BMJ No 7113 Volume 315
Education and debate Saturday 11 October 1997
An attempt to save money by using mandatory practice guidelines in France
Isabelle Durand-Zaleski, Cyrille Colin, Claudine
Blum-Boisgard
Two thirds of French doctors are in private practice and are paid
on a fee for service basis. The French social security administration
provides medical cover for 99.6% of the population, and 80% of the
fee the patient pays to a private physician is reimbursed by social
security. Private insurance companies reimburse the remaining portion
for most people. The social security administration and the doctors'
unions have negotiated contractual medical fees and these are paid to
most general practitioners and to 60% of specialists. Basic fees per
consultation in 1997 were F110 (£11) for a general practitioner and
F150 (£15) for a specialist but additional fees could be claimed for
medical or surgical procedures performed during the consultation. Thus,
physicians receive most of their income from social
security.(2)
The French healthcare system has historically provided freedom of
choice for patients and doctors. Patients can see any general
practitioner or specialist they choose, with no limit to the number of
doctors seen or the frequency of visits. Doctors have been free to
request any investigations or procedures and have prescribed as they
pleased - with the exception of a few drugs restricted to hospital use.
This combination of freedom of choice, the high proportion of medical
expenditure covered by social security, and the higher costs associated
with medical advances inevitably resulted in increased expenditure.
Traditional cost containment measures, including capping total hospital
expenditure, reviewing prescribing practices, and asking patients to
contribute to costs, combined with continuing medical education and
consensus conferences,(3-6) have not controlled the
increases in expenditure in France, in particular those of medical care
outside hospital.(2)
A containment policy for healthcare expenditure became law
in August 1993 (Loi Teulade 93-8). This introduced mandatory medical
practice guidelines, known as références médicales opposables.
These form a negotiated contract signed by the social security
administration and unions representing doctors in private practice
outside hospital. The guidelines aimed to limit the prescription of
redundant and costly drugs, tests, and procedures by fining physicians
who overprescribed and to improve the quality of care.
Redundant prescribing meant either that a harmless prescription was not
required, given the patient's condition, or that the benefit did not
justify the risk of harmful side effects. The costliness of a
prescription was considered from the viewpoint of the payer. The
guidelines attached two indices to each item, one for redundancy or
harm and the other for direct cost.(7-9)
The first guidelines, published in 1994, were produced solely by the
health department of the social security administration. New
guidelines were developed in 1995 and 1996, using a scientifically
va
Selecting topics Reviewing the evidence Implementing the guidelines The guidelines were applied immediately after their publication and the
enforcement procedures after a two month observation period. The number
of violations per doctor was determined by doctors from the health
department of the social security administration, who sampled
prescriptions over two months.
Each fine, varying from F1562 to F11 250 (£156-£1125), was
determined by a weighted combination of the indices of redundancy or
harm and cost and of the total number of violations. The number of
violations by each doctor was estimated for the doctor's entire
patient population (roughly 4,500 consultations per year). A threshold
for the minimum number of violations needed for legal action against a
doctor was established for each guideline. The lowest threshold before
a lawsuit (three violations in two months) concerned prescriptions of
non-steroidal anti-inflammatory drugs and sulphonamides. The highest
threshold (13 violations) concerned requests for investigations: full
blood investigations, ultrasound performed more than three times during
a normal pregnancy, repeated thyroid tests in the absence of clinical
symptoms or signs, repeated electrocardiography in patients with
moderate hypertension, repeated blood cholesterol and triglyceride
determinations in patients with no risk factors or in those on
cholesterol lowering treatment whose cholesterol concentrations were
stable, and preoperative tests (blood work, electrocardiograms, and
chest x rays).
Compliance Expenditure In 1995, data were drawn from a sample of 2,300 doctors participating in
a four year survey (1992-5) of prescribing practice and included the
name, type, and dosage of all drugs prescribed as well as
diagnostic codes (international classification of diseases, 9th
revision).(19) The annual total of prescriptions collected
varied between 154,000 and 218,000 yearly. The number of drugs
prescribed per doctor, the name of the drug, and costs were determined
for the drugs c Cost shifting Quality of care The main principle behind the guidelines was that reduction in
prescribing would cut costs while maintaining the same level of
effectiveness of patient care, a concept that needed the support of
strong scientific evidence to make it acceptable to the medical
profession and the public.(20-21) None the less,
enforcement was also considered necessary because doctors did not seem
to have sufficient trust in guidelines or had incentives to disregard
them.(21-22)
Legitimacy A survey commissioned by the social security administration found that
doctors considered that guidelines could be useful in explaining to
patients why tests or drugs could not be given and to counterbalance
any pressure from patients and their families. Medical unions expressed
the discontent among doctors in private practice, but doctors who
protested against the reform had little leverage, as they were paid by
the social security administration.
Difficulties and costs of enforcement Expansion and revision
For all three guidelines the index of redundancy or
harm was 0.5 and the cost index was 1.25. This meant that doctors who
violated guidelines 9-16 times (for their entire patient population) in
a two month period would be liable to pay a maximum of £400; for 17-24
violations the fine would be £800; and for were more than 24
violations it would be £1200.
Treatment for Helicobacter pylori was not mentioned;
this was included in the recommendations of the October 1995 consensus
conference held in Paris. Will it work elsewhere?
We are indebted to Professor A Durocher and to Dr H
Maisonneuve, ANDEM, as well as to Dr C Sermet and Dr P Le Fur from
CREDES for their helpful comments on the manuscript. We also
acknowledge the contributions of the anonymous reviewers.
(Accepted 25 April 1997)
Department of Public Health, Claude Bernard University,
Correspondence to: Dr Isabelle Durand-Zaleski
HITEQ,
email: zaleski@lmm.jussieu.fr
References
1 Service des Statistiques, des Etudes et des SystÍ4mes
d'Information. Annuaire des statistiques sanitaires et
sociales. Paris: Documentation FranÍcaise, 1996.
2 Fielding J, Lancry P J. Lessons from France.
JAMA 1993;270:748-56.
3 Kanouse D, Jacoby I. When does information change
practitioners' behavior? Intl J Technol Assess Health
Care 1988;4:27-33.
4 Lomas J, Anderson G, Domnic-Pierre K, Vayda E, Enkin M, Hannah
W. Do practice guidelines guide practice? N Engl J Med
1989;312:1306-11.
5 Lomas J. Words without action? The production dissemination,
and impact of consensus recommendations. Ann Rev Public
Health 1991;12:41-65.
6 Davis D A, Thomson M A, Oxman A D, Haynes R B. Evidence for the
effectiveness of CME: a review of 50 randomized controlled trials.
JAMA 1992;268:1111-7.
7 Avenant noSD1 à la convention nationale destinée à
organiser les rapports entre les médecins libéraux et les caisses
d'Assurance Maladie. Paris: Caisse Nationale d'Assurance Maladie,
1994.
8 Références médicales opposables: le guide pratique
d'utilisation. Journal Officiel de la République
franÍcaise, 1995 March 5.
9 Guide des références médicales opposables.
Panorama Médecin 1994;S 3934-78
10 Agene nationale pour le Developpement de l'Evaluation
Medicale. Recommandations et références médicales.
Paris: ANDEM, 1995.
11 Woolf S. Practice guidelines: a new reality in medicine.
Methods for developing guidelines. Arch Intern Med
1992;152:946-52.
12 Chalmers T C, Smith H Jr, Blackburn B. A method for assessing
the quality of a randomized controlled trial. Control Clin
Trials 1981;2:31-49.
13 American Medical Association. Attributes to guide the
development of practice parameters. Chicago: AMA, 1990.
14 Evidence-based medicine working group. Evidence-based medicine.
A new approach to teaching the practice of medicine.
JAMA 1992;268:2420-5.
15 Department of Clinical Epidemiology and Biostatistics,
MacMaster University. How to read clinical journals. 5. To distinguish
useful from useless or even harmful therapy. Can Med Assoc
J 1981;124:1156-62.
16 Oxman A D, Guyatt G H. Guidelines for reading literature reviews.
Can Med Assoc J 1988;138:697-703.
17 Benett K J, Sackett D L, Haynes R B. A controlled trial of
teaching critical appraisal of the clinical literature to medical
students. JAMA 1987;257:2451-4
18 Lalardrie S. Rôle des caisses d'assurance maladie dans
l'élaboration et l'application des références médicales
opposables. Arch Ped 1996;3(suppl):261-2.
19 Le Fur P, Sermet C. Les références médicales
opposables, impact sur la prescription pharmaceutique. Paris:
Centre d'Etude et de Documentation en Ec
20 Woolf S. Practice guidelines: a new reality in medicine.
Arch Intern Med 1993;153:2646-55.
21 Haines A, Feder G. Guidance on guidelines. Writing them is
easier than making them work. BMJ 1992;305:785-6.
22 Ellrodt A G, Conner L, Reidinger M, Weingarten S. Measuring and
improving physician compliance with clinical practice guidelines.
Ann Intern Med 1995;122:277-82.
In the five years up to 1996, expenditure on
healthcare in France increased at an average yearly rate of 4-5%; by
1996 it amounted to about 10% of the gross domestic
product.(1-2) Various cost containment programmes have been
proposed and implemented, many assuming that high costs are a result of
unnecessary tests and treatments. We describe and make a preliminary
assessment of the latest of these, introduced in France from 1994
onwards, which combines mandatory practice guidelines on procedures and
drug prescribing with a system of fines for doctors who do not
comply. 
French doctors who do not follow prescribing guidelines can be fined.
Summary points Mandatory practice guidelines were introduced from 1994 and 1995
as a way of cost containment and of standardising patient care The system was developed by an independent agency after 1994 and agreed
by representatives of the profession, the French social security
administration, and the government Practice guidelines apply to medicine and surgery, diagnosis and
treatment but hospital practice is excluded Doctors who do not comply can be fined up to £2000; in 1996, 186 were
investigated and 75 were fined
The rate of increase of expenditure on medical services outside
hospital has decreased since the system was introduced Limitations include the lack of outcome assessment, possible shifts in
expenditure from outpatient to inpatient services, and the fact that
enforcement is difficult and costly Background
Medical practice guidelines
Topics selected for mandatory practice guidelines 1994 guidelines:* 1995 guidelines:Ý 1996 guidelines: Prescription of non-steroidal anti-inflammatory drugs
Prescription of
antibiotics
Diagnostic imaging of back pain
Systematic blood
and urine testing
Testing during pregnancy
Treatment of
non-insulin dependent diabetes
Digestive endoscopy
Thyroid
function tests
Treatment of hypertension
Treatment of
hypercholesterolaemia
Magnesium blood and serum
concentrations
Preoperative tests
Mammography screening
UlcerCholecystectomy
Hysterectomy
Knee surgery and
exploration
Lower back pain
Prostate cancer (non-invasive)
Prostate adenoma (benign)
Carotid endarterectomy
Coronary
bypass surgery
Ocular implants
Laser in ophthalmology
Tympanostomy tube insertion
Surgery for deafness
Dental and
maxillary anomalies
Cervical smear screening for uterine
cancer
Diagnosis of pregnancy using chorionic gonadotrophic
hormone
Electroencephalography
Electromyography
Pulmonary
function tests
Exploration of gastro-oesophageal reflux in
neonates and children
Management of psychotic patients
Hypnotic
and tranquilising drugs
Prescription of neuroleptic drugs
Immunohistochemistry tests in pathology
Acne
Vasoactive drugs
in the treatment of arterial ischaemia
Skin cancerLipid lowering
treatment
Chronic venous insufficiency of the lower limbs
Invasive techniques for coronary artery diseases
Appendectomy
Treatment of gastric and duodenal ulcer
Diagnosis of viral hepatitis
Physiotherapy
Urinary infection
Current diagnoses
in haemotology
Hysterectomy
Sterility
Asthma
Long term oxygen therapy for chronic respiratory insufficiency
Antidepressant drugs*Developed by social security administration and
doctors' unions.
ÝDeveloped by Association Nationale pour le
Developpement de l'Evaluation Medicale, social security, and doctors'
unions.Methodology
The social security administration and representatives of
doctors' unions had the sole responsibility for selecting topics. The
criteria set for selecting topics were medical and economic, and
included high cost, high prevalence, high risk, and high variation in
practice patterns.
Practice guidelines were established by ANDEM with reference to
published work.(12-17) Medical societies, university
physicians, and experts were charged with reviewing and editing the
final wording. The social security administration then selected those
that were most feasible and relevant to medical practice outside
hospital.
The first group of guidelines was published as bylaws in
March 1994, six moths after the Loi Teulade. The indices of medical
redundancy or harm and of cost were published together with fines for
non-compliance as part of the bylaw. The index value of redundancy/harm
could be 0.5 (redundant but no iatrogenic risk), 1, or 1.5
(presence of appreciable iatrogenic risk). The index of costliness
could be 1, 1.25, and 1.5, coEvaluation
Doctors were asked to indicate on every prescription form whether
or not the item was covered by a guideline. A total of 13,000 doctors
(roughly 10%) was surveyed over two years. Altogether 1,278 were peer
reviewed, and proceedings were taken out against 186; 75 were
eventually fined.(18)
The average increase in expenditure for health services
outside hospital was 2.3% per year in 1994 and 1995 compared with 6%
previously. Total pharmacy expenditure rose by 1.3%, compared with
7.4% in 1993, and the total volume of tests requested fell by 15%
(tests had increased by 1% in 1993). Volumes of medical and surgical
procedures were not studied. Expenditure actually changed in 1993, when
the decision about forthcoming medical practice guidelines was
announced. We believe that doctors anticipated the controls and
voluntarily limited their prescriptions.
These guidelines do not apply to hospital practice, which means
that the current system does not address the issue of continuity of
care; inpatient and outpatient budgets and practices are clearly
separated. The increase in hospital expenditure was reduced by
government decree from 6% to 4% in 1994 and to 3% in 1995. There may
have been "cost shifting" to hospital services after the
implementation of the guidelines but the current information system
does not permit cross checking between patients' data and resource
use. However, because hospital expenditure was capped there was no
change in the trend after the guidelines were
introduced.(1)
No means of measuring the effects of the guidelines on patients'
outcome and satisfaction were planned. Laws (ordonnances) signed in
April 1996 mean, however, that healthcare providers must present data
on quality of care; a yearly regional assessment of the state of health
is to be implemented via medical surveys and questionnaires.
Issues for debate
The legitimacy practice guidelines being issued by the body that
is paying for medical care could be questioned. The first guidelines in
1994 were criticised because of methodological flaws and possible
conflicts of interest. The second group, issued in 1995 and 1996, were
written by panels of experts in each field and appeared first as
scientific recommendations. The procedures that rendered the guidelines
mandatory were the sole responsibility of the social security
administration and doctors' unions. In total 147 practice guidelines
have been published, 90 of them derived from the clinical practice
guidelines published by ANDEM.
As French law forbids collecting data on health that could be
traced back to an individual, details of tests and drugs may not be
entered on computerised patients' records. All checks on the
implementation of the guidelines had to be performed manually on a
sample of prescriptions written by doctors being surveyed. All
reimbursement claims sent to the social security archives are matched
against the original p
The rapid change in medical knowledge and technology limits the
credibility and medical relevance of the guidelines and the validity of
enforcing their use. For example, the 1994 guideline for ulcer
treatment (based on 1993 scientific data) had to be revised after the
importance of treatment for H pylori was acknowledged in
1995. Published guidelines need continuous revision; each year new
guidelines addressing 20-30 new topics should be
commissioned.
1994 guidelines on ulcer treatment "Ulcer treatment" was covered by three practice guidelines
on drugs for and duration of treatment (all categories of
anti-ulcer drugs were allowed but all antacids were
excluded). There are no grounds for:
The extension of mandatory guidelines to hospital medicine is
under study; most guidelines issued in 1995 were considered suitable
for hospital practice. Government policy was the route chosen for
Implementing change in physicians' behaviour because of the
centralised financing system in France. Replication in other countries
with centralised financing could be possible; in some countries the
regional level might be more appropriate. Only a centralised healthcare
system with relatively weak medical unions could introduce such a
system since guidelines have to be universal and the paymaster has to
have both leverage over doctors and government support. This may apply
to some European countries, but is more of a political than a medical
issue.
Henri Mondor Hospital,
Paris,
France
Isabelle Durand-Zaleski,
assistant
professor
Lyon I,
Lyon,
France
Cyrille Colin,
assistant
professor
Claudine
Blum-Boisgard,
professor
Henri Mondor Hospital,
51 Avenue
du Maréchal de Lattre deTassigny,
94010 Créteil,
France
Home | Current issue | Past issues |
Classified ads | Career Focus | Feedback
Collections |
About this site | About the BMJ | BMA | Medline