Patient-centred care has become an important aspect in health care delivery over the past years. It is seen as a core aspect of improving health care quality [1, 2]. The Institute of Medicine defines patient centredness as ‘health care that establishes a partnership among practitioners, patients, and their families (…) to ensure that decisions respect patients’ needs and preferences, and that patients have the education and support they need to make decisions and participate in their own care’ [3]. Nowadays, health care providers increasingly seek to improve patient-centred care in order to achieve better medical outcomes [4]. However, this goal cannot be achieved by single organisations solely. The integration of services and collaboration of health professionals and organisations is needed for improving health care quality and safety [5]. A primary reason why organisations in the health and social care sector should cooperate and implement integrated care concepts is the high fragmentation of support services and a strict separation of supply and care sectors. Health care providers are specialised in their tasks and offers so that patient-centred care and a comprehensive supply of care for people in need can only be established by collaboration and building integrated care partnerships [6, 7].

Organisations in the health care sector also have to manage costs for medical treatments. The challenge is to control costs, on the one hand, while enhancing the quality of health and medical care, on the other hand [8, 9]. Especially hospitals are affected by economic pressure due to a new remuneration system called Diagnosis-Related Groups (DRGs), which was invented in the USA and is implemented in many (European) countries [10, 11]. Since 2004, the costs for medical treatments in German acute hospitals are calculated according to this system. The former financing system covered the costs of how long a patient stayed in hospital, thus giving fewer incentives to discharge a patient as soon as possible. Now medical supply has to be achieved with limited financial resources due to the economic requirements from the DRG system. Hospitals have to stay within an average residence time for patients and leave further supply to other health care providers. Increasing costs and decreasing hospital-staying time require network-like co-operations. Thus, the DRG system fosters integrated care partnerships [12–14].

The challenge is to arrange and coordinate patient-centred care in a setting where many organisations are involved. Decisions on medical treatments, therapy options, in- or outpatient treatment, etc., are not made by a single central authority. Additionally, economic aspects may have an impact on these decisions.

Hence, the present paper addresses two questions: (1) How do integrated care networks effectively carry out control on collaborating health professionals and ensure patient-centred care under conditions of financial limitations due to the DRG system; (2) Which problems do collaborative networks solve with regard to patient centredness in health care delivery?

From a classical economic point of view [15], organisations and their structural characteristics can be located on a continuum between hierarchy and market. In the past, there was a ‘professional dominance’ (hierarchy) [16] in medical care where doctors decided on the completion of patients’ medical treatments and their hospital discharge. Handling the balance between medical and economical requirements was in the doctors’ authority [17–19]. This has changed dramatically with the introduction of the DRGs not only in Germany but also in many European countries [11], which led to a shift on the continuum from hierarchy towards market. The DRG system (market) ‘controls’ the discharge and transition planning in integrated care partnerships.

At first glance, it seems that nowadays market rules replaced hierarchy as predominating control mechanism in integrated care. However, having a closer look at organisational behaviour in integrated care partnerships, the hierarchy-market-continuum no longer seems to be suitable to describe the collaboration process. Organisations behave in a way that can be explained neither by referring to authority (hierarchy) nor by the logic of markets [20]. Rather, organisations are located in cooperation structures that rely much more on reciprocity, mutual interests and appraisal. Organisational behaviour cannot be sufficiently explained by either formal structures of hierarchy or profit-orientation. Powell suggests calling this control mechanism networks [20]. In this understanding, organisations appear as autonomous actors that cannot be controlled from outside [21]. This view is emphasised by systems theory research. Organisations, in general – and this applies to organisations within the health care and nursing sector as well, especially to hospitals, – can be seen as complex social systems, which have their own logic and dynamics [22, 23]. Systemic perspectives are helpful to get a better understanding of how collaborative networks ‘work’. The systems theoretical approach allows a comparison of networks and functional equivalent controlling mechanisms like hierarchical or market structures in order to underline advantages and disadvantages of different controlling mechanisms.

Unlike the Complex Adaptive Systems Theory (CAS) that considers systems as ‘open’ [7], in this paper, the systems theory by Niklas Luhmann [24] is used as theoretical framework to analyse the cooperation of organisations, supplemented by aspects of Powell's [16] and White's [25] network theory approaches. Luhmann adopted the concepts of self-organisation and self-reference of systems [26, 27] which means that systems are operationally closed. This perspective was emphasised later when Luhmann also integrated the autopoiesis concept, invented by the biologists Maturana and Varela [28]. Autopoiesis explains how a cell can reproduce itself without direct input from its environment, since the cell itself is separated from the environment via its membranes. Luhmann applied the concept of operational closure on social systems, which reproduce themselves by the basic operation of communication.

The concept of operational closure explains why control attempts (for instance, implementing the DRG system to control quality and costs in integrated care) do not automatically change an organisation's behaviour according to the primary goals of health politics. This problem becomes even more virulent in the context of integrated care: Decisions and routines made in one organisation are not automatically applied to or accepted by other organisations. There is no central instance or authority for controlling (steering) the integrated care and cooperation process. Different organisational structures have to be taken into account. In collaborative networks, each party has certain expectations of what the other network partners can or should do, and what can be expected from oneself.

In this sense, from a systems theoretical view, networks can be considered as attempts to develop, form and control identities and enforce own interests [25]. Organisational behaviour, as an attempt to form and control its identity, is influenced and ‘limited’ by other organisations’ attempts of controlling network-processes and vice versa. The network relations are self-conditioned by mutual observations. From this point of view, organisations no longer appear as purpose-rational institutions that are defined by their members, which pursue own interests. Rather, organisations are ‘purpose-seeking systems’ [29, 30] that evolve independent from its members’ actions and motives. The principle of self-organisation is a central point in this definition of ‘networks’.

The research interest of this paper was to investigate the control mechanism of collaborative networks in the context of integrated care partnerships and how this mechanism affects the challenge of high quality in health care delivery under limited financial resources. To pursue this question, a qualitative approach was chosen that facilitates access to the implicit knowledge and the social practise of health professionals in collaborative networks [31]. The data that will be analysed for this paper stem from qualitative, semi-structured interviews. An interview guide covering the topics of integrated care, patient centredness, barriers and facilitators of collaboration as well as economic aspects was developed and pre-tested. Seventeen persons in different acute hospitals of metropolitan areas (Hamburg, Bremen and Osnabrück) have been interviewed. Twelve of the seventeen hospitals were privately operated. Five were NPO. Due to the logistic effort of personal face-to-face interviews, the sample region was restricted to northern Germany. The sample size was chosen according the criteria of saturation. While about fifteen interviews are generally recommended for this kind of qualitative studies [32], after seventeen conducted interviews, no further insights into the issue under investigation were revealed and hence the stopping criterion for sample saturation was reached. The questioned persons are either case managers or social workers and work in the field of hospital discharge, i.e. their professional context is embedded into integrated care partnerships and interface management. None of these persons belonged to the nursing staff. The duration of interviews varied between 45 and 60 minutes. They were digitally recorded and transcribed with respect for keeping mentioned persons anonymous. Parts of the interviews were evaluated and interpreted collaboratively in data analyses workshops with colleagues.

According to Vogd [33] and Bohnsack et al. [34], the documentary method as a qualitative reconstructive research approach seems suitable to analyse organisational behaviour in collaborate networks. Following this method, the researcher tries to reconstruct the modus operandi of (organisational) practice. The question is, how social formations emerge, which means, the focus lies on the process of ‘sense–making’ in organisations [35, 36]. Three analytic steps are involved: the formulating interpretation, the reflecting interpretation and the constant comparative analysis.

In the first step of formulating interpretation, the researcher chooses specific passages of a text for further in-depth analysis. These passages are identified according to thematic relevance, where a possible interview-guideline can be used as rough framework to retrieve the formal structure of an interview. During this analytic step, the researcher tries to reformulate these passages in order to gain an overview of relevant topics for the following interpretation. The next two steps, reflecting interpretation and constant comparative analysis, are performed simultaneously in practise. In the reflecting interpretation, the aim is to identify different ways of how a specific problem is handled. By contrasting passages in different interviews or even different passages within the same interview that deal with certain problems, latent pattern of organisational structures are identified. This analysis step requires a constant comparative analysis. The analysed passages serve as comparing perspectives that identify the ‘contrast of similarities’ [37] and allow for methodological control, as the implicit knowledge of the actors becomes more and more explicit. ‘Consequently the case comparison is to ensure the inter-subjectification of the results’ and reduces ‘the risk that the interpreters are trapped in their own expectations of normality and thus approach the data with cultural norms that do not do them justice’ [38]. In case of sequential analytical approaches like that suggested by the documentary method, computer-aided data analysis (i.e. software like MaxQDA) does not provide much benefits over a ‘pen-and–paper’ procedure [39]. Thus, the data analysis was carried out by directly working with the text documents.

The following data analysis focuses on the structures of integrated care partnerships. It emphasises how the involved health care providers put patient centredness and quality of care into practice.

The core question of this paper was how professional health care providers deal with the problem of high quality in health care delivery under conditions of economic limitations (costs) and in which way collaborative networks (integrated care) can contribute to solve this problem.

From the empirical findings, we can derive that economic limitations especially affect the hospitals according to their discharge planning. There is no option getting more payments if a patient stays longer in hospitals. Increasing costs and decreasing hospital-staying time require network-like cooperations. These findings are emphasised by other studies that show how this remuneration system fosters the necessity of cooperation and coordination in health care delivery [12, 13]. However, looking at the data, it does not seem that the pure logic of market is a suitable instrument to control the integrated care process. The same seems true for hierarchic structures with clearly defined rules that determine the tasks and workflow in cooperation networks. All questioned health professionals have similar ideas of characteristics of integrated care partnerships. They point out the informal character of the relationship of network partners. They all are against consortiums and contractual relations. Networks, in their understanding, consist of ‘loosely coupled’ elements. This correlates with a systems theoretical conception of organisations, which are assumed autonomous in their decision-making due to operational closure [23]. Self-regulation (or self-organisation) of organisations (autonomous decision-making and operational closure) needs loosely coupled networks. First, this ensures that no conjoint binding rules, which can be considered as control attempts from the outside, work against a system's autopoiesis (survival). Second, lose couplings ensure high flexibility in networks [40].

These findings are in line with other studies that relied on systems theoretical frameworks. ‘Change cannot be forced and attempts to control the system or prescribe innovation are often counterproductive due to the potential to destabilize the system’ [7]. Rather, integrated care partnerships should rely on structures that ensure self-organisation. This kind of control mechanism is what Powell defines as ‘networks’ [20]. In this spirit, following the idea of Åkerstrøm Andersen [41], networks can be considered as ‘reflexive contract’. Reflexive awareness, mutual control and professional autonomy are the basis for sustainable network structures and keep these structures stable. Profit-seeking consortiums do not have the necessary reflexive awareness and adaptability. They may have a negative impact on patient centredness because inappropriate conditions and mal-supply cannot be eliminated by self-regulation. In this sense, collaborative networks can solve the problem of missing self-organisation. Therefore, integrated care partnerships should be based on autonomy (from the systems or organisations perspective) in the context of reliability (from the networks or environmental perspective).

Patient-centred care can be achieved in such kinds of integrated care partnerships. The possibility of alternatives according to the choice of network partners seems to be a tool for regulating mal-conditions. If two health providers cannot find a common basis for their relationship because they may completely differ in their understanding of patient-centred care, other network partners will be chosen. To remain attractive as network partner and benefit from health care delivery (in terms of making money), health providers have to behave cooperatively in a way that each party in the network has to stick to the ‘reflexive contract’. The interviewed health professionals called this a ‘give-and-take basis’ for their relationship. Networks appear as reflection strategy according to the stability and durability of integrated care networks – also taking into account the point of staying in a network. From a network theoretical perspective, the health professionals’ behaviour can be described as interplay of ‘identity and control’ [25]. They enforce their claims towards other network partners, negotiate tasks in the context of hospital discharge planning and fathom advantages and disadvantages (identity). They reflect reactions from other network partners and adjust their future behaviour in order to stay in the collaborative network (control).

This study uses a relatively small sample of case managers and social workers in the context of integrated care. It is difficult to say how far these findings are transferable to integrated care partnerships in general. Interviewees were mostly involved in informal collaborations. For deeper analyses of advantages and disadvantages of contract-based cooperation, other cases should be included in the sample. Moreover, the sample was taken from metropolitan regions of Germany, where many health care providers exist and compete. In rural areas, there are often less alternatives for choosing cooperation partners, so network structures may differ in such regions. Finally, only the perspective from the hospital staff was included in the data analysis. The recommendation for further research is to enhance the ‘network perspective’ by contrasting different views of all involved professional providers.

Nevertheless, even having a limited sample size – which is, however, typical for qualitative study designs [32] – the findings are based on a reconstructive research method. This type of method allows conclusions on specific social patterns even in case of smaller sample sizes, given that the steps of data analysis follow the rules of the applied methodological approach [34].

In conclusion, the recommendation from the perspective of systems and network theory as well as the empirical data is to be careful when planning integrated care based on formal contracts and hierarchical structures. Limiting the professional autonomy, limiting the competition and implementing inflexible co-operation structures may have negative impacts on patient-centred care and health care quality. Contractual bindings in networks that suppress the competitive environment may have negative consequences for patient-centred care. This finding may not apply in general to all formalised integrated care partnerships. However, there are empirical examples that show how hierarchical structures led to malpractices in patients’ medical treatment and its related financial charging. In the so-called ‘Helios affair’ in Berlin [42, 43], the competition of health providers was limited by regimented network structures that included just a few actors. This affected the quality of patient provision because missing reflexive awareness and self-regulation of the network could not eliminate this drawback.

Collaborative networks based on informal relationships solve this problem of missing ‘feedback system’. Evaluation criteria in terms of good patient centredness evolve from the reciprocity and self-organisation in such networks. Hence, reflexive awareness in networks becomes a quality indicator that promotes health care quality and patient centredness. The same applies to competition, which also improves the quality of care [44]. This is also because health care delivery embedded in formalised hierarchical structures provokes a behaviour, which mostly has a negative impact on the patients’ autonomy of decisions [45].

Integrated care partnerships may benefit from relying on networks considered as reflection strategy based on professional autonomy in the context of reliability. Contract-based collaboration concepts therefore should define quality standards and include methods to ensure feedback and evaluate the work of network partners.

In this sense, ‘the parties to a network agree to forego the right to pursue own interests at the expense of others’ [20]. This may lead to improved health care quality and better patient-centred care.