these networks. This chapter aimed at identifying three perspectives on the governance of integrated care networks, describing network governance as structure, process and practice. By doing so, it has hopefully added important “conceptual sensors” (Weick 2016, p. 339) to increase the meaning of our experience with this complex social phenomenon.