Constructing Knowledge on Hospital Activity in France

Transcription

1 Renaud GAY PhD student in political science Institute of Political Studies Grenoble University, France XXIInd World Congress of Political Science RT 25 - Policy Analysis, Health Politics and Policy Change : Arguments, Evidence and Policies Constructing Knowledge on Hospital Activity in France An exploratory historical sociology of measuring devices Introduction The hospital sector has been the most central piece of the French healthcare system since the 1960s so that cost containment has focused on hospitals for more than thirty years. Several economic and sociological studies can have however highlighted difficulties in controlling this organization whose activity is not precisely known. For example, hospital can be described as a professional bureaucracy, which is characterized by the lack of managers control and influence over the clinical-professional activity (Mintzberg, 1982). Furthermore, the hospital regulation is confronted with the agency dilemma according to economic analyses (Mougeot, 1994). In other words, the incomplete and asymmetric information between hospitals and state authorities yields problems of moral hazard, of adverse selection and of free rider. These scientific analyses have sharply influenced the governmental authorities, which have implemented several devices supposed to improve their knowledge on hospital activity (Serré, 1999 ; Benamouzig, 2005 ; Savoir/Agir, 2008). The paper aims at examining these various measuring devices of hospital activity, which have been designed, implemented and sometimes abandoned from the 1970s onwards. The paper only deals with the public hospitals, although the private hospitals play a large part in the French healthcare system. These devices are technical instruments intending to measure the hospital activity. More precisely, they are based on a specific meaning of the hospital activity, which can differ between devices. Lastly, these are not only tools of knowledge but they are designed as a part of a broader policy. Consequently, they pursue political purposes that governmental authorities define more or less clearly. The paper refers to two types of scientific studies. First, several analysts stress the current development of a new way 1

2 of governing based on the quantification. This way of governing appeared in the French hospitals at the beginning of the 1980s so as to contain the hospital expenditure by controlling the healthcare delivery (Mayere, Grosjean, Bonneville, 2009). The quantifying devices are mainly a way of describing the social world so as to act on it (Desrosières, 1993). Besides, these are particularly developed by the managerial state, which tries to freeze political decisions into technical tools (Ogien, 1995). Second, a new research program means to understand public policy through its instruments (Lascoumes, Le Galès, 2004). This political sociology approach to instruments asserts that these are not neutral devices. They comprise a specific idea of regulation and produce specific effects (Lascoumes, Le Galès, 2007). Lastly, the policy change can be analyzed through its instruments. As this approach suggests it, the paper focuses on the construction dynamics of the measuring devices 1. The purpose is to scrutinize the social, political and historical conditions of their construction and to understand their trajectory (Bardet, Jany-Catrice, 2010). Consequently, the social practises resulting from the device implementation are not analyzed. In addition, the paper follows a socio-historical perspective (Déloye, 2003 ; Laborier, Trom, 2003). It pays attention to institutionalization or no institutionalization of these devices and to the various temporalities of their construction. The methodology is mainly based on the study of historical records from the health administration and ministers of health. A part of the records is located in the National Archives, while the latest ones are still in the Ministry of Health. Additionally, a few senior civil servants were interviewed (n = 9). Several questions underlie the paper. What do the various measuring devices of hospital activity exactly measure? What political ends do they more or less explicitly follow? How can the trajectory of these devices be explained? Part one describes the attempts of determining the right hospital costs in the 1970s and their failure. In the second part, the implementation of a major measuring device in the 1980s is analyzed. This controversial and competed tool aims at medicalizing the measurement of hospital activity. Part three demonstrates how this new measuring instrument is used to fund hospitals and how it strengthens the state position in the hospital system. 1 The paper is based on a PhD dissertation in progress, which deals with de French hospital reforms between the 1960s and the 2000s in a socio-historical perspective. 2

3 I Health administration looking for a transparency of hospital charges (1960s- 1970s) State authorities have faced since the 1950s a great increase in hospital spending. To control it, they design a funding device based on an accurate knowledge of hospital activity. This system intends to fund hospitals at the right costs (1). Technical and institutional obstacles nevertheless hinder this project (2). 1. Funding hospitals at the right cost From the 1950s on, the increase in hospital spending is large. The annual growth rate of medical consumption is 6.3% between 1956 and 1960, 10.3% between 1961 and 1965, and 7% between 1966 and The share of hospital consumption amounts roughly to 40% of medical consumption at this time. This is mainly explained by technical progress in the medical field, the development of health system and the shift in the hospital role, from a social role to a more medicaloriented one (Valat, 2001). Health administration and social security are concerned about this increase during the 1960s for they fund a growing share of hospital costs: 78% in 1955 to 90% in In addition, state authorities and social security have poor information about the hospitals functioning. They act like blind payers, while several administrative reports denounce the ineffective management of hospitals. Two types of information on hospital activity are available to the public authorities: - First, four indicators informally called the four old figures (les quatre vieux) are supposed to mirror the hospital activity: the number of days of hospitalization, the number of admissions, the average length of stay, and the bed occupancy ratio. These indicators are based on two measurement standards, which are the bed and the day of hospitalization. These standards are however considered ineffective to evaluate an ever more technical activity. On the contrary, they are founded on a hotel conception of hospital, which does not take into account the medicalization of the hospital product. - Then, the price per day (le prix de journée) is the main mechanism of hospital funding and also the second source of information on hospital activity. Set by prefectorial decree, the price per day is obtained in each hospital with 3

4 the division of estimated annual spending of each department (medicine, surgery, obstetrics) by the number of effective days. It is above all a funding tool since payments from state and health insurance are calculated according to the number of days and their price. The price per day is also an evaluating tool as a high price would be the evidence of mismanagement and vice versa. The price per day is yet criticized for two reasons. It only provides a loose estimation of the real cost of service as its calculation blends spending of accommodation, of diagnosis, of treatment and of administration. The calculation of price is also biased because it prompts administrators and practitioners to uselessly extend the length of stays in order to multiply the days to increase financial resources. In this context, the law on 31 December 1970 plans a new funding device, which has to reveal the real cost of services. The objective is to optimize the spending by funding hospitals at their right cost. Three interrelated reforming initiatives try to determine the real hospital costs: the reform of funding system, the insertion of the management analysis, and the computerization of hospitals 1. The reform of funding system is designed by the new Direction des Hôpitaux (the administrative office in charge of hospitals), which is created inside the Ministry of Health in Its reform called fragmented price per day (prix de journée éclaté) aims at refining the knowledge and at strengthening the cost containment but also at creating incentive mechanisms improving the hospital productivity 3. The reform has therefore two aspects. It plans to clarify the calculation of the price per day by distinguishing an accommodation price from a care price. This distinction requires the reorganization of hospital budgets in order to isolate the medical expenditure. The fragmented price per day is coupled with incentive mechanisms, 1 These initiatives are linked to the program called «Rationalisation des Choix Budgétaires» which is inspired by the American «Planning, Programming, and Budgeting System» program and is launched in 1968 in France. Cf. : Bezes Philippe, Réinventer l État. Les réformes de l administration française ( ), Paris, Presses Universitaires de France, 2009 ; Benamouzig Daniel, La santé au miroir de l économie. Une histoire de l économie de la santé en France, Paris, Presses Universitaires de France, Décret du 13 novembre 1970 portant organisation de l administration centrale du Ministère de la santé publique et de la sécurité sociale. The Direction des Hôpitaux brings together the former Service des établissements and Direction de l équipement social. Cf. Valette Jean, Etude sur l évolution des structures de l actuelle direction des hôpitaux de 1945 à 1976, Revue française des affaires sociales, n 1, janvier-mars 1977, p National Records (NR), Collection of the Direction de la Sécurité Sociale (DSS), , Direction des Hôpitaux, Projet de décret relatif au fonctionnement financier et comptable des établissements assurant le service public hospitalier, 15 novembre

5 which have to give the hospital agents a sense of managerial responsibility: profitsharing system, digressive care price for the long stays, etc. The Direction des Hôpitaux also promotes the insertion of cost-containment and cost-measurement tools in the hospitals. This initiative is deeply linked to the reform of funding system. On the one hand, the new organization of the hospital budgets is supposed to make the cost analysis easier. On the other hand, these managerial tools have to improve the hospital productivity like the fragmented price per day. The Direction des Hôpitaux establishes an ambitious program of hospital management 1 that the Minister of Health Robert Boulin 2 supports. Units of management analysis are created in the Ministry of Health and in the regional hospitals. Their main purpose is to analyse the shaping of hospital costs and to compare the hospital costs to each other. These managerial innovations require a computerizing of hospitals that the Division Organisation et Méthode Informatique (DOMI the administrative office in the Ministry of Health in charge of information technology) undertakes from 1971 onwards. Taking advantage of the Plan Calcul launched in 1966 (Baudot, 2011) 3, the DOMI develops several computer programs called national applications 4, which are supposed to improve the hospital functioning. These programs should ease not only the collection of data necessary for the management analysis but also could automate this analysis. The national application named SHAGE precisely tries to automate the management analysis through the regular production of performance indicators The impeded determination of the right costs First of all, the uncertainty around the reform of hospital funding system jeopardizes the plan of defining right hospital costs. From 1971 to 1978, the reform is negotiated by various administrations before being experimented in a few hospitals. The slowness of the reform is explained by the institutional structure of the 1 NR, Collection of R. Boulin, , Jean Baudouin, directeur des hôpitaux, Note à l attention de M. le Ministre, 15 mars R. Boulin is the Minister of Health from 1969 to The «Plan Calcul» is a governmental program, which intends to guarantee the national independency in the computer sector. 4 Three national applications can be noticed : GRAPH for the staff management ; GEMCO for the inventory control ; SIGMA for the inpatient management. 5 NR, Collection of the Direction de l Administration Générale, , DOMI, SHAGE : Système Hospitalier d Analyse de Gestion. Présentation, mai

6 administration: the Direction des Hôpitaux is isolated and has not the enough authority to impose its device. Although the Direction des Hôpitaux is characterized by large technical capacities, it has to reach a compromise with the three other administrative offices of the Ministry of Health: the Direction Générale de la Santé, the Direction de la Sécurité Sociale and the Direction de l Action Sociale. Whereas all the administrative offices agree with the idea of fragmented price per day, they all oppose the incentive mechanisms that they consider costly and complex. Though the reform of the Direction des Hôpitaux is barely supported by the health administration, Boulin defends this plan, which comes up against the veto of the Ministry of Finance and of the Ministry of Interior 1. These two institutional veto players fear that the reform results in an increase of hospital spending (Tsebelis, 2002). The economic stagnation in the middle of the 1970s strengthens the administrative coalition opposed to the reform of the Ministry of Health. The control of health care expenditure becomes the new priority on the governmental agenda (Palier, 2005). The Direction de la Sécurité Sociale designs an alternative plan competing with the device of the Direction des Hôpitaux and backed by the Ministry of Finance. It proposes to replace the price per day system with an annual resource allocation handed out to each hospital. Imported from Quebec, this tool makes a macro-economic control on the hospital expenditure possible. In 1978, a simplified version of the fragmented price per day and the annual resource allocation are experimented in several hospitals until the adoption and the extension of the annual budget allocation in A second factor explains the administrative failure to determinate the right hospital costs. It refers to the heteronomy of the administration, particularly of the DOMI, which impedes the diffusion of managerial and computer tools. The heteronomy results from the technical and institutional dependency of the DOMI toward its environment: - The DOMI defines the objective of the computer policy but its lack of means forces it to contract out the development of the computer programs to private companies such as STERIA, SLIGOS or AVA, which are founded in France 1 NR, Collection of the DSS, , Le Ministre de la Santé publique et de la sécurité sociale R. Boulin, Note au cabinet du Premier ministre, 21 avril Décret du 11 août 1983 relatif à la gestion et au financement des établissements d hospitalisation publics et privés participant au service public hospitalier 6

7 - from the 1960s onwards. A computer department is set up inside the National Centre for Hospital Equipment 1 in 1975 so as to reduce the administrative technical dependency. The Centre is yet relatively independent from the state authorities and leans on private companies to maintain the computer programs. The institutional dependency of the DOMI is linked to the legal status of hospitals, which are public organizations with managerial and financial independence. The implementation of the computer policy is performed not by the DOMI but by Regional Centres for the Hospital Information Technology (Centres régionaux de l Informatique Hospitalière) that the regional hospitals run. These Centres unevenly deploy the national computer programs and can create their own computer systems. This centrifugal dynamics explains why the administration does not manage to set up the SHAGE program, which is though the centrepiece of its managerial computer system 2. II The invention of a medico-administrative language (1980s-1990s) Despite the failure of the reforming initiatives in the 1970s, administrative and hospital agents do not give up the idea of measuring the hospital activity. Contrary to the former initiatives, they think about a tool of measurement independent of the funding device. What henceforth matters is to medicalize the measurement so as to determine not only the activity costs but also the hospital product (1). So, the health administration forges a new tool of measurement called Programme de Médicalisation des Systèmes d Information (PMSI) whose aims are ambiguous and implementation is challenged (2). 1. Medicalizing the measurement of activity At the end of the 1970s, the Direction des Hôpitaux and the DOMI want to medicalize the measurement of hospital activity. The objective is not to determine precisely expenditure in hospitals any more. It is henceforth to make a link between 1 Created in 1974, the Centre is in charge of informing public hospitals about the available medico-technical equipment and of rationalizing its manufacturing process. 2 NR, Collection of the Direction de l Administration Générale, , DOMI, Note à l attention de M. le Directeur des Hôpitaux, 29 mai

8 expenditure and features of the medical activity (type of diagnosis, type of treatment, etc.) 1. The administration has the required tools to appraise the financial and material means of hospitals but it has no device that collects medical data. Devices are however set up independently by several hospitals with the assistance of Regional Centres for the Hospital Information Technology. Instead of imposing a central standard system, the health administration decides to support and to develop local current experimentations, particularly two of them: - The DOSTAM 2 system is used in Grenoble from It intends to centralize the medical records of all the inpatients, which enable to carry out a statistical study on the hospital morbidity. The medical records supply three types of information: on the inpatient (age and sex), on the stay (its length), on the medical care (diagnosis, treatment). This system is yet based on the World Health Organization s classification of diseases whose the complexity prevents a global analysis. Moreover, it only supplies medical information without economic dimension. - The Gustave Roussy Institute in Villejuif tries to solve these two problems through a new measuring device of hospital activity, which is backed by the Direction Générale de la Santé 3. The experimentation is based on the production of boards of activities in each department, blending a clinical and an economic vision of the hospital activity, These boards comprise three items: information on the inpatients and their medical care, information on human and material means in each department, and deviation analysis between predicted activity and effective activity. They aim at defining pathological types and at connecting these types to used means in order to determine precisely medical care costs and to optimize the means 4. In 1983, the administration sets up a new measuring tool of hospital activity, the PMSI 5. While the measuring innovations such as the DOSTAM system did not 1 Records of the Ministry of Health (RMH), Collection of the Direction de l Hospitalisation et de l Organisation des Soins (DHOS), , Le Directeur des Hôpitaux, Stratégie globale pour une évaluation de l activité hospitalière, 8 novembre DOSTAM is the acronym for Dossier médical résumé et Statistique de Morbidité. 3 The Direction Générale de la Santé is directed by P. Denoix from 1974 to 1978 who is the chief of the G. Roussy Institute from 1956 to NR, Collection of the DSS, , DGS, Mesure de l activité hospitalière à l Institut Gustave Roussy, novembre Circulaire n 16 du 18 novembre 1982 relative à l informatisation des hôpitaux publics. 8

9 jeopardize hitherto the hospital autonomy, the measuring standardization enables public authorities to reassert their control on hospitals. The PMSI aims at measuring the hospital production by connecting medical activity and expenditure. This connection is set by a new classification of stays based on the Groupes Homogènes de Malades (GHM, Homogeneous Patient Groups). Each stay is translated in a Résumé de Sortie Standardisé (RSS, Standardized Exit Summary), which consists of a few indications on the stay: principal diagnosis, secondary diagnosis, age and sex of patient. The RSS enables to include a stay in a GHM, which is characterized by a medical and an economic homogeneity. Thus, the activity of each hospital and department can be summarized by its case-mix which amounts to the number of stays per GHM. The PMSI results from two processes: the importation of an American device and the mobilization of academics and administrative agents. The GHM are the French version of the Diagnosis Related Groups (DRGs) that Prof. Fetter of Yale University created in the 1970s. This system of classification intends to identify the products that the hospitals provide. In other words, the provision of medical care is reduced to some products, which are homogeneous in economic and medical terms. From 1983 onwards, DRGs are used to determine how much Medicare pays the hospital for each product. The importation in France of the DRGs classification is carried out by the chief of the Direction des Hôpitaux between 1981 and 1986, Jean de Kervasdoué. The latter belongs to academic, hospital and political networks. He meets Fetter at a conference on the healthcare systems that he organizes in In 1982, de Kervasdoué sets up a PMSI team into the Direction des Hôpitaux. The small team leans on researchers in management, mainly from the Centre of Management Science in the École des Mines de Paris, in order to adapt the DRGs tool to the French context (Serré, 2001). These researchers provide an expertness to the administration and find a scientific interest in their collaboration. After trying to implement managerial engineering inside hospitals, they are indeed interested in the global regulation of the healthcare system (Moisdon, Tonneau, 1999 ; Engel et al., 2000). The researchers come to two conclusions 1. First, the implementation of the DRGs classification is technically possible. Actually, the 1 RMH, Collection of the DHOS, , École Nationale Supérieure des Mines de Paris, L analyse des systèmes d informations hospitaliers dans le cadre du projet PMSI. Le cas des quatre établissements de la région de Grenoble, août

10 DOSTAM system provides the required data to the shaping of French DRGs. Second, the GHMs cannot be used in a financial way so as to protect public hospitals. The GHM classification has only to be used to improve the hospital management. 2. Controversy and uncertainty around the PMSI device Until the beginning of the 1990s, public authorities are not clear about how they want to use the PMSI. Although the tool is initially thought as a mean to suit the resource allocation to the hospital activity, public authorities seem to hesitate between two objectives: using the PMSI as an internal tool allowing a better resource control by the hospital managers or using it as an external tool improving the allocation of resources by the state authorities (Pierru, 2009). The ambiguous purpose of the state authorities is largely explained by the hospital agents suspicion about the tool. They criticize it for its strictly economic conception of the hospital activity since care quality and social role of hospitals are not considered. Moreover, academics do not agree on the use of the PMSI tool. The researchers of the Centre of Management Science suggest a managerial use of it: the PMSI provides information to hospital agents, which is supposed to help them to improve their organization. Conversely, the neoclassical economists defend a financial use of the tool founded on the theory of yardstick competition. Like in the United States, health insurance should pay a hospital according to its GHMs and their binding prices. This device should prompt the hospital agents to reduce their functioning costs in order to provide services under or at the fixed price. Despite the controversy over the tool, the PMSI is gradually implemented (Lenay, Moisdon, 2003). From 1982 to 1983, it is experimented in several hospitals, while a computer engineering propping up the PMSI is developed (RSS, classification of GHMs). From 1989 onwards, the PMSI is implemented in all the hospitals. The latter are compelled to collect the RSS for all the stays in medicine, surgery and obstetrics by 1992 and to set up a Department for Medical Information (DMI). A major institutional innovation, the DIM has to collect the RSS and to encode them into GHM. In 1992, the classification of GHMs is completed by the National Cost Study. This tries to determine the cost of each GHM on the basis of a sample of about fifty hospitals having the same cost accounting. 10

11 Uncertainties around the PMSI favour the development of other measuring tools of hospital activity. These intend to complete or to compete the PMSI. Two prominent initiatives in the 1990s can be highlighted: one supported by the Groupe pour une Amélioration de la Comptabilité Analytique Hospitalière (GACAH, the group for improving the hospital cost accounting) and the other one by the French Hospital Federation (FHF): - Created in 1991, the GACAH is formed by a sample of hospitals, which experiment a standard and improved cost accounting 1. This group does not oppose the PMSI device but tries to improve it. The group specially criticizes the PMSI for providing no information on the real costs of hospital activity. On the contrary, the cost accounting that the GACAH wants to spread enables to analyse the shaping of the hospital costs on the basis of main activities such as laundry or medical imaging and not of clinical stays. The ultimate purpose of the plan is that each hospital can compare its operating costs with those of the GACAH sample so as to identify potential extra costs. - In 1995, the FHF creates a Statistical and Financial Data Bank (SFDB), which is expected to compete the PMSI 2. The FHF is an association founded in 1924, which is comprised of the hospital administrators and is regarded as one of the main players in the health policy-making process. The FHF primarily tries to defend the self-governing organization of the hospitals that it considers as threatened by the PMSI device. Developed by and to the hospital agents, the data bank supplies management dashboards to the hospitals contributing to the SFDB plan. These dashboards are designed on the basis of secondhand data (budget accounts, state surveys) that the FHF collects. These are described as managerial tools, which enable the hospitals to benchmark their activity. Besides, the dashboards try to mirror complexity and characteristics of hospital activity: they are grounded on a cluster analysis according to the type of hospitals (local or regional hospitals) and on several indicators, on contrary to the PMSI, which provides a single indicator, the GHMs case-mix. 1 RMH, Collection of the DHOS, , GACAH/DH, Utilisation de la comptabilité analytique d exploitation hospitalière Calmon Michel, Hantour Zinna, La Banque hospitalière de données statistiques et financières, Revue hospitalière de France, n 5, septembre-octobre 1995, p

12 III Quantification of the hospital activity and state expansion The PMSI progressively becomes the main quantifying tool of hospital activity. Requiring a major state reinforcement, this work of large-scaled quantification allows hospitals to be funding on the evaluation of their activity (1). The new funding system based on the activity is supposed to trigger off internal reorganizations of the hospital activity that the state authorities directly support and run (2). 1. Basing the hospital funding system on the PMSI The launching of the National Cost Study in 1992 confirms the financial use of the PMSI. It defines the cost of each GHM, which is translated into a point score on a Synthetic Activity Index, the Indice Synthétique d Activité (ISA) 1. The production of each hospital is valued by the addition of the ISA points corresponding to the GHMs that the hospital produces. In 1996, the health administration decides to use the PMSI and the ISA index to reduce budget disparities between regions according to their hospital productivity. This is measured through the value of an ISA point, which is get by the division between the regional budget allocation and the number of ISA points produced by the hospitals in the region. State authorities cautiously use, however, the PMSI, which weakly impacts the internal organization of hospitals and their activity (Moisdon, 2000). This new major use of the PMSI requires a hard technical work that the health administration and its PMSI team cannot accomplish alone. Consequently, it decides to include a few academic agents in this quantifying process at the end of the 1980s. The Centre de Traitement de l Information du PMSI (CTIP, centre for date processing from PMSI) is founded in 1989 in the University of Paris VI. It is mainly composed of doctors proficient in hospital computer technologies and it determines the GHMs costs with the assistance of academics from the École des Mines and from Polytechnique. It also processes the PMSI data that hospitals convey to it. Then, the Pôle d Expertise et de Référence National des Nomenclatures de Santé (PERNNS, a centre in charge of health nomenclatures) is created in Controlled by the medical profession, the PERNNS is designed to improve the GHM classification so 1 The Synthetic Activity Index is a relative scale of costs. The calculation of the point score for each GHM is based on the assumption that the GHM n 540 (a standard childbirth) is worth 1000 points. 2 The PERNNS is headed by Marion Girardier-Mendelsohn who supported the DOSTAM system in Grenoble at the end of the 1970s. 12

13 that it might fit with the new medical practices. This technical device can be regarded as the outsourcing of the PMSI data processing by state authorities. This device is yet criticized because of its weak legal statute, which prevents, in addition, any administrative control 1. Consequently, the health administration undertakes the reform of this organization at the end of the 1990s. The Direction des Hôpitaux decides to merge the PMSI team, the CTIP and the PERNNS into a single public administrative institution, the Agence Technique pour l Information Hospitalière (ATIH, a technical agency responsible for hospital information) 2, which can be described as a quantifying state apparatus (Ogien, 2009). Actually, the state authorities largely control the agency in that they rule its governing board and some administrative agents into the Direction des Hôpitaux permanently monitor the agency s work. The ATIH plays a role of expert and gives a strong technical capacity to the health administration (White, 2009). Besides, it looks like a peripheral bureaucracy insofar as this technical agency reintegrates private expert activities into the state (Benamouzig, Besançon, 2005). The ATIH plays a central part in the implementation of a new hospital payment system called tarification à l activité (T2A) in for the latter is based on the PMSI data. The new centrality of the PMSI in the hospital regulation results in a marginalization of the other measuring tools such as the SFDB of the FHF. The new funding system creates a technical and economic reliance of the hospital agents on the PMSI tool. The T2A is a piece of a broader reform program entitled Plan Hôpital 2007 launched in 2002 by the neoliberal Minister of Health, Jean-François Mattei. This program has been analyzed as a corporatization of the hospitals and a marketization of their regulation (Domin, 2004 ; Pierru, 2007) 4. The structure of the new funding system is complex. On one side, the stays in departments of medicine, surgery and obstetrics are paid according to the national predefined price of their GHMs 5. The T2A can be therefore described as a prospective payment system. On the other side, the missions of public utility (teaching, research) are funded by annual budget allocations. A mechanism of macro-economic regulation is introduced so as 1 Cour des Comptes, Rapport annuel au Parlement sur la sécurité sociale, Paris, Journal Officiel, Décret du 26 décembre 2000 portant création de l ATIH 3 Loi du 18 décembre 2003 de financement de la sécurité sociale pour The reform plan is organized around four consistuents : boosting hospital investments ; simplifying the hospital planning ; reforming the internal governance ; paying hospitals according to their activity. 5 The mechanism of ISA points disappeared with the implémentation of the T2A. 13

14 to prevent any increase of hospital expenditure. It refers to the setting of an objective of hospital spending in the medicine, surgery and obstetrics departments. If this objective is overspent, the prices of GHMs are lowered. 2. Getting in the hospital black box The new hospital payment system can have been analysed as the state unwillingness to get in the hospital black box. Inspired by the Foucauldian concept of neoliberal governmentality (Foucault, 2004), some studies regard the T2A as a tool allowing state authorities to govern hospital from a distance and to impose, by minimizing the political costs, reorganizations of the healthcare system and inside hospitals (Pierru, 2008). Grounded on utilitarian principles, the T2A is supposed to reveal potential extra costs of any activities and to urge hospital agents to change their activities. Linked to the T2A, a new internal organization of hospitals is besides set by the reform plan Hôpital 2007 in order to improve their efficiency 1 : the clinical departments have to be grouped together into several activity centres (pôles d activité) having a significant management autonomy and the structure of governance is deeply modified. However, the idea of a government from a distance can be qualified in that this organisational concern leads state authorities to reinforce their presence into hospitals. Indeed, they create a new structure named Mission nationale d Expertise et d Audit Hospitalier (MeaH) in 2003, which is in charge of helping hospitals in facing these new organisational challenges 2. The MeaH reminds of the agencies that de new public management promotes (Belorgey, 2009). Actually, it is a small, technical and autonomous structure. Headed by a hospital manager, it employs about fifteen people only. Attached to the Ministry of Health, the MeaH has the responsibility of auditing hospitals, which request it and of drafting organisational standards. More precisely, it organizes investigations around specific managerial problems (beds management, cost accounting, emergency department) into a few hospitals with support of consulting firms and some researchers from the Centre of Management Science. Then, the MeaH draws from these audits general organizational standards available to all hospitals. The approach of the audits is based on the principle of co-construction of organizational norms by hospital agents and the MeaH members so as to make their 1 Ordonnance du 2 mai 2005 simplifiant le régime juridique des établissements de santé 2 Loi du 20 décembre 2002 de financement de la sécurité sociale pour

15 implementation easier. This principle can have been described as a rhetoric way for the MeaH to conceal its dependency on the health administration. The Meah would be an institutional tool that the administration would use to impose its norms (Belorgey, 2010). In fact, the relations between the MeaH and the administration are looser. The Direction de l Hospitalisation et de l Organisation des Soins (DHOS), which substitutes for the Direction des Hôpitaux from 2000 onwards, blames the MeaH for acting independently 1 and criticizes its working method 2. While the MeaH supports the voluntary participation of hospital teams, the health administration wants to compel all of them to participate in its audits. This deep disagreement partly explains the creation of the Agence Nationale d Appui à la Performance des établissements de santé (ANAP, national agency for support of the hospital performance) in The ANAP is formed of the MeaH and two other organizations in charge of the hospital information system and hospital investment. Although the ANAP looks like the MeaH in its functioning (autonomy, involvement of consultants), it strongly strengthens the device of state penetration into the hospital black box. Conclusion What empirical and theoretical lessons can be drawn from this quick history of the measuring devices of hospital activity? Five concluding points can be highlighted: 1. Empirically, policies of quantification in the hospital sector and the state managerial rationality do not appear in the 1980s-1990s. Although they largely fail, they can be, indeed, spotted at the end of the 1960s and in the 1970s, long before what is called the neoliberal turning point at the beginning of the 1980s (Jobert, Théret, 1994). This remark is reinforced by the studies demonstrating the emergence of a managerial rationality inside the French state, called state self concern, from the 1960s onwards (Bezes, 2009). This rationality is embodied in a reform plan 1 RMH, Collection of the DHOS, , Courriel de J. Castex, directeur de la DHOS, à E. Beau, directrice de la MeaH 2 RMH, Collection of the DHOS, , Note d E. Beau pour M. le directeur de la DHOS, relative au déploiement des bonnes pratiques organisationnelles identifiées par la MeaH 3 Loi du 21 juillet 2009 portant réforme de l hôpital et relative aux patients, à la santé et aux territoires 15

16 named Rationalisation des Choix Budgétaires launched in 1968, which directly impacts the Ministry of Health (Benamouzig, 2005). 2. The various measuring devices of hospital activity, which have followed one another for forty years, do not measure the same thing. While the measuring device in the 1970s tries to determine the right costs of hospital activity, the PMSI mainly aims at defining the clinical product. The devices can even contradict each other. One of the PMSI by-effects is, actually, the deterioration in the hospital cost accounting, which is the basis of the measuring device in the 1970s but is not necessary for the PMSI implementation. It confirms the idea that instruments are not neutral and can produce specific effects, which are not linked to their initial purposes (Lascoumes, Le Galès, 2007). 3. Examining the measuring devices gives information about the state conceptions of the regulation and the state capacities. The analysis of the measuring tools of the 1970s reveals the weakness of the health administration, which suffers from its technical and institutional dependency. Conversely, the examination of the measuring device in the 1990s-2000s demonstrates the expansion of the health state and the reinforcement of the state authorities into the hospital system. The relations between state and measuring devices have to be thought in a reciprocal way. Whereas strong state capacities can be necessary to impose and develop a tool, a measuring device can also strongly reinforce the state position. 4. More precisely, the managerial state uses the measuring tools in an ambiguous way. The PMSI enables the implementation of market mechanisms into the regulation of the hospital sector through the prospective payment system. However, this funding system results in reinforcing the state position since it controls more strictly the hospital funds and it tries to supervise directly the organisational changes that the funding system prompts. Thus, this analysis confirms the idea that marketization and state strengthening are not opposed (Hassenteufel et al., 2000 ; Barbier, Théret, 2009). 5. Lastly, the paper questions the linear historical interpretation of the measuring tools, which considers the PMSI as the first measuring device and asserts that it 16

17 intended to implement a prospective payment system from the start. Actually, measuring devices come before the PMSI, which is challenged by other tools for a while. Besides, the trajectory of the PMSI can be explained by a phenomenon of path dependency (Pierson, 1993). The new payment system based on it can particularly be analyzed as a functional mechanism producing reinforcement of the measuring device (Mahoney, 2000). The central role of the measuring devices in the regulation of the hospital system raises the problem of activities, which cannot be measured and then cannot be financially valued, such as the public utility activities. Protecting these activities requires thus other ways of evaluating the hospital activity, which are released from the demand for quantification. References Barbier Jean-Claude, Théret Bruno, Le système français de protection sociale, Paris, La Découverte, 2009 Bardet Fabrice, Jany-Catrice Florence, Les politiques de quantification. Introduction au dossier, Revue française de socio-économie, n 5, 2010/1, p Baudot Pierre-Yves, L incertitude des instruments. L informatique administrative et le changement dans l action publique ( ), Revue française de science politique, vol. 61, 2011/1, p Belorgey Nicolas, L hôpital sous pression. Enquête sur le «nouveau management public», Paris, La Découverte, 2010 Belorgey Nicolas, Réformer l hôpital, soigner les patients. Une sociologie ethnographique du nouveau management public, Thèse de sociologie, École des Hautes Études en Sciences Sociales, 2009 Benamouzig Daniel, Besançon Julien, Administrer un monde incertain : les nouvelles bureaucraties techniques. Le cas des agences sanitaires en France, Sociologie du travail, vol. 47, 2005, p Benamouzig Daniel, La santé au miroir de l économie. Une histoire de l économie de la santé en France, Paris, Presses Universitaires de France, 2005 Bezes Philippe, Réinventer l État. Les réformes de l administration française ( ), Paris, Presses Universitaires de France, 2009 Déloye Yves, Sociologie historique du politique, Paris, La Découverte,

18 Desrosières Alain, La politique des grands nombres. Histoire de la raison statistique, Paris, La Découverte, 1993 Domin Jean-Paul, La nouvelle gouvernance sauvera-t-elle les hôpitaux publics?, Mouvements, n 32, mars-avril 2004, p Engel François, Kletz Frédéric, Moisdon Jean-Claude, Tonneau Dominique, La démarche gestionnaire à l hôpital. Le PMSI. 2. La régulation du système hospitalier, Paris, Seli Arslan, 2000 Foucault Michel, Naissance de la biopolitique. Cours au Collège de France , Paris, Seuil/Gallimard, 2004 Hassenteufel Patrick et al., La libéralisation des systèmes de protection maladie européens. Convergence, européanisation et adaptations nationales, Politique européenne, n 2, 2001/1, p Jobert Bruno, Théret Bruno, France : la consécration républicaine du néo-libéralisme, Jobert Bruno (dir.), Le tournant néo-libéral en Europe, Paris, L Harmattan, 1994, p Laborier Pascale, Trom Danny (dir.), Historicité de l action publique, Paris, Presses Universitaires de France, 2003 Lascoumes Pierre, Le Galès Patrick (dir.), Gouverner par les instruments, Paris, Presses de Sciences Po, 2004 Lascoumes Pierre, Le Galès Patrick, Introduction : Understanding Public Policy through Its Instruments. From the Nature of Instruments to the Sociology of Public Policy Instrumentation, Governance, Vol. 20, No. 1, January 2007, pp Lenay Olivier, Moisdon Jean-Claude, Du système d information médicalisée à la tarification à l activité. Trajectoire d un instrument de gestion du système hospitalier, Revue française de gestion, n 146, 2003/5, p Mahoney James, Path Dependence in Historical Sociology, Theory and Society, Vol. 29, No. 4, August 2000, pp Mayere Anne, Grosjean Sylvie, Bonneville Luc, Rationalisation des organisations hospitalières : des incitations économiques aux injonctions à collaborer, Sciences de la société, n 76, février 2009, p Mintzberg Henry, Structure et dynamique des organisations, Paris, Les éditions d organisation, 1982 Moisdon Jean-Claude, Quelle est la valeur de ton point ISA? Nouveaux outils de gestion et de régulation dans le système hospitalier français, Sociologie du travail, vol. 42, 2000, p Moisdon Jean-Claude, Tonneau Dominique, La démarche gestionnaire à l hôpital. 1. Recherche sur la gestion interne, Paris, Éditions Seli Arslan, 1999 Mougeot Michel, Systèmes de santé et concurrence, Paris, Economica, 1994 Ogien Albert, L esprit gestionnaire. Une analyse de l air du temps, Paris, Editions de l EHESS,

19 Ogien Albert, L hôpital saisi par la quantification. Une analyse de l usage gestionnaire de la notion de qualité, Sciences de la société, n 76, février 2009, p Palier Bruno, Gouverner la sécurité sociale. Les réformes du système français de protection sociale depuis 1945, Paris, Presses Universitaires de France, 2005 Pierru Frédéric, Hippocrate malade de ses réformes, Broissieux, Editions du Croquant, 2007 Pierru Frédéric, La santé au régime néolibéral, Savoir/Agir, n 5, septembre 2008, p Pierru Frédéric, Un instrument de gestion controversé : la tarification à l activité à l hôpital, Regards sur l actualité : la réforme de l hôpital, Paris, La Documentation française, n 352, juillet 2009, p Pierson Paul, When Effect Becomes Cause : Policy Feedback and Political Change, World Politics, Vol. 45, No. 4, July 1993, pp Savoir/Agir, Dossier: économiser la santé, Savoir/Agir, n 5, septembre 2008 Serré Marina, La santé en comptes. La mise en forme statistique de la santé, Politix, n 46, 1999, p Serré Marina, Le «tournant néo-libéral» de la santé? Les réformes de la protection maladie dans les années 1990 ou l acclimatation d un référentiel de marché, Thèse de science politique, Université Paris I, 2001 Tsebelis George, Veto Players: How Political Institutions Work, Princeton, Princeton University Press, 2002 Valat Bruno, Histoire de la sécurité sociale ( ). L Etat, l institution et la santé, Paris, Economica, 2001 White Joseph, Capacity and Authority : comments on governing doctors and health care, Health Economics, Policy and Law, No. 4, 2009, pp