ENCC or ad hoc model

Hospital Managerial Accounting, a field to explore: ENCC or ad hoc model? Pierre Mévellec, Professor Emeritus, University of Nantes [email protected] Benoît Nautré, Director of the Hospital St Augustin [email protected]

Summary: The new unified financing system, based on the production measures, place hospital structures vis-à-vis an alternative: to implement the model suggested by supervision (ENCC) or to develop its own model. This choice is particularly crucial for not-for-profit establishments, their originality being any more taken into account by the law HPST. The Hospital Group St Augustin made the choice of an autonomous instrumentation. Choice based on a participatory approach associating the health professionals to the structuring of the economic model and the definition of the refunds intended for control. This text puts in perspective the development of an instrumentation and debate on the governance of the hospital structures and offer a track to bypass the traditional conflict between medical and administrative power by calling upon the concept of boundary-object. Keywords: healthcare costing, boundary-object, governance,

Cf. version française pour les annotations

Hospital Managerial accounting, a field to explore: ENCC or ad hoc model? Management tools, in the broad sense, are built on the basis of representations of the functioning of the organizations in which they are located. These representations themselves integrate with an overview of the functioning of the society context in which these organizations fit. The development of non-indigenous organizations from the dominant model in a society is problematic in the long term. Indeed, the management instrumentation that they tend to incorporate draws them imperceptibly towards the dominant model. It is therefore important to maintain and operate its originality, so that the not-for-profit hospital movement (Nautré, 2007) acquires a critical reading of the standard management tools and invests in the development of specific variants of the management instrumentation consistent with its vision. The introduction of cost calculation and its supports within the information system is seen here from the organizational learning perspective. If the cost concept has always been convened in discussions on the financing of the health care system (Moisdon and Tonneau, 1999), it has changed over the last thirty years. Theoretically, one can distinguish three phases: daily historical full cost, estimated global cost, and finally standard unit cost per pathology. This evolution in cost variants used for the management of the health system has given rise to two complementary and overlapping questions: is cost a good indicator of pilotage for a health care system? If so, is the current variant likely to be efficient for the management of health care organizations? The answer to these questions cannot be disconnected from the global environment of the health system, subjected to a succession of reforms whose effects cannot be limited to the financing of hospitals. Indeed, beyond moving from global allocations to funding on the basis of activity (T2A), reform requires closer liaison between medical units and administrative and nursing logic, accountability of actors through the dissemination of a medico–economic culture, and further, looking for appropriation by the actors of the piloting tools. In line with the principles of the New Public Management model, orders of April 1996, the Hospital Act of 2004, Hospital Plan 2007 and finally Hospital Patient Health Territory Act (HPST), all these reforms require 'corporate governance' as a means of driving performance.

We have structured our study as follows. The first part includes a brief presentation of the field and the methodology of intervention followed by an exploration of consistency between modes of control and instrumentation in cost calculation. The second part seeks to describe three different stages followed, in order to provide the development of a piloting open tool.

A: Health care organizations looking for new piloting tools. 1 Field study and research methodology The French hospital system includes two broad categories of institutions, private organizations and public organizations or organizations associated with the public sector. The first have been profoundly reformed over the last twenty years, to adapt their offer to economic and technical developments in medicine. In fact, their mode of financing, unchanged since their inception and only related to activity, places them in a highly competitive market. Concentrated movements have marked the end of a medical or family ownership for the benefit of specialized health groups, refocusing the decisionmaking and management tools into the hands of senior management. The latter are either governed by public law, and in this case governed by relatively cumbersome and inertial administrative logic, or belong to not-for-profit organizations (mutual entities, religious congregations, foundations...). The mode of financing of these institutions, public and private not-for-profit, has substantially evolved over the last two decades. In 2004, a "global allocations" scheme, which allocated institutions an annual budget relatively independent of their activity or the usefulness of their care offer was abandoned and replaced by "pricing per activity" (T2A). This marked change of logic places these structures under direct competition from the commercial sector and highlights a serious management shortage. Private not-for-profit institutions, more fragile than their public counterparts, because less protected by their private status, are thus subject to the difficult challenge created by any change of environment: disappear or innovate. We thus justify the choice of our field, the Augustines Clinic in Malestroit (CDA). The property of a religious community, CDA experienced in 1996 a serious economic and structural crisis, which led to the redefinition of its provision of care as well as the creation of a new managerial model. This model, described in a recent research work (B.Nautre, 2007), has led CDA to move from a local hospital (emergency, medicine, general surgery, obstetrics) towards an average capacity hospital (140 beds), specializing in geriatrics, palliative care, rehabilitation and follow-up care. CDA is part of an integrated Group including a retirement home, a training center, and a home-care hospital. Two other developments are of interest for our work, on the one hand the concern shared by the management team and the medical community for the development of a management tool, on the other hand the desire to base their strategy on strong ethical values and a sense of action. Since 2004, CDA has been

subject to T2A funding for 50% of its beds (medicine) and is preparing for the transition of all of its beds to this mode of funding in the near future.

The methodology adopted is intervention-research, in the sense proposed by David (2000): "intervention research is to help on the field, to design and implement models, tools and processes for managing [...], with goal of producing knowledge useful for action and theories of different levels of generality in management science". The choice of a single field, despite the bias that it can induce in terms of objectivity and external validity of the findings, promotes strong interaction between research and reaction of the Organization (Yin, 1994). This work aims to participate more with a capitalization of knowledge than to create or overturn a theory; it is part of the logic described by Wacheux (1996), a logic which "it organized an active and methodical consultation in the Organization, to produce knowledge on a process of change notified and enabled". Organized over a period of 3 years, the research has been structured by a pilot group, a researcher, managers and doctors from CDA, as well as specific skills required by the approach (IT management, nurses, functional managers). A journal, bringing together all the records of meetings, documents, bibliography, is viewable by all members involved (on the CDA intranet and on paper). This text renders the work carried out during the first 18 months.

2 Hospital governance or management by values In many research papers and publications dealing with hospital management, published during the last decade, two major trends can be identified and with them, inseparably, two approaches to management instrumentation. But the two paths meet regarding the specific difficulty of this type of organization to establish an effective relationship between the choice of management tools and their impact on the improvement of the performance of the structures (Clement, 2001, Cremadez, Jules, 1992, Claveranne, 2003).

2.1 Corporate hospital governance A first path, relying on the contributions of corporate governance (CG) in the field of health, studied the abilities and limitations of this model for a transposition to hospital governance (GH). It tends to promote the concept of rates, intrinsic to the T2A, to incite institutions to homogenize their practices. It requires internal behavior that matches global cost structures to the constraints of these unit rates. The question of the historical conflict between managerial and medical power has induced here the problem of the latitude of managerial power as well as location of power (Charreaux, 2010). This approach emphasizes the multiplicity of stakeholders that represent as many principal actors in the

agency theory. The measurement of performance and by way of consequence, the development of tools for measuring costs poses a series of questions. Measure for whom? The model of the ENCC is interesting here, in the sense where a number of health care organizations have volunteered for experimentation without any real involvement of medical communities (the language of the ENCC is strongly accountancy-based and models hospital activity into "units of analysis"), and without clarifying the purpose of the process, informs an external database to set rates or participate in the development of a new internal management approach. How does one measure? By relying on standard costs, which amounts to asserting the independence of costs by activity or to relativize their imprecise nature or even political breakdown of indirect costs. By relying on the marginal cost, which would imply that the increased activity in a speciality did not substantially influence the distribution of the burden of structure and induces only homogeneous short-term spending... Finally, who is involved in the process? Here again, confrontational relations between managerial and medical power are at the heart of the debate, management tools language increases the asymmetry of information thus the specific power of the management community, but at the same time justifies the withdrawal of other stakeholders (including the medical corporation) who oppose accounting logic to their professional logic. The same work also stresses non-independence, in organizations which are public service providers between authority and controlled organization (Muselin1999). The CG and subsequently the new public management model poses, as a precondition to any incitement to performance, independence between control and execution. The results from this approach are the centrality of the contract and the underlying idea, inspired by the agency theory, of a contract as a way to reduce the costs of monitoring the agent (the

hospital director) and encourage him to align his interests with those of

the principal (the regulatory body becomes representative of the interests of the user). But, in the real world, the definition of the terms of the contract is facing great difficulty in defining the borders of the organization (De Pouvourville, Letourmy, 1995), where does the mission of the hospital commence and stop, what about the competition, even the "agreements imposed"? The HPST law (Hospital, Patient, Health, and Territory) does not simplify this problem, it complicates it even, to the extent that it strengthens the capacity of interference by the ARS (Regional Health Authority) in the structure of markets, partly against the model of framed competition prevailing in the model organized by the orders in 1996. The Manager, in the agency logic described previously, tends to focus on the instrumentation model that strengthensspecific information and therefore its power over the organization. Thus, a "standard costs" (base of the T2A) model will be more comfortable in the sensethat it is disconnected from the

real cost structure. Management tools preferred by this model differ so little from the existing tools developed under the overall budget arrangements. These mainly budgetary tools are based on a representation of the organization in units of analysis: these, with regard to the volume of shared resources, leave the manager with a very wide latitude to meet its two major constraints. The first is to adapt to what is most likely to be funded by a regulatory body (Mimetic attitude) playing on the breakdown of its support costs. The second is to promote any particular activity (through the display of a return on investment or a partial loss), playing with indirect cost allocation rules to justify policy decisions.

2.2 Management and values at the hospital A second path is interested in the nature of the organizations studied, classified in the category of professional bureaucracies (Mintzberg.1991) or pluralistic organizations (Denis et al., 2004). More than the pitfalls associated with bureaucratic mechanisms, we focus here on one of the specificities of this type of organization which is the predominance of the institutional framework, patterns, values, serving as a landmark for actors in action (Valletta, Denis, 2005). In this reading, the performance of the hospital would be to search more within the capacity of the strategist to take into account and translate these values into action with the actors and stakeholders of the organization, and less to compel these actors and stakeholders to adapt their behavior to make their activity compatible with tariffs imposed ex ante. In this neo-institutional reading, we admit that players agree to engage in the search for performance only if the economic representation of their activity is in tune with their own meaning developed under their own schemes (Valletta, Karaki, 2005). The doctor in geriatrics will focus on correspondence between the homogeneous group in which his patient is classified and the duration of stay, when this attitude fosters that which is contained within its own value framework (thus it gives meaning to its action). He must understand that this mastery will increase the capacity of its service to meet its mission, and then increase the means at its disposal. The tools measuring costs resulting from this approach are quite different from those of the first approach. Firstly the need arises for internal design and not adaptation of a model imposed from outside. Here each professional or group of professionals should define how they contribute to the added value related to a stay or an activity. For example: the social service of a hospital contributes to the realization and the quality of a stay, in the same way as a care service. At the same time, the way the service perceives its contribution, and especially how it will be involved in the qualification of this contribution, will act directly on its interest to intervene in the process and buy into a learning logic related to the global strategy of its organization.

Even if it appears somewhat simplistic to oppose the two approaches, it is useful for the debate. On one side there is the construction of a new representation of the health care organization imposed from the outside based on a new corporate governance. On the other side, the development of an autonomous representation specific to each structure, inducing continuous learning, representations which would take into account the expectations of stakeholders and turn them into tools supporting performance incentives. One can say that hospital reform is today facing a challenge in terms of choice. In the first case, too much confidence in the possibility of transfer of the mechanisms of corporate governance into hospital governance, transposing without qualms management tools in respect of the results they have shown into industry or market services, may undoubtedly lead to the pitfall described by Or and Renaud (2009) 'the chosen mechanism, which is unique to the France .provides for a reduction of tariffs in case of increase of the overall hospital activity and not according to the evolutions of activity of each institution. This device, which makes no distinction between the different activities produced and which does not take into account the individual effort of institutions, is problematic and may cause adverse effects. This generates a highly opaque system for settlements with little foreseeable market developments. In addition, at level and equivalent activities range, an institution may be "sanctioned" in its funding because of strategic decisions made by other institutions'. In the other direction, the approach of autonomous self-construction, taking into account specific values of the studied organizational field, and most researchers agree to recognize the highly institutionalized nature of the health sector (Valletta, 2004), opens a way to make sense of cost control at the hospital. Where the market does not tell the value, it seems that it is starting from the value problem and its representation by stakeholders, that one can develop a shared vision of the strategy, and derive cost measures contributing to performance improvement.

3. The technical options available Let us examine in more detail, technically, these two tracks taking as reference the standard cost proposed by the regularity authority model. Is this representation built for the needs of the global regulation, relevant to the development of management tools specific to each structure? Can each structure embark on the development of a specific model? The functional model used by the regulatory authority distinguishes 3 functions: clinical function, the medico-technical function, the logistics function and 3 other frameworks: the mixed units of analysis, royalties from practitioners and subsidiary activities. The functions are divided into functional units or medical units. The allocation base between units of analysis are varied even if the duration of stay and the consumed euros are the two main bases. Specific work units, especially for the medico-technical

and logistics functions are proposed. It is a model of

health care organization based on craft

perspective. The coexistence of three subsystems, the public hospital, the for-profit hospital and the not-for-profit contradicts the Ashby law which states that to control a complex system one requires a control system of the same complexity. Therefore, there are obvious risks of perverse effects using a single cost system as a tool for piloting three diverse subsystems. However, it probably isn't for the supervisory authority to anticipate these risks but for the structures concerned to document these risks, to be able, through negotiations, to achieve convergence between the collective interest and the specific objectives of each system. So it's worthwhile looking at the inherent characteristics of the cost used as piloting instrument in an attempt to deduce some recommendations on its use within health care organizations. To quote Edmonson (1996), no doubt inspired by a Japanese proverb, "most practitioners carry a hammer and assumes the presence of nails ' one should avoid making the ENCC cost version the manager’s hammer. Whatever the system, the piloting is based on the ability of the pilot to measure both objectives and achievements, to guide decisions so that the latter maintain or bring back the system to the desired path. This is done in human organizations by an information system. The latter collects, processes and allows the linking of multiple measures, an expression of the variety of the system to be controlled. Used to controlling physical systems not reacting to the measure itself, except down to the subatomic level, we tend to think that measures in the social field are of the same nature as usual scientific measurements. But it is not so because the measure in social environment can be neither separated from its producers nor its users (Meason, Swanson, 1979). It is in this context that the concept of cost must be interpreted, as used in health system management. Management tools, and the cost system is a major one, are the result of a modeling process carried out by stakeholders who want to control it. This begs the question: does the notion of DRG cost express a sufficient representation of the real world to act effectively on it? Management control in any organization must be at the service of the strategy. If one accepts this hypothesis, representation proposed by the regulatory authority may be admitted as relevant only if one accepts the lack of strategy within the hospital structures (forced public service) or the neutrality of the information with regard to the strategy system. The lack of strategy can be expressed positively in the form of 'the only possible strategy is alignment with the standards of the Ministry model. The tariff based on ENCC indicates the "one best way". One must use this model as a functional benchmarking to progress towards best practices. The ENCC provides both economic constraints and an organizational target, in a word, it overrides local strategies to guide the actions within each of the

functional units belonging to the health care organization. We can easily see that this substitution can only eradicate the originality of the not-for-profit sector. It is vital for this sector to develop a model capable of supporting dialogue with the ENCC, and not be consumed by the underlying model represented by the ENCC.

The neutrality of management tools that prevailed until the 1970s is no longer defended. On the contrary, the evolution of tools has become one of the major levers of change in organizations. One cannot ignore the impact of the introduction of the concept of DRG cost on the operation of healthcare organizations. If we defend the idea of a minimum of institutional autonomy then we must accept that each should build its own representation, more or less close to that proposed by the Ministry, according to its degree of assumed autonomy and strategic willingness. Distancing oneself from the standard model opens doors to other dimensions than the economic dimension. It offers the opportunity to reflect on its own functioning, on the relations between social groups, between specialists, between communities of practice, in short, to launch bases of a learning structure, potentially innovative instead of being guided by a hypothetical 'one best way' that would emerge, as by magic, from the ENCC and the pro format P&L. Finally, it is also the only means, collectively, of getting closer by trial and error to the optimum scenario which may not arise mechanically from the application of the ENCC as stressed by Or and Renaud (2009). It is by confrontation and debate between the standard cost of the ENCC and specific costs of the various institutions that the regulatory authority can improve its management of the health system. As in all organizations, one can imagine, in health care organizations, the construction of an integrated information system such as an ERP, or a more flexible approach that accepts multiple business approaches reflected in specialized software. The multipolar nature of health organizations would act against the success of an ERP (Romeyer, 2003, Nobre 2009). The needs and expectations of the medical profession, those of the administrative staff, and those of external partners or branches will find it difficult to be formalized and to be adjusted during the implementation of the ERP project, inevitably bound in time and by available resources. Romeyer (2003) highlights the technical difficulties of overcoming and obtaining traceability among

the three areas to cover: information flow, physical flows and support activities for patients. Added to these technical difficulties are the difficulties associated with the implementation process, slow and complex as well as difficulties related to both external stakeholders (lack of experience in the health sector) and internal (difficulty of changing work habits). On the other hand, the acceptance of specialized software development, each with its own rate of evolution, risks depriving senior management of the transversal and synthetic vision necessary for developing a strategy for the institution. There comes a time where abundance must be channeled within a global model. The latter will have a dual function, on the one hand of offering a global vision of the functioning of the organization on the basis of the existing IS, and on the other hand, of contributing to the piloting of the future developments of the IS, to respond better to the expectations of the various stakeholders. The idea is then to allow the specialized development systems to expand their horizon but to develop their interfaces so as to foster a genuine central dashboard. The passage of local and specialized software to a global template is the necessary condition for the establishment of a central control device capable of articulating the different logics promoted by stakeholders and translated partly in different sub-systems. The development interfaces must respect a number of principles to deliver the expected service. The first is the completeness of data from heterogeneous sources. The second is the consistency and completeness of traceability. Finally the selection and construction of indicators which should be adapted to the needs of each institution. The encompassing model must have an economic 'layer' to ensure a minimum interfacing with the regulatory authority. Its construction can be organized according to two different logics: a logic of objects or an activity logic. Conventional models are built rather according to the first option. They try to answer the question: what (service) object has consumed what resource? The approach is analytical. Faced with an impossibility of resource allocation one will be sought by a technical detour (more or less sophisticated and more or less realistic) to operate the allocation of the resources concerned (Mevellec, 2005). Models built on a logic of activity are interested first in realized productions, whatever they are. Once all production and activities have been identified, one addresses the consumption of resources, which are consumed by the activities and incidentally by objects for which one wants to calculate the cost. Focusing on production allows the global template to respect local business logic while integrating it. It seems that it is therefore on this basis that the global model of health care organizations should be built if they wish to develop a forum for strategic debate. It is only on this basis that the economic

model can enables the different expertises which cooperate within institutions to become visible, instead of having them melt into the anonymity of the accounting categories representing various resources consumed by DRG. The ENCC retains as a basis: « In accordance with the principle of homogeneous unit of analysis, institutions must proceed with the splitting of their activity in units of analysis (UA). UA is a subsystem of cost assignment which implies a homogeneity of the activity. It must therefore allow the rapprochement between clearly identified resources and a precisely measured activity. "(p.13) The ENCC is placed in the second logic but things become less clear when it comes to the allocation of resources 'focusing direct assignment to the stays of a series of medical costs, in addition to their assignment to the UA' (p.36). Pictured here is clearly the first logic which aims to directly affect maximum costs to the single object that is the duration of the stay by DRG. This methodological floating does not in itself contribute to making the ENCC model a clear support for a strategic debate between actors of the most heterogeneous skills in economic and accounting. We believe that a model should convey only a single logic, although it should be able to answer multiple questions.

B. Design of tooling and confrontation in the field 1. The ad hoc model design 1.1 Construction of the model Two options are available to the designer of the model of a health care facility. A prescriptive approach and an empirical approach. The normative approach simplifies the work and will generally go much faster in the construction of the model. The standard is of course provided by the model of the ENCC. It is for the designer to check which activities are present in the organization, among those provided for in the ENCC. This task can be achieved in a confined area with minimal contact with the operational, be they medical or administrative. The model is presented as the reference to which the organization has to become closer. The risk is proportional to the gain of time and resources obtained during the construction: a non-appropriation of the model by the operational and consequently its difficult use for the management of the organization. The model will be the tool of the accounting department, representative of the 'accounting logic', the regulatory authority logic and not the image of the functioning of the organization because its design has not been a collective learning exercice.

The empirical approach involves the actors themselves in the design of organization representation, a representation in which they will recognize themselves. This condition seems fundamental to us to have the information from the model accepted and used. The counterpart is the unwieldiness of the approach and its cost (Thurnston, Kelemen, MacArthur, 2000). Nevertheless, these two critical viewpoints, largely reiterated in the professional press, must be relativized because they ignore an important side effect of the approach: the information and training of a very large number of actors in the principles of economic modelling, the demystification of the conventions and sharing of the choice of inevitable conventions of representation involved in modeling. This work with the operational will avoid, for example, separate maintenance of the buildings for their use to focus on the service provided: provision of functional m2 for other services. According to the organization and the size of the facility, cleaning, maintenance, can each be the subject of a separate analysis but what is important is to agree whether medical and administrative services should discuss the maintenance cleaning, heating, etc. or if it is the provision of the m2 which is the support of intra-organizational relations. In the medical sections one may be wondering, for example, about patient flow-related activities and those related to the length of their stays. Does one opt for two units of analysis to better identify the causality of resource consumption or can one accept some approximation to match the department and the unit of analysis? It is also at this stage that one decides how to qualify the value created by the different units of analysis and their selected production measure. Some units of analysis are identical to those of the ENCC, but it is interesting to compare the choices of the production measure. The case of the laundry is an example of the difference in points of view. On the one hand, an ENCC template centered on the technical function in the narrowest sense defines the volume of activity by the input, i.e. Kg of dirty clothes. On the other hand, our model deals not with the technical function but with the process that integrates the supply and delivery and defines the volume of activity by Kg of clean linen provided, which represents the value created for internal clients. The value support is more rewarding for staff and is gaining its support on the need to measure, while the same claim for the arrival of dirty linen met passivity and resistance under various pretexts. The case of catering also deserves a comment. While the accuracy of resource allocations and the fineness of the costs is a goal, the ENCC voluntarily neglects the cost of breakfasts and snacks. If all patients are fully hospitalized, this convention has no impact on the costs of the DRG. However if day hospitalization is equal to full hospitalization, the phenomena of subsidization is no longer negligible.

If we differ on the choice of certain choices, we appreciate the effort made in the context of the ENCC to move away from pure volumetric allocation bases. The choice of prescription lines for the breakdown of pharmacy operating expenses seems to us particularly appropriate.

1.2 The architecture of the model Two questions arise at this stage. Would one take into account the reciprocal benefits between units of analysis (Mevellec 1995): does the IS department provide services to the human resources department which in turn provides services to IS? On this point we suggest having the wisdom of the ENCC, i.e. not taking them into account. This increases the complexity of the model for a very relative benefit. This technique is to be used only if it is likely to highlight local costs significantly very different from those obtained without reciprocal benefits (direct cost). The second question concerns the organization of the units of analysis into a hierarchy. One aspect of this problem has already been mentioned above, it being the decomposition of a process into its activities. The second problem is the consumption of the production of one unit of analysis by another. When all consumers of the service provided by a unit of analysis are internal to the Organization, prioritization is necessary except to accept arbitrary allocation of costs to customers. Typically, support functions such as human resources management, an information system or housing service fall into this category. These choices having been made, it is possible to organize different activities depending on their role and their behavior within the organization. We propose to distinguish 5 types of resources, resources being understood here in the sense of production capacity at the service of the business.

Capacity resources

Resources linked to patient volume and duration of stays

Resources linked to patient flow

Prescribed medico-technical resources

Medical activities

Figure 1: Typology of resources The concrete translation of the proposed hierarchy can, for example, be as follows:

Human resource management

Housing

Laundry

Radiology/ cardiology

Administrative support

Kitchen

Balneotherapy

Geriatrics

General support activity

IS

Physiotherapy

Palliative care

Front desk

Ergotherapy

Medecine

Spiritual accompaniyng

Pharmacy

Functional and physical medecine

Social service

Follow-up care

Figure 2: Partial view of the model The units of analysis included in this hospital (with no technical platform) collect all costs. There is never double-billing on a unit analysis and on a stay. All costs generated by the operation of a unit of analysis are full costs as capability resources are consumed by all remaining units of analysis. The only costs which are allocated other than on a duly measured service consumption are the general support costs. Each unit of analysis is characterized by its production and it is the unit cost of that production which is the allocation base used to value the consumption of services by remaining units of analysis or cost objects. These measures of production are hardly different from those proposed by the ENCC, it is the mode of assessment of their cost and their interpretation which is different.

2. The model for the structuring of refunds. By opting for extreme logic one can imagine two antagonistic reviews. The first which places the patient at the heart of care organization, leads to the calculation of cost of a single object: the cost of the individual stay. The second, which adopts the point of view of doctors and medical knowledge leads to the presentation of the care institution in the form of a grouping of medical responsibility centers within which all other services merge. Before developing intermediate solutions, let us examine these two extreme situations.

2.1 The patient at the heart of the care structure. The relationship between the cost of the unit of analysis and the cost of stays is rightly at the heart of the approach of the ENCC. If this appears legitimate insofar as the ENCC must provide a basis for the tariff, this procedure can be considered risky and clumsy for an isolated institution. Certain expenses are simultaneously attributed to units of analysis and to patient stays. This approach blurs the total cost of the units of analysis involved because, according to the type of analysis to which we refer, the amount of expenses to be taken into account is no longer the same. This parallel allocation, in addition

to its computational complexity, increases the risk at the moment of the imputations and compromises the reliability of the cost of stays. It would have been more logical to build the cost of stay as a process during which the patient consumes various services, both administrative and medical. This would have allowed us to replace, for each stay, the double-billing by a detailed technical follow-up. Any medical act or any service from which the patient benefits should be documented in the patient’s record. It is on this technical basis that the cost of each patient's stay will be calculated here. Traceability must be one of the structuring concepts of the information system (Mevellec2005). Allocation on all units of analysis Based on their service consumption

Laundry

Length of stay

Kitchen

Administrative support

Front desk

1

consum ptions

Cost of stay

Margin

T2A for the DRG of the patient

Figure 3: Patient-centered modeling The economic review of the care provider is modeled on the calculation of a full cost. The patient is hidden behind the DRG and the contribution of the various services, whatever they may be, is treated in the same way. Kitchen and laundry are no different from physio-therapy or occupational therapy or medical activity per se. Patients are also independent from each other and nothing in the model underlies the solidarity which is at the heart of the French health care system and in particular the strategy of our not-for-profit hospital.

2.2 The medical unit at the heart of the information system Changing the way of looking at the hospital structure leads quite naturally to the development of a different economic review. The economic balance of the organization is no longer to examine every patient stay but to look at each identifiable medical unit within the hospital structure. We will not discuss here the detail of this display that is part of an internal power game. We are concerned only by the modalities for the review of the economic data which reflect the functioning of the organization. In this context all data must focus on medical units. The patient, in the same way as service providers, which are all considered as ancillary services, disappears behind the main services i.e. the medical units. The question is no longer whether the patient YYY... has consumed 3 sessions of physiotherapy and a consultation with a dietitian, but now the question relates to

how many sessions of

physiotherapy and how many dietary consultations have been prescribed by medical unit X. At the

same time, management no longer questions the profitability of this or that category of patient but the viability of such or such medical unit. The customized processing of YYY patient is no longer visible. What appears is the medical behavior globalized at the medical unit level. These are black boxes for senior management, which perceives only the overall economic result. This architecture mapping is shown below.

Housing

Laundry

Radiology/ cardiology

Geriatrics

Human resource management

IS

Kitchen

Balneotherapy

Administrative support

Physiotherapy

Palliative care

Ergotherapy

Medecine

Front desk

Pharmacy

Functional and physical medecine

Social service

Follow-up care

T2A income Margin by medicinal unit of analysis

Figure 4: Modeling focused on medical units

2.3 The result of a first Exchange with physicians. The presentation of the two extreme solutions to the hospital doctors led to a reaction whose source is in the truncated representation of their activity. "How is the patient's stay valued in our service?", "is there a traceability of the time that we devote to each patient?" 'I do not see pharmacy spending' "How do your costs take into account the quality of care and the quality of the environment that we offer to our patients?", "if one does not see patients how can one act to help to return to the economic balance if it is endangered? '' "all our patients are different and it is our duty to treat them as best as possible given our knowledge, how can this be justified?”, «the T2A rates are not the result of theoretical work but are based on an examination by doctors of best practices, we cannot be, without exception, very far from the implied standard that constitutes the T2A tariff. "You must give us information which would enable us to react to the causes of our potential overruns", "we are currently under DMS (average length of stay) pressure and alerts; if you put still other pressures on us, I despair of doing my job", "If the DMS is not an appropriate lever, then what is - it or they? '' At the end of more than an hour and a half of exchanges between physicians and between physicians and the project leader, it is clear that the agreement between the economic vision contributed by management accounting and the medical vision contributed by doctors could find strong support not for one tool, but for a combination of two new management tools for the hospital: a cost per medical unit and one technical dashboard by DRG. The chronological approach, which had been considered for cost calculation and the development of dashboards must be replaced by a synchronic approach of the

two management control tools. Non-financial dashboards, based on the analysis of all the medical requirements and the length of stay by DRG, both satisfies and reinforces, through new information, the logic of medical care, at the same time offering support for dialogue with managers, because the medico-technical parameters are all cost drivers. Income per medical unit simultaneously allows management to 'visualise' the contribution of each medical team to the overall functioning of the organization and offers senior management the possibility of better targetting its demands and improving the allocation of any free resources.

3. Dashboard as a boundary object The above management models should not be designed as purely economic models (Briers, Chua, 2001). A model provides us with a representation of the overall operation of the organization and its multiple productions. Each of these productions constitute in itself an area of management and a source of reflection on the production of value within the organization. He who speaks for the production of value must identify one or more clients as stated quality analysis and more widely, its stakeholders. We make the assumption that, on this basis, it is possible both to strengthen management on a business basis and transversal management basis of the strategy. Beyond the costs of stay, the model allows the preparation of multiple reports in an economic dimension, but also in a medical dimension, since all acts and syntheses are dealt with in the model. We believe that this capacity can be stimulated by the notion of boundary objects (Trumpet, Vinck, 2009) that will combine both medical and economic dimensions. Here we shall resume the theoretical perspective opened up by sociologists (Vink 1996, Star and Griesmer, 1989) in the study of the functioning of the networks of researchers and engineers. Even if our objects do not have all the features of the objects studied by the aforementioned sociologists, they are close enough be able to make reference beyond the traditional categories of management control (Zeiss, Groenewegen, 2009, Wenger 2000)

Accounting /OSII/GA/P MSI

Standardization of the translation

Questions/answer

P&L by medical unit

Dashboard by DRG

Questions/answers

Boundary object Abstract representation of the organization Community of medical practice

Management dialogue

Managerial group

Figure 5: The piloting tools as a boundary-object The boundary object formed by the coupling of a DRG dashboard

and a contribution margin per

medical unit allows us to support efficaciously both the dialogue within each community of practice and between the two communities through the various mechanisms of standardized translation available in the infrastructure of the organization’s information system and within the two communities. The patient record, OSII in the figure, is a good example of this effort to standardize medical information. This standardization allows a strengthening of internal exchange within the medical community while enabling it to interface with administration and management tools. We intend to continue to explore this track in the continuation of our work.

Conclusion Our purpose is less to criticize the ENCC than to show that

focusing on costs and not on overall

management is a risky approach to the organizations providing health services. Everyone recognizes the complexity of costs and the difficulty of organizing dialogue between management and the medical community. It would be a pity that responses to the T2A in hospital structures reinforce this problem instead of serve as a springboard towards a leap into the management of complexity. To make this leap when implementing cost systems, each institution must take the opportunity to think collectively about real functioning, joint productions and the social value thus produced. In exchange, the medical community should derive benefits in the form of a workable data base for improving practices, and senior management would be able to collect material for strategic reflection on its positioning and its alliances which serve its territory. In addition, that different categories of staff will be able to add an economic dimension to their professional thinking. In short, we believe that the

modelling issue

suggested through the ENCC is vital for the transformation of the health system and that it would be a tremendous waste to restrict it only to cost calculation and not to use it as the starting point of a

collective learning process (Carlisle, 2002). The use of the concept of the boundary object for this reflection appears to us extremely promising, as shown in our initiative, which is still in its infancy, and that we will continue in the context of our longitudinal case study. Finally, this research, like all intervention-research limited to one organization, cannot claim the generalization of the model proposed for all hospitals. However, depending on the categories of organizations that we describe in the first part, it appears that private not-for-profit sector structures, in particular those whose strategy is based on the attachment to value systems, institutions within the meaning of Granovetteur (1994), would be particularly suited to take advantage of this model: in fact, we want to show that the construction of a piloting model only finds its usefulness when it is understood and translated into sight structuring or networking (Callon, Latour, 1991), around the strategic object that the model wants to have emerged. Thus we can issue the principle whereby, in a type of organization that we have described as highly institutionalized, in an organizational field characterized by the absence of relationships between value and market, this strategic object is neither natural nor predefined. It is instead the result of a collaborative construction. The meaning attributed to it by the organization stakeholders, that is to say, its management effectiveness and its role in inciting performance, results more from its ability to comply with their individual expectations than from any specific power linked to any particular management logic. It is, therefore, on this strategic object, in the sense of a boundary object,that we should focus. The fact of being specific to each organization, with its evolving, unstable character, thus inciting the construction of fitted and flexible models, must be recognised. Thus, whether it is the ENCC or an ad hoc model, the field to explore is vast, and the debate is at the center of contemporary issues in the context of hospital system reform.

Bibliography AITH (2010) Guide méthodologique de l’ENCC, avril, document PDF Briers M. Chua W.F (2001). The role of actor-network and boundary objects in management accounting change: a field study of an implementation of activity-based costing, Accounting, Organizations and Society, n° 26

Brown J.S and Duguid P (1991). Organizational learning and communities of practice: toward a unified view of working, learning and innovations, Organization Sciences, 40-57 Carlisle P.R, (2002). A pragmatic view of knowledge and boundaries: boundary objects in new products development, Organization Sciences,vol 13 Charrea ux G. (2010) La gouvernance hospitalière : quelques réflexions à partir de la

gouvernance d'entreprise ; http://leg.u-bourgogne.fr/wp/1100705.pdf Claveranne J.P. (2003), L’hôpital en chantier, du ménagement au management, Revue Français de Gestion, septembre octobre, vol 29, n° 146, pp.125-129, 2003 Clement J.M. (2001), 1900-2000. La mutation de l'hôpital, Les Etudes hospitalières, Bordeaux, Cremadez, Grateau (1992), Le management stratégique hospitalier, cités par KRIEF. pp. 111-119, Inter Editions David A. (2000), La recherche intervention, un cadre général pour les sciences de gestion, Actes de la IXème Conférence de l’Association Internationale de Management Stratégique, Montpellier, 24-26 mai Denis J.L., Langley A., Rouleau L.(2004), La formation des stratégies dans les organisations pluralistes : vers de nouvelles avenues théoriques, actes des 13ème conférences de l’AIMS, Normandie, De Pourville, Letourmy (1995), dans : les transformations des systèmes de santé en Europe, vers de nouveaux contrats entre prestataires, payeurs et pouvoirs publics, éditions ENSP, pp. 7-19, Edmondson A.C. (1996). Three faces of Eden: the persistence of competing theories and multiple diagnoses in organizational intervention research, Human relations, 49 (5), 571-595, cite par Virkkunen et alii Enthoven A.(1985), Systèmes de santé, comparaisons internationales, Paris, Institut la Boetie, Le Moigne J-L.(1977). La théorie du système général, PUF Paris Levit et March (1988). Organizational learning, Annual review of Sociology (14), 319-340 Mason R, Swanson E.B. (1979). Measurement for management, décision and perspective, California Managment Review Mévellec P. (1995). Calcul de coûts dans les organisations, Collection Repères, La Découverte, Paris

Mévellec P. (2005). Systèmes de coûts, Dunod Paris

Musselin C.(1999), comprendre les systèmes universitaires, publications du CSO, juin Moisdon J.C, Tonneau D. (1999) La démarche gestionnaire à l’hôpital, T1 , Seli ARSLAN ed. Nautré B. (2007). Hôpital privé à but non lucrative et nouvelle gouvernance publique : de l’organisation pluraliste au projet stratégique, 6ème conférence internationale de gouvernance d’entreprise, Genève Nobre T. (2009). Le système d’information hospitalier : quelles caractéristiques privilégier pour favoriser le développement des outils de contrôle de gestion, Congrès de l’AFC Strasbourg Nonaka I. and Takeuchi H. (1995). Managing the firm as an information creation process ; Advances in information processing in organizations, 239-275 Or Z. et Thomas R. (2009). Principes et enjeux de la T2A à l’hôpital, enseignements de la théorie économique et des expériences étrangères, papier de recherche IRDES Romeyer C. (2003). Analyse des obstacles à l’implantation d’un système d’information hospitalier traçant les activités : quels enseignements pour les SI et la traçabilité AIM, Grenoble Star S.L. Griesmer J, (1989). Institutional ecology, “Translation’, Boundary objects: amateurs and professionals on berkeley’s museum of vertebrate zoology, Social Studies of Sciences Stiles R. (1997). What is the cost of controlling quality? Activity-based cost accounting offers an answer, Hospital and Health Services Administration, summer. Thurnston K.L Kelemen D.M and MacArthur J.B,(2000). Providing strategic activity cost information : cost for pricing at blue cross and blue shield of Florida, Management Accounting Quarterly, spring, Montvale, Fl. USA Trompette P. Vinck D.(2009). Retour sur la notion d’objet frontière, Revue d’anthropologie des connaissances, vol 3, n°1 Valette A, Denis J.L.,(2005) Quand le nouveau management public ouvre la boite de pandore des restructurations, actes du colloque de l’AIMS, Angers, 2005

Valette A., Karaki H. (2005), Dynamiques professionnelles et projets économiques de changement : questionnements autour d’un cas de restructuration hospitalière français, XIème conférence de AGRHParis Dauphine, 15 – 16 septembre Virkkunen J., Kuutti K, (2000). Understanding organizational learning by focusing on “activity systems”, Accounting, Management and Information Technologies, vol 10 Wacheux F. (1996), Méthodes qualitatives et recherche en gestion, Economica Gestion, page 93, Wenger E. (2000) Communities of practice and social learning systems, Organization, vol 7 Yin R.K. (1994), Case study research: Design and methods, second edition Beverly Hills, Sage, 1994 Zeiss R. Groenewegen P (2009. Engaging boundary objects in OMS and STS? Exploring the subtleties of layered engagement, Organization, vol 16,