**1. Introduction**

Integrated care is a concept used extensively to refer to advanced collaboration in the context of health and social care. To put it simply, integration means 'bringing together different actors or activities' (Axelsson and Axelsson 2009). Cross-sectoral collaboration, referring here to both multi-professional and inter-organizational collaboration of professionals and collaboration between clients and professionals, is needed to make care services for people more co-ordinated, flexible and continuous within the care systems, in order to produce better care for people. Person-centeredness is at the center of integrated care (World Health Organization 2016) and a holistic professional-client relationship forms a salient basis for cross-boundary collaboration. Multiprofessional refers here to collaboration between professionals (doctors, nurses, social workers) and interorganizational to collaboration over sectoral boundaries (primary health care, specialized health care, and social care).

Integrated care is expected to be of particular benefit for people with multiple complex health and social care problems who need services from several care providers in both health and social care (Goodwin 2015; Rijken et al. 2017). These people, from a collaboration perspective also called 'shared clients' (Oksman 2017), consist of somewhat different groups, such as those with multimorbidity, older persons regularly needing acute health services for several reasons, families with children who have special problems, drug users, young people or adults who suffer from mental problems. What these

client groups have in common is that they need services from several care providers at the same time: from hospitals, from health centers or general practitioners in primary health care and also from social care. The existing disease-based and fragmented care systems do not meet the needs of people with multiple complex problems (Hujala et al. 2017). Collaboration between clients and professionals and among various professionals and between the sectors of care–specialized care, primary care and social care–is crucial.

Person-centered integrated care has been offered as a solution to solve the problems described above regarding care for people with multiple complex problems (Ahgren 2012; Van der Heide et al. 2018). This reflective paper draws on experiences and insights from a Finnish study addressing care professionals developing integrated care pathways for 'shared clients'. The emotional and other relational dynamics that the study participants highlighted in relation to these clients inspired us to take a closer look at the emotional side of integrated care and to consider if an approach addressing emotions could give any added value to the concept and approach of integrated care.

Collaboration among care professionals has been studied extensively (Cameron et al. 2014; D'Amour et al. 2008; Schepman et al. 2015), often specifically addressing the barriers and difficulties of multi-professional collaboration (see Axelsson and Axelsson 2009). Previous research has shown that successful collaboration between professionals requires a shared goal, commitment, trust and respect between the participants (e.g., Willumsen et al. 2012; see also Schruijer 2008). These issues may sound quite familiar, but it is not always so easy to realize this ideal in practice. The reasons why cross-boundary collaboration is so challenging may be better understood by also paying attention to emotional tensions, which tend to pervade all interaction between human beings.

In the context of integrated care, challenges related to interaction between human beings, such as emotional dynamics, are seldom explicitly addressed. This is not exceptional given that emotions have long been 'the neglected side' of organizational research in general (Fox and Spector 2002). At least, integrated care is most often described as a phenomenon taking place at system, organizational, professional and clinical levels, including functional and normative dimensions (Valentijn et al. 2013). Although numerous approaches and models of integrated care are available (see, e.g., World Health Organization 2016), the focus of research on integrated care is often on the structures, processes and tools of integration. Processes of integrated care refer to collaborative processes, but the main focus of integrated care is functional addressing the concrete ways in which collaboration should be organized and paying less attention to the human side of integrated care. Normative integration, which perhaps has the closest relationship to emotional dimensions, does indeed address the different values and cultures of professionals. These aspects have recently been emphasized by an approach highlighting value-based integration (see Minkman 2016), aiming at a better understanding of the values underpinning integrated care. In addition, person-centeredness has become a core aim of integrated care, emphasizing the importance of taking into account patients' needs and wishes and focusing more on person-relevant outcomes of integrated care (Van der Heide et al. 2018).

However, in spite of the recent diversification of the perspectives on integrated care we venture to claim that mainstream research and practice focus on the conventional points of views of rationality, where human beings tend to be rendered as rational actors in a technical and practical framework. Less attention is paid to the social, emotional and affective dimensions of integrated care. As in planning processes in general, these may even be considered irrational and irrelevant (see Osborne and Grant-Smith 2015). Emotional and other relational dynamics are present in particular when implementing integrated care, because ultimate implementation takes place at grass-roo<sup>t</sup> level in interaction between human beings.

As a phenomenon, emotions can be approached from a variety of perspectives. Traditionally emotions have been considered to be individual and internal phenomena (Osborne and Grant-Smith 2015). Psychodynamics emphasize the nonconscious nature of emotions at individual level; systems psychodynamics (developed by the Tavistock Institute of Human Relations) link nonconsciouos dynamics with organizational structures and see that routines and practices can serve as a social

defence to manage the emotions of organizational members (Pratt and Crosina 2016). According to the constructionist approach, emotions do not exist inside us; they are not only an individual or internal phenomenon, but emerge from social interaction (Harré 1986; see also Gergen 1999). Our emotions are very closely connected to our relationships to other people (Burkitt 2014, p. 2); Brotheridge and Lee (2008, p. 109) claim that emotions are 'at the heart of all working relationships'. Further, emotions can be considered as emotional intelligence (e.g., Morrison 2007) or emotional competencies (e.g., Kinman and Grant 2011).

In this paper, we apply a broad understanding of emotions, recognizing both the individual and social dimensions related to them. One potential option to approach theoretically the presence of emotions in integrated care is the concept of emotional labor (Hochschild 1983; see also Zapf 2002; Zapf and Holz 2006). The concept was originally introduced by Arlie Hochschild in 1983 and has since also been applied in the field of health and social care (Mann 2005), especially in nursing (e.g., Gray 2009; Hunter and Smith 2007). Hochschild's 'sociology of emotions' reveals the taken-for-granted nature of interaction between people. It pays attention to what people feel, how they make sense of their feelings and how people have to regulate and manage their feelings not only in their individual lives (emotion work) but also at work (emotional labor) (Hochschild 1983; Garey and Hansen 2011).

In the emotional labor approach (Hochschild 1983), the client-employee relationship is emphasized, which matches well with the nature of person-centered integrated care. In this paper we do not focus so much on the ways care professionals display the appropriate feelings in given (face-to-face) situations with clients. Instead, the focus here is on the assumption that feelings related to the multiple complex problems of clients affect professionals' actions in any case—whether they are aware of them or not—in all their doings, not only in interaction with clients but also with other professionals. In emotional labor research, the original concepts of surface acting (faking emotions) and deep acting (trying to feel actual emotions required in the situation) have been complemented by a third form of emotional labor, natural and genuine emotional labor (Humphrey et al. 2015), which we think suits well in the context of care. A fruitful concept matching with care for people with multiple complex problems is emotional dissonance. Further, so-called feeling rules (display rules), i.e., organizational norms of feeling (Hochschild 1983; see also Diefendorff et al. 2011; Humphrey et al. 2015; Grandey and Melloy 2017) link the approach interestingly to cross-boundary collaboration in the context of shared clients.

This reflective paper is based on a Finnish study on integrated care. The aim here is to describe and illustrate the emotional and other relational dimensions of integrated care in light of the experiences of care professionals in the context of care for people with multiple complex problems. We hope that the insights presented here will raise further discussion on whether it is worth addressing 'the affective turn' (Burkitt 2014; Greco and Stenner 2008) and emotional dimensions more profoundly and in more detail in future research on integrated care.

### **2. Materials and Methods**

The empirical material reflected here comes from a Finnish research project, Onnistu sote-integraatiossa/ Successful Integration of Health and Social Care (2016–2018), funded by the Finnish Foundation for Municipal Development (KAKS). The study was conducted in close collaboration with Parempi Arki/Better Everyday Life (BEL) development project (2015–2017), funded by the Ministry of Social Affairs and Health (Finland). The aim of the BEL development project was to support clients with multiple problems in everyday life by developing a person-centered and integrated care approach including services in both health care and social care. The overall aim of the research project was to add to the understanding of the prerequisites for successful integration. The BEL project was a pilot project addressing implementation of integrated care, connected to Finland's national reform aiming at the complete integration of health and social care in 2021.

In total, 250 care professionals from primary health care, specialized health care, social care, and the education sector were involved in the BEL development project. These same professionals

were also the participants in the research project. The study participants were professionals working in the field of health and social care: nurses, doctors, social workers, therapists, teachers etc. The BEL project was an intervention project in which the professionals from different sectors (primary health care, specialized health care, social care and the education sector) and organizations (health centers, hospitals, social care organizations, schools) were joined together into 37 local cross-boundary development teams. The aim of these teams was to develop integrated care pathways for clients with multiple complex problems and multiple care needs–pathways linking care professionals from different organizations and different sectors to work together for the good of the people with multiple problems. Each team developed care for a chosen target group, for example for older persons regularly needing health services or for families with children who have special problems. The work of the BEL teams was supported by project coaches and through seminars (applying the Breakthrough method and other Leanmethods, see Bhat et al. 2014). Theoretically the development work drew on the extended Chronic Care Model (Barr et al. 2003).

The close collaboration between the research project and the development project enabled access to all materials produced by the BEL development project. The researcher (first author) was present in the seminars and workshops of the BEL project and the project manager of it (second author) also worked as a researcher, taking part in some of the interviews and in the data analysis. Together with the professionals involved in the development project, these arrangements enabled a fruitful dialogue between research and practice.

This reflective paper draws on the overall material and findings of both projects, in particular on the following data: (1) Lean-based Fishbone problem analysis (see Bhat et al. 2014) regarding clients with multiple complex problems, done by 21 BEL project teams (about 100 professionals). In these fishbone analyses the professionals analyzed the critical points in providing care for people with multiple complex problems. The outputs of the teams' analyses were transcribed and Atlas.ti software was used for preliminary thematic analysis of the fishbone problem analysis; (2) Research interviews with professionals and managers. Three cross-boundary BEL project teams were interviewed. Altogether, 14 care professionals (nurses, social workers, physiotherapists and doctors) from primary health care, social care and specialized health care participated these group interviews. In addition, individual interviews with nine managers connected to the BEL project, representing primary and specialized health care and social care were interviewed. Interviews were conducted by the authors of this paper (the researcher and the BEL project manager).

The interview themes concerned the professionals' experiences with clients with multiple complex problems, the challenges in providing and developing care for them, experiences of cross-boundary collaboration related to these clients, expectations of the collaboration outcomes, and managemen<sup>t</sup> issues related to cross-boundary collaboration. The research material was analyzed by means of inductive content analysis and the original findings are presented in the research report (Hujala and Lammintakanen 2018). For the purposes of this reflective paper, the findings of the original analyses are used selectively to highlight the emotional and other relational aspects of the research material. The themes described below (1) emotional burden; (2) professionals' fear of emotions and (3) 'emotional territories' are based on the holistic reflection of the whole research material and illustrated here by purposefully selected quotations from the data.
