Introduction

This paper builds on extant perspectives and approaches to understanding clinical directorates (CDs) in acute settings. CDs are managerial devices by which to coordinate clinical and organisational aspects of care. They are part of a broad reform agenda to improve the management of, and in, acute settings. To date, CDs have been assessed mainly through health services theoretical lenses and as such have been understood largely within the paradigm from which they have emerged rather than from the perspectives of social or organisational theory. In going beyond this paradigm, a more theoretically rich conceptualisation of CDs is realisable.

Clinical Directorates as Conceived in Health Services Literatures

After decades of organising hospitals along functional lines, policy-makers and managers have re-configured traditional designs. This appears to be an attempt to emphasise the products or services hospitals produce rather than the inputs used in their production. Until relatively recently the prescribed framework of hospitals was functionally isomorphic, with a separation by profession into four main divisions (medicine, nursing, allied health and everyone else).

Now, wards, clinical units and departments that exhibit characteristics which suggest some form of synchronous fit have been grouped together [23, 30, 32, 39]. Two broad types of arrangement can be identified, although there is considerable local variation. One is the divisional design, in which the organising principle is based on existing medical domains (doctors, nurses, allied health and support staff in internal medicine are grouped together organisationally, as are the respective staff in surgical, mental health and paediatric services). The other is the institute design, in which the organising principle is predicated on illness or body systems (staff are grouped along the principal service lines, e.g. cancer services, cardiothoracic services and neuro-services). The essential purpose is to organise hospitals into a coherent sub-structural arrangement through which to manage care processes.

Typically, in both of these kinds of arrangement formal leadership is provided through a tripartite management grouping consisting of a part-time medical clinician (often titled the clinical director), full-time nursing manager (frequently construed as the operational manager) and business manager (in many cases taking responsibility as the financial and information manager). Sometimes, the medical and nursing leaders are co-directors. Depending upon circumstances, and how much control is exercised by the more senior executives or policymaker ranks, this management group is said to be responsible to a greater or lesser degree for the administrative and financial performance of the organisational sub-entity. It is also deemed responsible for aspects of clinical processes, outputs and outcomes, although it is not always clear to what extent. While these new designs have been given many titles [10] they are most often referred to as clinical directorates (CDs).

The origin of the CD form of hospital design dates from the mid-1970s at the Johns Hopkins Hospital in Baltimore, United States of America [38, 39]. At Johns Hopkins each CD was managed by a medically qualified chief with a nurse manager and business or administrative manager in support. This management grouping was allocated a budget, and the CD essentially became a small-scale, clinically focused acute care entity with defined boundaries based on product line management principles, yet it was integrated within the larger organisational configuration.

This idea was imported into Guy’s Hospital in Britain’s National Health System (NHS) in 1984 [17] and into Australia at St. Vincent’s Hospital in Sydney in 1989 [8]. Once it took hold the idea radiated rapidly. There has been strong support for it among managers and clinicians in large hospitals on both sides of the Atlantic [48, 64] and in Australia [51] as well as the Netherlands [54], Denmark [21, 53], Switzerland [31], Sweden [7] and Canada [30]. The trend appears to be widespread in Western countries. Throughout the 1980s, 1990s and 2000s CDs emerged, seemingly in all, certainly in most large, acute settings.

This paper’s task is to consider some of the effects of this shift from one to another way of organising activities, and in particular, how social processes are expressed in terms of the structural and cultural changes that CDs ostensibly enable. Conceptually, the focus of much of this kind of research has in health services research circles been on the redesign of the organisational chart, what will be referred to as the prescribed organisational framework. As will be seen, this notion of organisational structure is incomplete [57]. The paper thus will pay attention to a second aspect of structure, which will be discussed under the label structuring. This is the social structure of the teaching hospital, the interwoven, patterned interaction of enduring organisational relationships and activities. But first, in order to provide context and to set up the subsequent analysis, we will reflect further on what the CD literature to date has said.

Claimed Effects of the Shift to Clinical Directorates

There are two major streams of extant CD literature [12]. This paper seeks to contribute a third and fourth literature set, mainly by drawing on theoretical perspectives from outside of health services domains. The first stream, and by far the largest by volume, is the literature promotional of CDs. The second is the literature critical of CDs. The third deconstructs the first two streams, having recourse to a distinction between prescriptive and descriptive accounts. The fourth offers a critique of CDs, particularly using Giddens’ and Weick’s theoretical work, as well as Martin’s.

Most commentators in the health services literature have tended to be anecdotal, as opposed to empirical, and generally positive about CDs and the effects they are purported to have had. When they have been empirical, case studies are the preferred research method. Claimed benefits of CDs have centred on greater efficiencies [9, 58], better, more localised managerial decisions [39, 40, 59], a sharper patient focus [45], more targetted resource allocation [33, 46, 52] and more generally, that CDs represent a major step forward in managing hospitals [23, 44, 51].

The amount of criticism is dwarfed by the volume of supportive literature such as this. Where criticism has been levelled at CDs, it is typically of two kinds. The first is that CDs represent a managerialist attempt to subordinate clinicians by incorporating them into management hierarchies at the expense of their own interests and at the risk of supplanting professional by managerial values [29, 37]. The second round of criticism is in contradistinction to the first: that clinicians have used or even stimulated the emergence of CDs to consolidate their place in the health system by “colonising managers’ agenda and effectively neutering any serious challenge to their position” [42]. On this account the establishment of CDs signals that it is the doctors and nurses who occupy CD managerial and concomitant positions of influence who are triumphant, not those health sector policymakers and managers who might seek to overshadow them or curtail their power.

A comparison of three studies, one each from Britain, the United States and Australia [9, 23, 58], suggests that there are common response patterns when participants within reorganised hospitals are asked for the reasons for embracing clinical management designs. These are summarised in Table 1.

Table 1 Reasons for moving to a CD design reported in three studies
Full size table

According to these studies, the reasons for engaging clinicians within management processes and hierarchies via CDs are fourfold. They are improved technical efficiency, i.e. heightened financial focus and tighter budgetary control; enhanced managerial performance and involvement of clinicians in management leading to better organisational focus on patients and services; improved information technology (IT) systems and management; and enhanced quality of care and patient outcomes.

Whether these claimed benefits have eventuated, or the scale to which they have, is not apparent from an inspection of most literature. It is also unclear whether there are satisfactory theoretical grounds, or good evidence, that the establishment of CDs has eroded or will erode professionalism over time in favour of managerialism.

Prescriptive and Descriptive Literature

These are empirical questions, and largely are yet to be answered. As alluded to, one theoretical problem is that in dealing with CDs from within their own paradigm, the possibility of evaluating CDs from traditions outside health services or clinical perspectives has been missed. It is to this task that we now turn.

One line of critique is to deconstruct the limitations of the CD literature by exposing its prescriptive nature. A major proportion of general management and health care management literature, including much of the literature about the CD presented above, is prescriptive rather than descriptive. It assumes the necessity, legitimacy and rationality of management. Such literature is managerialist, instrumental, hortatory and normative. By managerialist is meant that it takes the perspective of the organisation’s ruling hegemony, chiefly represented by senior managers at the corporate level and board members including, in private organisations, shareholders. Accounts of this type assert management’s primacy and advocate managerial solutions to problems. The interests and concerns of people in management positions have ascendancy over those of others. This is a view of the world that proceeds mainly from one standpoint—management’s—and tenders management-oriented responses to matters at issue. Much of the CD literature cited above reflects this managerial perspective.

Instrumental approaches view the organisation as a tool for alleged societal purposes. Such approaches assume that organisations exist for functional reasons. Hospitals, schools and banks are instruments under the control of managers for the provision of vital community needs—in these cases, acute care, education and financial services respectively. According to this perspective it is the managerial coalition’s task, inter alia, to determine strategy, organisational design and responsibilities, to allocate resources, reward and punish, hold people to account and prescribe policy in order that organisations fulfil their charter. Managers do so by virtue of their knowledge, skill and pre-eminent position in the hierarchy. This entitles them to shore up and extend their position, privileges and prerogatives via managerial strategies. Most CD literature proceeds from instrumental assumptions.

The hortatory concept draws another distinction. Much management literature espouses what should be the case rather than what is the case. It prescribes a view of what managers ought to do to be successful rather than describes and analyses existing states-of-affair. A considerable amount of the CD literature is hortatory, pronouncing on issues such as how to be a better clinician manager [18], and portraying how a well organised CD would be run, and what sort of characteristics it should exhibit [56].

The word normative here is used in two senses: in terms of prescribing an ideal and in propounding the expectations of managers by favouring the establishment of a norm. Normative literature postulates an idealised world and, for example, formulates or assumes standards and policies. Normative literature implies that there is an appropriate way of thinking about the world. Within conventional literature it is evident when the organisational configuration, encompassing hierarchy and managerial dominance, is presented as unquestionably good. It suggests that management’s task is to devise and enforce adherence to these norms. Management thus projects expectations about how organisational agents should act. Departure from these is seen as an unwarranted distortion. Most writers in the CD literature speak from a normative position: the assumption is that clinical governance designs of this type are an improvement, and they are part of a hierarchically desirable, and correct, even obvious way of working. There is an unspoken expectation that organisational agents should and will uncritically accept this idea for rearranging the hierarchy of large hospitals.

Consideration of managers’ (especially senior managers’) language and behaviour suggests that many proceed from managerialist, instrumental, espoused and normative assumptions (for further critiques in this vein, see, for example [1, 3, 47, 62]). The interests of managers are served if the organisational discourse is characterised thus, and their assumptions are unquestioned. It is also in the interests of managers that the managerial literature is couched in this way. They are the main consumers of this literature and it is to their advantage if they can exercise control and influence over other members of the organisation. They can do so more securely if the literature to which they have recourse and if managerial training and development programs are written in terms of—and take a perspective that—bolsters their position and serves their interests.

Organisational Social Structure

Thus a criticism of the received wisdom on CDs to date is that it is prescriptive rather than descriptive. We can extend the points of critique about CDs qua CDs if we consider them through the lens of Anthony Giddens [34–36], a sociologist of social structure, and Karl Weick [67–69], a social psychologist of organising. One of Giddens’ long term projects has been to reconcile, and connect, structure with agency through analysis of the macro and micro of social settings. His structuration theory engages with social practices across time, and argues that social structure is emergent, and produced, reproduced and changed via practice. Social structures are “constituted by human agency” (i.e. action), and “action is constituted structurally” [34, p. 161]. Social behaviours are underpinned by rules, or generalisable procedures that guide behaviours, and resources, or material and non-material capacities. Social structures in health sector organisational terms are the ongoing, recurring patterns of behaviours exhibited by clinicians, managers and patients. Extrapolating this view, CDs are attempts to alter social practices, actions and behaviours for improved service provision.

Weick’s long run mission involves considering organisations and institutions as processes: organising not organisation, structuring not structure and managing not management. He sees institutional behaviours as non-linear, evolving, ambiguous, often chaotic. The world, and organisational life, is essentially confusing and unpredictable. There are few if any simple answers, there is always noise and variability in the system, and organisations are evolving unpredictably all the time. In order to understand what is going on in organisations people ‘sensemake’—i.e. use rolling hindsight and retrospective understanding to try to account as best they can for all the stimuli they constantly encounter. Weick would no doubt predict that hospitals can be represented as complex institutional processes, and whether before or after the introduction of CDs, these processes would be disjointed, hard to read, and full of blind spots, wrong turns and roundabouts of uncertainty.

Taken together, Giddens–Weick can be apprehended with reference to several conceptual contingencies by which to picture organisations. These include that organisations, although hard to pin down, are compositions of social structures and sub-structures; they manifest in patterned interactiveness; and they both enable and constrain behaviour. Social behaviours in organisations can be best understood as processes that inhere with uncertainty, unpredictability and continuous disorganisation. Action and structure are mutually constituted. We may not be able completely to square Weick with Giddens, but they have both called to attention the emergent, constraining and sometimes enabling aspects of organisations, and have brought to centre stage the place of human actors’ recursive behaviours in organisations and their ongoing interpretations and constructions of the social phenomena they encounter.

Using the earlier health services literature to understand CDs as a structuring device, we can go about the task of distinguishing, following Giddens–Weick and other contemporary social theorists’ positionings, between two overarching notions of organisational structure [14, 24–27, 34, 36, 66, 67]. This distinction draws on a philosophical position about the ontology of structure from Searle [61]; (see also [14, 41]). In this paper, we are concerned to express this not so much in terms of the dual nature of structure, but to make the point that the views people hold of them are dichotomisable. The first of these, what we might call the type A structural account, takes it that structure is concrete and tangible, and leads to the treatment of the organisational framework as a noun, as identifiable entity: the structure. Structure here is that which is prescribed, usually in terms of formal variables. Structure is thought of as the projection of a framework by senior managers, and is accompanied by an insistence on rules and reporting arrangements. Key concepts are hierarchy and formalised roles. So conceived, the organisation’s form is often thought to be designed intentionally by senior managers so as to guide (i.e. facilitate or impede) certain kinds of behaviour, although there is a longstanding debate about the role of strategic choice. This field is usually considered the orientation of the structural contingency theorists [20, 28, 55]. Their approach is to examine how organisations and their environments relate. Donaldson in particular has argued these must be in fit, otherwise there will need to be a structural accommodation via contingency factors such as technology, size, strategy or innovation. When in misfit, organisations adapt over time by changing structurally. CDs might emerge for example from a misfit between an environmental factor or factors, and are a structural accommodation to it.

The second view, labelled the type B structural account, relies on much of Giddens–Weick, and takes it that structure is that which is recursively enacted by organisational players as they create and recreate their social setting [6, 35, 67]. Structure is social structure, fashioned through a continuous flow of organisational conduct. People initiate and perpetuate social forces which bind them [4]. Social structure is centrally about the taken-for-granted, continually evolving and routinised behaviours and meanings of social actors. Structuring on this account is not a noun but a verb, a process. This is the orientation of the social action domain [19, 22, 36, 63, 66]. This latter concept is complex, and not well delineated by prescribed organisational charts, i.e. two-dimensional depictions with lines linking boxes purporting to represent responsibility, accountability and other clarifying relationships of power.

Relating Social Structure to Culture

Views about the patterns of developmental interaction—the on-the-ground, interlocking behaviours of workplace participants—provide a gateway to an account of organisational culture (see, for example [2, 5, 49]). Observing the behaviour and listening to the talk of organisational participants provides information about ‘the way things are done around here.’ This is classically what organisational culture is thought to be: the collectively held and transmitted history, meanings, customs, symbols, artifacts, language, values and behaviours [15]. Schein [60], taking a perspective of group culture, for example, defines it as “a pattern of shared basic assumptions that was learned by a group as it solved its problems of external adaptation and internal integration, that has worked well enough to be considered valid, and, therefore, to be taught to new members as the correct way to perceive, think and feel in relation to those problems.”

What are the conceptual relationships between culture construed this way, and structure? For some scholars there may be theoretically technical reasons for supposing that organisational social structure and culture are separate and distinct but there seems no ready advantage conceptually in maintaining cleavage. Investigate social structure and you seem to be, pari passu, examining organisational cultural characteristics. Alvesson [2] considered the distinction between culture and social structure. For him, culture is centrally about meanings and symbols, while social structure is primarily concerned with behavioural patterns and interactive flows. Both, for Alvesson, “represent different abstractions of the same phenomenon.”

Given the present paper’s concern with hospital structural improvement via CDs, at issue is what is meant by social structural or cultural change? To answer this question raises another, about what view of structure is envisaged. In the first, type A structural case, change occurs when, for example, there are attempts to alter formal relationships (who now reports to whom?) which are then encoded in re-arranged boxes-on-the-chart. This is one manifestation of what CDs are, within the logic of the type A perspective. In the latter, type B case, change becomes apparent when an observable shift occurs in what social psychologists would call the patterns of group behaviour and sociologists would refer to as action.

It is tautological to state that the emergence of CDs is associated with an alteration to the prescribed framework. New organisational charts have appeared in hospitals everywhere, reflecting that CDs are now the privileged structural arrangement. What is unclear is whether the appearance of CDs has changed the social structure of hospitals, and with that the way people act and relate culturally. In short, have new forms of social structuring occurred as a result of CDs? On the broader front, has the establishment of CDs affected the organisational cultural landscape of acute settings? If this has changed, in what way and to what degree?

In truth we cannot provide a definitive response, but we can speculate. Martin’s map of culture in three modes—integration, differentiation and fragmentation—provides one key to conceptualising social structure, culture and change ([49]; see also [15]). The integration perspective is often normative, and looks for what is held in common throughout an organisation. The organisation is some kind of coherent, consistent entity, so observers can think of an entire hospital or NHS Trust, for instance, as exhibiting certain broadly defining features.

Look at an organisation through the differentiation perspective and sub-units come into focus. Hospitals are replete with all sorts of formal and informal groups, professionalised networks, friendship and convenience cliques, departments, units, teams and other demographic sub-sets. Viewed this way, the hospital is now some sort of loosely-coupled, composite arrangement comprising various sub-cultures, silos and meso-clusters, situated above individual wards, units and departments and below the whole-hospital-as-an-entity. CDs are an organisationally-sanctioned and prescribed, sub-cultural, meso-clustered arrangement.

Now attend to the organisation through the fragmentation perspective, and search for micro ambiguities, tensions and contradictions. Organisations are in constant upheaval with shifting coalitions, politicking, in-fighting and the ebb and flow of constantly forming and re-forming diverse groups, cliques and clans. Within CDs, according to this vantage point, we will see the shifting sands of fine-grained, conflictual, human interaction.

CDs could be construed, observed and analysed under each of these theoretical headings. From an integration standpoint, CDs might be thought of as normative solutions to attempt to induce professional leadership and interconnectedness across a large acute setting, i.e. the whole hospital. Move to a differentiation position and each CD could be expected to be a relatively independent sub-structure, be distinguishable from other CDs, and perhaps compete for resources as against other CDs. Shift once more to a fragmentation perspective, and the micro behaviours within CDs will appear as untidy, political and muddled, sometimes disorganised, always iterative and often chaotic.

Type A and Type B Social Structures

Which brings us back to Giddens–Weick and the alternate views of social structure. In the first, type A rendering, CDs have emerged, the new framework has been settled on, and organisational charts have been re-drawn to reflect the CD designs. Few acute settings in developed health systems have not gone through these changes. But if they are type A structural views, what can we make of them in the light of the type B, Giddens–Weick social structural perspective, and of Martin’s view of culture, as collectively representing a different gateway to understanding the CD phenomenon? These social and organisational theorists speak from outside of the heath literature, yet can enable insights into CDs, despite them not having specifically considered CDs within hospitals.

What are some of these insights? A priori it might be predicted that CDs are set up with integration in mind by the authorities who establish them. They will be construed as some sort of concrete structural solution in order to emphasise, inter alia, efficiency, multidisciplinary teams and cooperative service delivery. They are intended to change behaviour and practices, to improve care and focus attention on specific aspects of acute service delivery [12, 13]. However, in describing them this way, they are a boxes-on-the-chart, type A solution.

Yet whether there is any relationship between such formally prescribed structures and the actual practices of clinical professionals is extremely doubtful on several bases. First, a range of empirical observations show that organisations do not function according to formal prescriptions, and behaviour on-the-ground, including managerial behaviour, is much more inchoate, more emergent and less planned [16, 43, 50, 65] than it seems from the coherency suggested in formal structural accounts. Second, there is evidence that CDs have been established as formal entities and prescribed as new boxes-on-the-chart, yet they have not affected social practices to the extent envisaged in the formal documentation [14]. The organisational chart says one thing but the behaviours and practices manifest differently. Third, CDs are attempts at imposing type A order on practices and behaviour which are, for a Weickian analyst, actually type B phenomena: often unfathomably complex, iterative, dynamic and disjointed, and dependent on retrospective sensemaking for understanding. It is unremittingly hard to see how drawing up new boxes-on-the-chart will change this. Fourth, CDs, to use Martin’s cultural classification, may be construed as an integrationist or differentiationist tool by those who establish them, but they are also fragmentationist. No-one amongst the cohort of supporters of CDs in the health services literature has made a convincing case for how the fragmentationist behaviours will be influenced by the attempts at creating integration or differentiation, i.e. by establishing CDs. Similarly, fifth, those who have established CDs have not mounted a plausible argument (indeed, any case much at all) as to the circumstances in which, in Giddens’ terms, new rules are stipulated and different resources are mobilised in support of type A CD structural aims. Sixth and finally, it does not appear to be the case that those who have advanced or ushered in CDs have factored into their thinking a deeply philosophical-sociological apprehension of hospital organisation. Hence, it is not likely that the intrinsic features of hospital organisation, understood through a frame like that which can be mobilised via Giddens–Weick, or type B accounts, were taken into consideration by those designing and instituting the new CD order of things.

The paradox is that if you realised and factored in Giddens, Weick, Martin and type B structural features, you might come to believe that formally reorganising into CDs is not the solution needed to create the changes you believe are desirable. So all in all, restructuring into CDs does not appear to be able to realise the changes required by the proponents of them: plus ça change, plus c’est la même chose.

Conclusion

It is doubtful on theoretical arguments whether CDs could change on-the-ground behaviour as extensively as hoped by some of their advocates, or in the direction required according to the admonitions of those who support them as mechanisms for change. Both Giddens’ structuration and Weick’s organising-sensemaking draw attention to a weakness in these propositions: Giddens, because restructuring along CD lines is not guaranteed to affect agency or the rules and resources that govern social situations; and Weick, because hospitals restructured into CDs are likely not to be any less ambiguous, imprecise or chaotic than they were previously, despite the rationale for CDs in the first place as vehicles for making acute settings more coherent and organised. In the end, these are not only theoretical but also empirical questions; and the companion paper to this one [11] attempts a response by considering some ethnographic data to shed further light on this question.