Background

As in many industrialised countries, the health care quality movement began in France because of rising health expenditure and the necessity to contain costs¹ but, recently, the public has become more aware of issues relating to quality. Serious public health problems such as the contaminated blood scandal of 1984, when blood that was strongly suspected of being contaminated by HIV was knowingly transfused to haemophiliac patients, and greater visibility of routine medical practice through regular publications in the lay press² has led to a crisis of public confidence in the ethics of the medical and political worlds and a strong demand for accountability and greater transparency. Important reforms in the organisation of health care and public health have therefore been undertaken, of which accreditation is one. Its objectives reflect this historical background—namely, “to assess the quality and safety of health care, to assess a health care organisation's ability to ensure continuous improvement in the quality of overall patient care, to formulate explicit recommendations, to involve professionals at all stages of the quality initiatives, to provide external recognition of the quality of care in health care organisations, to improve public confidence”.³

Accreditation was enacted in France as part of the 1996 health care reform by ordonnance, a government decision that is taken without consulting Parliament. Governments in France under the Fifth Republic use ordonnances when they feel there is an urgent need for reform that could be delayed by parliamentary discussions. Many healthcare reforms have been enacted by ordonnance and, in 1996, it was felt that the magnitude of the deficit of the national health insurance fund and the public health situation was sufficiently serious. The ordonnance of 24 April 1996 reforming public and private hospitalisation stipulates that “in order to ensure continuous quality and safety improvement of health care, all public and private health care organisations must submit to an external evaluation procedure named accreditation” (Article 710–5). Accreditation applies to public and private hospitals and healthcare networks but does not include general practice. However, in time the fields of application for accreditation are likely to be extended.⁴

Accreditation is new in France so we are not yet able to describe the implementation of the programme, the uses that will be made of the findings of the accreditation surveys, and the changes it will eventually bring about in the management of French hospitals and the quality of healthcare services. We will therefore limit ourselves to the description of the accreditation policy in France, evaluate its similarities and differences with other accreditation programmes, and try to estimate its chances of success.

The policy

The instruments of accreditation in France consist of an agency, a manual, and surveyors.

THE ACCREDITATION AGENCY

In order to implement accreditation the 1996 ordonnance created what it defined as “an independent and professional organisation”, a national agency for accreditation and evaluation in health care, the Agence Nationale pour l'Accréditation et l'Evaluation en Santé (ANAES). The mission of the new agency is “to help develop quality assurance of medical practice, in public and private hospitals as well as in private practice, and to implement the accreditation procedure” so that it manages both quality assurance and accreditation.⁵ ANAES consists of a Board, a Scientific Council, and an Accreditation College and is headed by a General Director. Its members are appointed by the Minister of Health and it is financed one third by the Department of Health, one third by the National Health Insurance fund, and the remainder by the survey fees of the hospitals. A notable feature of ANAES is the importance of professional representation: the health professions comprise at least three quarters of the board and more than half are medical doctors. The present Chairman of the Board is a professor of neurology and the General Director of the agency is a professor of public health. Members of the Scientific Council are experts in the areas of quality assurance and accreditation, and/or belong to medical or other scientific societies. The Council is presided over by a professor of intensive care and is responsible for the scientific quality of all guidelines and other documents produced by the Agency. In particular, it supervised production of the accreditation manual. The importance of professional representation is intended to guarantee the Agency's independence and credibility.

A key feature is the Accreditation College which is responsible for examining the survey reports, attributing accreditation, defining recommendations for improvement for every hospital, and publishing an annual report. It is composed of 11 members and a similar number of deputy members. Members are hospital managers, hospital doctors, pharmacists or allied health professionals, and two medical doctors with recognised expertise in quality assurance and/or accreditation. All are experienced professionals who have been in practice for at least 15 years. They are appointed by the Minister of Health on the proposal of the Scientific Council of the ANAES, after approval by the Board.

THE ACCREDITATION MANUAL

Two specialties (psychiatry and cancer) have developed their own accreditation manuals, but there is only one official manual that is applicable to all healthcare organisations. The manual was compiled during 1998 by ANAES with the help and contribution of 150 professionals, 57 of whom were medical doctors. The working groups included nine patients' representatives. The actual writing was preceded by a literature search on the topic of accreditation, foreign experiences, and various manuals compiled privately in France, in particular the accreditation manual of the National Federation of Cancer Hospitals.⁶ Foreign manuals were also examined, including those of the Joint Commission for the Accreditation of Healthcare Organisation, the Canadian Council on Health Facilities Accreditation, the Australian Council on Health Care Standards, the King's Fund, and CASPE Research. A first version of the manual was tested in 12 hospitals, amended, and tested again before it was ready for official use in February 1999. The whole procedure was supervised by the Scientific Council of ANAES.⁷ An English version of the manual is available online on the Agency's website (http://anaes.fr).

The manual is divided into three sections (box 1) and each section is subdivided into chapters. Every set of standards begins with the definition of a general policy concerning the objective of the standard—for instance, on the organisation of patient records (DPA):

DPA standard 1: The healthcare organisation formulates and implements a patient record policy for all its activity sectors and ends with the evaluation of the level of quality achieved:

DPA standard 7: The patient record is the subject of a strategy of assessment and continuous improvement.

The standards concern management processes and procedures. Outcome standards are expected to be introduced at a later stage. The purpose of the manual is for all professionals in the hospitals to implement continuous quality improvement (CQI) by means of an explicit quality management system, and for clinicians to use practice guidelines and protocols. It responds to the main principles of CQI—that is, involving everyone in the organisation, being concerned with all the internal organisational processes, and focusing on external needs (in this instance, those of the patient).⁸

A management system concerned with all the organisational processes

In France the hospitals are divided into specialised departments or firms (services), each led by a medical chef de service who exercises complete power on its medical policies. Every service tends to pursue its own interests with no real consideration for the hospital as a whole. Departments compete for their fair share of the hospital's global budget, and the culture is more one of competing interests than of constructive solidarity.⁹ A further rift is the one which often exists between clinicians and managers whose culture, objectives, and language often differ. The hospital has therefore been compared to a “mosaic” of decentralised decision centres.¹⁰ The accreditation manual takes the radically different approach that quality is the product of the cooperation of everyone in all the management processes. It is not divided by departments but by processes: implementing patients' rights and information, managing patient records, organising patient care, managing human resources, the information system, quality and risk prevention, etc. The chapter on “organisation of patient care” (OPC) follows the patient's route through the hospital from access, admission, assessment of the patient's condition and needs, coordination of care, discharge, to quality assessment of the patient's care. It prescribes organisational processes that can ensure continuity of care by involving all concerned professionals:

OPC standard 6: Patient care is coordinated within the various clinical activity sectors.

OPC standard 7: Continuity of care is ensured.

OPC standard 8: Health professionals involved in operating theatres, other intervention sectors and clinical activity sectors work together in formulating their operational procedures.

At this stage, only two standards address quality of care:

OPC standard 14: Clinical and ancillary medical activity sectors use diagnostic and therapeutic protocols.

OPC standard 15: Clinical and ancillary medical activity sectors assess professional practices and their results.

Assessment of professional practice is to be achieved through the definition and use of performance indicators (OPC 15b) and sentinel events (OPC 15c). It is the task of every hospital to define its own.

The quality management and risk prevention section (QPR) focuses on the cross-cutting prevention of the major risks (transfusion (VST), infection (SPI)) at the hospital level:

QPR standard 1: The healthcare organisation initiates, leads and maintains a quality policy based on quality management and risk prevention.

VST standard 1: The authorities and professionals concerned are involved in drawing up and implementing the organisation's policy concerning specific prevention programmes, including transfusion safety.

SPI standard 1: The healthcare organisation establishes and operates a coordinated infection control policy among patients and professionals.

THE SURVEYORS

All are experienced and active healthcare professionals who have been trained in accreditation procedures by ANAES. Training is both initial and ongoing. Surveyors are part time and cannot spend more than one third of their working time on accreditation. In order to ensure consistency of the accreditation surveys, surveyors must follow the principles of the accreditation survey listed in an “accreditation surveyor's charter” planned to evolve as experience of the procedure is acquired. Obviously, surveyors can have no professional ties or interests of any kind with the organisation they survey.

The procedure

Initially, accreditation was a private and voluntary procedure but, as its objectives evolve towards public regulation, it is tending to become compulsory.¹¹ In France the voluntary/compulsory issue was resolved as follows. The 1996 ordonnance instituted a 5 year period (1996–2001) during which hospitals can apply to enrol for accreditation. This is the “voluntary” aspect of the procedure. However, if at the end of this 5 year period any hospital has not yet volunteered, it will be compulsorily enrolled by the local hospital authority so that no hospital, public or private, can escape a procedure that is, in fact, compulsory.

The procedure itself consists of self-assessment followed by the survey visit and a report, the purpose of which is to assess the compliance of the hospital with the standards defined in the accreditation manual. Self-assessment is considered to be the most important part of the process and a special guide was written by ANAES to help hospitals with it. Self-assessment is intended as a preparation for the survey visit during which the whole organisation assesses (or discovers) its baseline state of compliance with the standards contained in the manual. At the end of the self-assessment the hospital sends to ANAES a report in which it estimates its degree of compliance with the manual standards and defines what improvement measures it has undertaken where necessary. The actual accreditation survey follows the self-assessment by less than 3 months. Depending on the size of the surveyed organisation, the survey team is composed of at least three professionals including a doctor, a member of the allied health professions, and a manager. The survey visit lasts a varying number of days. The surveyed organisation must be able to answer the question “What do you do in order to comply with ... the concerned standard?” The hospital must then be able to produce the documents proving that they have the corresponding policy. Where any deficiencies have been noted during the self-assessment, the organisation must be able to demonstrate what improvement measures they have undertaken to correct them. Surveyors can also interview professionals and patients to verify the answers they have received.

After the survey visit is completed the surveyors write an expert report in which they compare their own conclusions with those made by the hospital itself after the self-assessment. This is sent to the Accreditation College which examines both reports and then writes an accreditation report in which it attributes an accreditation level, with or without reservations, and eventually decides on recommendations for improvement. The College also considers the methods used for self-assessment and for the survey visit. A summary of the accreditation report is made available to the public. At the time of writing about 200 hospitals are in the process of self-assessment and 10 have been accredited. All completed survey reports are accessible on the ANAES website.

A CONTINUOUS PROCESS

This is an important feature of the process. Hospitals are being made to understand that accreditation is not a one time procedure and that it is not obtained once and for all. On the contrary, it is presented as a means of ensuring a continuous improvement process within the hospitals, showing changes from the baseline quality estimated by self-assessment before the survey visit and the progress as surveys are repeated at least every 5 years, thus ensuring the continuity of the improvement in quality. Compliance with the standards will be assessed by the College of Accreditation on the basis of the survey report. No one will fail, at least this first time, but for hospitals with recommendations the next accreditation level will depend on their implementation of the recommendations. Survey visits will normally be repeated every 5 years but, for hospitals with reservations, a survey visit focused on deficient areas may take place at an earlier date as decided by the College. A cyclical process is thus being installed, instituting a systematic and prospective risk prevention policy in hospital management which constitutes a fundamental departure from the retrospective crisis management of quality problems that has prevailed up until now in the healthcare system in France.

THE CENTRAL IMPORTANCE OF THE PATIENT

An important objective of the current French health policy is to refocus health care on its ultimate object—that is, the patient—rather than the process of care per se. Explicit concern for the welfare of patients has been the object of many legal texts in recent years. In particular, in 1988 an important law defined the conditions of the protection of patients in medical research.¹² A Charter of the rights of the hospitalised patient was issued in 1995.¹³ Symbolically, the 1996 ordonnance's first title is “Patients' rights”. It contains a whole array of measures aimed at providing greater attention to the rights and needs of patients, and ensuring a greater participation of the patient in the life of the hospital; the patient's Charter must be made public and included in the information booklet that every patient is given on admission. Professionals must recieve specific training in patients' rights and confidentiality. Access for deprived populations and immigrants must be organised specifically.¹⁴ Representatives of patients now sit on every hospital board and on hospital committees such as the infection committee (Comité de Lutte contre les Infections Nosocomiales, CLIN). Every hospital must organise a commission de conciliation where conflicts between patients and the hospital may tentatively be solved amicably before reaching the legal stage. Furthermore, although not an explicit reference of the manual, the assessment of patient satisfaction is an important issue in the accreditation procedure. The ordonnance requires that every hospital should “proceed to regularly assess their patients' satisfaction”. The results of these assessments will count for accreditation.¹⁵ Patients' rights and information are the subject of the first chapter in the manual (DIP) and management of patients' complaints is the subject of a specific standard (DIP standard 8).

THE INVOLVEMENT OF ALL PROFESSIONALS

Healthcare policy in France is characterised by the strong influence of the State and the weak bargaining power of a divided medical profession.¹⁶ Most major reforms, such as national health insurance, were introduced unilaterally by the government against the opposition of the medical profession. Although the accreditation policy was introduced in the same authoritative fashion, and despite the fact that the process is decidedly public and government owned, an essential point is the involvement of professionals. We have described the importance of professional involvement in the composition of ANAES. A similar multiprofessional involvement is expected at the hospital level. This can sometimes prove more difficult where doctors are concerned; indeed, doctors have often been the missing link in healthcare quality assurance systems because their definition of quality in medicine differs from the one adopted by regulators and managers. For doctors, improvement in medical quality consists in the accomplishment of medical progress through clinical research. The majority have not participated up to now in the managerial culture of audit and accreditation. In France, as elsewhere, quality assurance as medical audit is viewed by many as an intrusion and a waste of time. The difficulty is enhanced in France by the fact that the concepts used are all of Anglo-Saxon origin and do not always find easy translation into French and acceptance in the French culture.¹⁷ The fact that concepts relating to quality assurance and management are not integrated in the medical curriculum contributes to doctors' disinterest. One reason why the structures of ANAES include so many professionals, apart from their necessary technical contribution, is the importance of obtaining professional legitimacy and credibility. However, at the local level it often remains difficult to obtain medical participation in meetings, and the very time consuming work for the self-assessment prior to the survey visit is mostly carried out by nursing and management personnel.

Discussion

Does accreditation in France respond to the definition given by the ExPeRT project of an external quality mechanism: “a regional or (potentially) national process voluntarily entered by service provider organisations for the improvement of organisation and delivery of health services assessed against explicit, published standards by peer group teams moderated by a non-partisan authority involving (but impartial to) users, providers, purchasers, and government”¹⁸? In France accreditation is a national process which healthcare providers enter with the view of improving the management and delivery of healthcare services in their organisation by having them assessed by peer group teams against explicit published standards and moderated by an authority that involves government (in France the main purchaser), providers and, to a lesser degree, users. This process, however, is not voluntary but is made compulsory by law. Is the moderating authority non-partisan and is it equally impartial to government, providers, and users? How does accreditation take its place in the quality movement in France? Does it have a chance of succeeding in improving the organisation and delivery of healthcare services in France?

The answer to these questions probably lies in the answer to the question: “Who wants to influence whom to achieve what?”¹⁹ The history of healthcare quality assurance in developed countries can be roughly divided into three periods.²⁰ The first took place in the USA around 1917 when, following the initiative of E A Codman, surgeons decided that they would not operate in hospitals that did not provide them with a minimum standard of quality in their working conditions. Eventually this initiative developed into accreditation. It is important to point out that this first period of quality assurance was (a) the result of a professional initiative, private and voluntary, and (b) mainly structure orientated—that is, aimed at ensuring doctors with satisfactory working conditions. At that time the question of the quality of medical procedures was not an issue. It was not even a concept. Quality was “what we (doctors) do”,²¹ and the purpose of accreditation was to provide doctors with an appropriate environment to do it. In other words, doctors wanted to influence providers (hospitals) to obtain satisfactory working conditions.

When the economic crisis of the 1970s attracted the attention of Western governments to the increasing costs of health care, economists, epidemiologists, managers, and regulators (that is, the “external users” of health care²²) started examining medical practices.²³ With no evidence that health indicators improved in parallel with health expenditures, and comforted by the extent of unexplained variation that had started to be documented in doctors' practices,²³ they developed a rhetoric for introducing so called “professionally led quality assurance of medical practice”. But the “quality” argument was promptly viewed by the medical profession for what it was—namely, a method for rationalising medical prescriptions through utilisation review, consensus conferences, and practice guidelines. The focus had shifted from structure to process and, whereas the first period of quality assurance was induced by and for the medical profession (the “original users of health care”²⁴), the second one was initiated outside the medical profession and, one could say, against it: today's quality assurance questions doctors' decisions and practices. So it is not surprising that it never gained professional legitimacy in the eyes of the doctors—as opposed to other health professionals such as nurses—and that the biggest difficulty in implementing quality assurance of medical practices in most industrial countries except, perhaps, the Netherlands has been the doctors' constant indifference at best and, more often, the outright opposition to it. During this second period of quality assurance purchasers and regulators wanted to influence providers (doctors) to achieve better quality care for less money. However, in France, as in many other western countries, they largely failed because they did not succeed in getting the medical profession to feel concerned. Will accreditation succeed where quality assurance failed?

Accreditation, although authoritatively introduced in France by law, is characterised by an explicit involvement of the medical profession, both in the structures of the programme and in its implementation which consists of peer review. It is mainly concerned with management, but management that includes both the organisation of healthcare delivery that mainly concerns managers, and the more professional aspects of medical practice—such as making informed decisions about patients with the help of evidence based practice guidelines—which essentially concerns doctors. In the UK these two aspects of medical management have been united in the new concept of clinical governance.²⁵ Undoubtedly, accreditation will considerably improve prospective risk management in French hospitals and therefore will hopefully reduce the hazards of hospitalisation. But will the medical profession follow? Will its participation in accreditation go beyond a formal involvement in structures and surveys? Will they engage at the local level in new behaviours that would result in an improvement in the quality of medical care and of patient satisfaction? Such behaviour would, for instance, involve a systematic review of the quality of patient records, a systematic assessment of the organisation of every patient stay, and an explicit policy for patient information in every service. In an ideal world doctors would want to influence themselves or each other in order to improve the quality of everyday practices as well as their participation in medical research and publications. Does accreditation contain the necessary incentives?

At this stage of the procedure it is not known what the public authorities intend to do with the results of accreditation in France. If the official theory is that accreditation is mainly destined to encourage quality monitoring systems in hospitals, a common fear is that unsatisfactory scores would encourage authorities to close down hospitals in these times of restructuring of the healthcare system (France supposedly has about 60 000 hospital beds in excess) so that, if the process today is largely professionally driven, the continuity of professional participation may depend on the nature of the utilisation of its results by government. It will also depend on the evolution of the content of the standards: will they remain strongly management orientated, in which case the doctors may not feel really involved (only two standards explicitly concern quality of care today), or will they evolve to include visitatie like procedures, thereby explicitly including the quality of medical practices in the evaluation of performance?¹⁸ Given the political system in France, the main users of accreditation are likely to be, firstly, the government and, secondly, if all goes well, the users. Whither the provider?