 |
|
Despite Financial Penalties, French Physicians' Knowledge of Regulatory Practice Guidelines Is Poor
Pierre Durieux, MD, MPH;
Bertrand Gaillac, MD;
Bruno Giraudeau, PhD;
Michel Doumenc, MD;
Philippe Ravaud, MD, PhD
Arch Fam Med. 2000;9:414-418.
ABSTRACT
| |
Objective To evaluate the level of awareness and knowledge of regulatory practice guidelines (références médicales opposables [RMOs] or regulatory medical references) implemented to control ambulatory care costs among French family physicians.
Design Observational study. Participants were asked to identify RMO topics among a list of actual and fictitious RMO topics and the RMOs themselves among a list of actual and fictitious RMOs.
Setting General practice in France.
Subjects Three hundred twenty-one family physicians.
Main Outcome Measure Average score of 100 (95% confidence interval [CI]) on the awareness of RMO topics and knowledge of the RMOs.
Results The average overall score was 55.8 of 100 (95% CI, 53.3-58.3) for the awareness of the RMO topics and 50.5 (95% CI, 48.3-52.7) for knowledge of the RMOs themselves—53.2 (95% CI, 51.1-55.3) for diagnostic RMOs and 47.8 (95% CI, 45.6-50.0) for therapeutic RMOs. Chance would have yielded an expected mean score of 50. A statistically significant difference was noted between the average score for actual (62.2) and fictitious (43.2) RMOs, P<.001. None of the respondents correctly identified all 24 correct answers.
Conclusion Despite implementation of RMO policy, the awareness and knowledge of RMOs among French family physicians seem weak. The number of RMOs and the difficulties in controlling physicians probably explain these results. Thus, it is doubtful that the RMO policy will have a long-term effect on physicians' behavior.
INTRODUCTION
IN AN EFFORT TO control ambulatory care costs, regulatory practice guidelines known as références médicales opposables (RMOs) or regulatory medical references were introduced by law in France in 1993.1 Regulatory medical references are defined as "recognized scientific criteria that make it possible to define inappropriate care and prescriptions, and the frequency with which such care or prescriptions are received by the patient.''2(p6349) Regulatory medical references are presented in the form of clearly stated, short, prescriptive recommendations, negatively formulated (eg, "it is inappropriate to treat systemic hypertension before having measured blood pressure 3 times over a two-month period''). Regulatory medical references cover medical and surgical topics as well as diagnostic and treatment procedures. Regulatory medical references are selected by representatives of French insurance funds and physicians' unions from specific guidelines drawn up by a national publicly funded agency, Agence Nationale d'Accréditation et d'Evaluation en Santé (formerly known as the Agence Nationale pour le Développement de l'Evaluation Médicale). These guidelines, based on scientific evidence and expert opinion, are developed by panels that include family physicians and individuals from all relevant medical disciplines. Since the introduction of these RMOs, physicians (family physicians and specialists) who do not comply with RMOs can be fined. The fine is determined by a weighted combination of indexes of redundancy, harm, cost of each RMO, and the number of violations.3 A list of applicable RMOs is published every year. In 1994, the first year of implementation of the RMO policy, a total of 66 RMOs covering 15 topics was published in the Journal Officiel de la République Française (the French equivalent of the Federal Register).4 Each year, the list of applicable RMOs is revised and new RMOs about new topics are published. Thus, in March 1997, the number of RMOs applicable for family physicians reached 197, covering 46 topics (between 1 and 10 RMOs per topic).5 In addition to annual publication in the Journal Officiel de la République Française, RMOs are mailed by the major national health insurance fund (the Caisse nationale d'assurance maladie des travailleurs salariés), to the 110,000 French physicians working in private practice. They are also widely published and discussed in French professional medical journals. Thus, each family physician was exposed at least 1 time to the more recent RMOs and 3 times to the oldest RMOs. All RMOs were accepted and endorsed by family physicians' representatives (unions) before publication in the Journal Officiel de la République Française.
The RMO policy was questioned in 1997 when the reform of the French health system changed the rules.6 According to this reform, French physicians working in private practice could also be collectively fined at the end of each year if they overspent the budget prescribed by the French Parliament. On the contrary, they could receive a "bonus" if they stayed within this budget. Many physicians protested that this principle was unethical because a physician should not be rewarded for prescribing less.7-8 This reform created intense conflict among the medical unions, the French government, and the national health insurance fund.
It has been suggested that incentives or penalties could have some influence on physicians' behavior if they are linked to an explicit clinical objective (eg, to reduce inefficacious procedures).9-10 If any effect of the RMO policy on medical practice is to be expected, physicians must first be aware of the RMOs.11 Thus, we conducted a survey to evaluate the level of awareness and knowledge of RMOs among family physicians 4 years after the implementation of this policy.
MATERIALS AND METHODS
QUESTIONNAIRE
Two family physicians involved in the study (B. G. and M.D.) selected 4 RMO topics, 4 diagnostic RMOs, and 4 therapeutic RMOs from the list published in 1997. In addition, they selected a list of 4 topics not covered by the RMOs. Then, 2 of us (P.D. and B.G.) created 4 fictitious diagnostic RMOs and 4 fictitious therapeutic RMOs. Fictitious RMOs were recommended practices that would be reasonable to follow but that had not been selected as RMOs (Table 1). Both actual RMOs and fictitious RMOs concerned clinical problems faced by family physicians (eg, expected high frequency or high cost of prescription).
|
|
|
|
Table 1. Responses of 321 French Family Physicians to Actual and Fictitious RMO* Topics in the Questionnaire
|
|
|
From a list of 8 topics including 4 actual and 4 fictitious RMOs in each of 3 sections, participants were asked in yes or no questions to identify the following: RMO topics, diagnostic RMOs, and therapeutic RMOs (Table 1). Participants were unaware of the number of actual items among a total of 24 items.
Respondents were also asked to give brief details about themselves (age and sex), to state whether they received RMOs and read them, how often they consulted them at the time of patient consultation or not during patient consultation, and whether they are a member of a continuing medical education group.
SURVEY
The questionnaire was administered in June 1998 to 3 groups of practicing family physicians:
- Physicians attending a session on guideline development methodology that occurred at a conference on therapeutics in general practice (group 1, n=159).
- Members of a group of physicians who participate regularly in training activities organized by Agence Nationale d'Accréditation et d'Evaluation en Santé at a regional level (group 2, n=120).
- A random sample drawn from the list of members living in the Ile de France region of a national medical society of family physicians (the Société de Formation Thérapeutique du Généraliste) (group 3, n=200).
The questionnaire was distributed to all group 1 physicians attending the session after a short presentation about the study. In groups 2 and 3, the questionnaire was mailed with a cover letter signed by the investigators and a self-addressed stamped return envelope.
DATA ANALYSIS
The questions required yes or no answers. We first considered the 24 items as the unit of analysis to estimate for each of them a percentage of correct answers among the physicians. Then we focused on the physician as the unit of analysis, and individual scores were obtained by totaling the scores obtained by the physician for each item (1 point per correct answer). These individual scores were then divided by the number of items considered and multiplied by 100, thus leading to scores of 100 (95% confidence interval [CI]). Comparisons between groups were made by means of analysis of variance or Fisher exact tests. Comparisons of average scores to the theoretical value of 50 (expected average score obtained by chance) were made using t tests. Let us consider a series of 8 yes or no items (eg, items related to the awareness of RMO topics). Assuming the physician always answers randomly, the probability of being right (and thus obtaining 1 point) is 0.5 for each of the 8 items per section. The score obtained when answering randomly is, therefore, binomially distributed with parameters of 8 and 0.5. The expected mean score obtained by chance is then 4 of 8 or, equivalently, 50 of 100. Values are expressed as the mean (SD).
RESULTS
Three hundred fifty-two physicians answered the questionnaire. The response rates were 92.5% for group 1, 78.3% for group 2, and 55.5% for group 3. After exclusion of 31 questionnaires completed by physicians who were not practicing family physicians (retired family physicians and specialists), 321 questionnaires were analysed. Table 2 provides the main characteristics of the respondents in the 3 groups surveyed. Overall, 93.8% of the respondents declared that they received the text of RMOs. Respectively, 44.3% and 80.0% declared that they sometimes read them, either at the time of patient consultation or not during patient consultation. There were more female physicians in group 3 and more respondents who declared that they sometimes used the RMOs in group 2.
|
|
|
|
Table 2. Characteristics of French Family Physicians Completing the Questionnaire*
|
|
|
AWARENESS OF RMO TOPICS
The average overall awareness score was 55.8 of 100 points (95% CI, 53.3-58.3). Group 3 performed worse than the 2 other groups (mean score, 47.8 points compared with 59.5 for group 1 and 60.0 for group 2). The percentage of awareness per topic varied from 26.8% (95% CI, 22.0-31.6) to 88.5% (95% CI, 85.0-92.0) (Table 1). The topic "medical treatment of arteriosclerosis of peripheral arteries''(actual RMO topic) had the highest percentage of recognition in the 3 groups of physicians, and the topic "insulin-dependent diabetes'' (fictitious RMO topic) had the lowest percentage in the 3 groups.
KNOWLEDGE OF RMOs
The average overall knowledge score (16 questions) concerning RMOs themselves was 50.5 points (95% CI, 48.3-52.7). This score was 53.2 points (95% CI, 51.1-55.3) for diagnostic RMOs and 47.8 points (95% CI, 45.6-50.0) for therapeutic RMOs. Results of the 3 groups are listed in Table 3. Again, the results of group 3 were worse than those of the 2 other groups for diagnostic and therapeutic RMOs.
|
|
|
|
Table 3. Mean Score of Correct Answers per French Family Physician*
|
|
|
Only 3 diagnostic and 1 therapeutic RMOs were identified by more than 75% of the respondents. These 4 RMOs were published for the first time in 1994. They concerned the following subjects: tumor markers (76.6% correct; 95% CI, 72.0-81.2), diagnosis of osteoarthritis (84.1% correct; 95% CI, 80.1-88.1), diagnosis of non–insulin-dependent diabetes (86.3% correct; 95% CI, 82.5-90.1), and treatment of hypertension (90.0% correct; 95% CI, 86.7-93.3). Less than 25% of the correct answers was observed in 1 diagnostic and 1 therapeutic sentence: diagnosis of urinary tract infection (24.4% correct; 95% CI, 19.7-29.1) and antibiotic treatment in otitis and pharyngitis (20.6% correct; 95% CI, 16.2-25.0), respectively. These fictitious RMOs were identified as true by most of the respondents (Table 1). These results did not differ among the 3 groups surveyed.
When the 24 questions were considered together, divided into actual and fictitious RMOs, we observed a significant difference in the average (SD) scores, 62.2 points (19.5) for actual RMOs compared with 43.2 points (25.0) for fictitious RMOs (P<.001). None of the respondents correctly identified all 12 correct items. Fifteen (4.8%) of them correctly identified the 8 RMO topics, but none correctly identified the 16 items concerning the RMOs themselves. There was no difference in the average scores of physicians who indicated participation or not in a continuing medical education group (52.4 points [16.5] vs 54.2 points [16.4], P=.43) or who declared that they sometimes used the RMOs during patient consultation (52.0 points [16.8] vs 53.7 points [16.8], P=.32).
COMMENT
Comparing average physicians' scores to the score of 50 points that would have been obtained by chance showed statistical significance for awareness of topics, for knowledge concerning diagnostic RMOs in groups 1 and 2, and in the whole population surveyed (Table 3). However, the average overall score (24 questions) concerning actual and fictitious RMOs was not different from the score of 50 points (50.5 points; 95% CI, 48.3-52.7). When there were differences between average physicians' scores and 50 points, they were not very large and small P values were calculated only as a consequence of large sample size. Despite the fact that 80% of the respondents declared that they sometimes consult RMOs, 44.3% at the time of patient consultation, this survey did not provide evidence of a strong effect of the implementation of RMO policy on the awareness and knowledge of RMOs among French family physicians.
Regulatory medical references were widely distributed and nationally endorsed, but diffusion alone has been proven to be ineffective.12-18 As an example, in Canada, Lomas et al13 found that 67% of obstetricians were aware of guidelines on cesarean delivery, widely disseminated by the Canadian Society of Obstetricians and Gynecologists, 1 year after their release. However, only 3% of them could correctly identify 4 recommended actions and 4 nonrecommended actions. Similarly, Fowler et al14 showed that 51% of British general practitioners reported having received guidelines on smoking cessation counseling, 28% of them reported having read it but only 9% could name the 3 major recommendations of the guidelines. In the case of RMO policy, the difficulties of physicians in remembering these RMOs are probably explained by the number of topics (n=46), the number of RMOs (n=197), and their 4-fold increase from 1994 to 1997. The fact that awareness and knowledge of RMOs was highest for those published in 1994 and lowest for recent RMOs published in 1996 (Table 1) reinforces this hypothesis.
The 3 groups recognized significantly better actual RMOs than fictitious RMOs (average score of 62.2 points [19.5] vs 43.2 points [25.0], P<.001). This could mean that physicians see the RMOs as overly intrusive and they overidentify them. This result is consistent with the belief that when this policy was implemented in 1994, physicians anticipated the controls and spontaneously limited their prescriptions, leading to a reduction of drug expenditures estimated in 1995 to be about Fr337 million.19 The threat of financial disincentives using RMOs could have an effect on physician behavior even if physicians do not know these RMOs. However, it is also possible that physicians paid no attention to RMOs and answered with their current practice patterns that happened to be in agreement with the true and the fictitious RMOs. In other words, when they are unable to discriminate, physicians assume the more restrictive option.
A recent study showed that the influence of the RMO policy on drug expenditures that was observed in 1994 was not observed for therapeutic RMOs published after 1994.20 The long-term results of financial disincentives on physicians' behavior depend on trust, legitimacy, and the quality of controls.21-22 Most French physicians resent the financial penalties. In 1998, 60% of French physicians declared that the RMO policy could affect the quality of care.23 However, at the end of 1997, only 151 physicians (0.14%) had been fined. It has been stressed that, due to the lack of pertinent information systems, it took 300 hours to check prescriptions ordered by 1 physician over a 2-month period.24 Thus, the difficulty of controlling family physicians limits the credibility of this policy, which was not perceived by physicians as a real threat.
Our study has some limitations. The RMO topics and RMOs chosen to be included in the questionnaire concerned problems faced by French family physicians. However, it is impossible to know if the results would have been different if we had chosen other RMOs. Another limitation of our study is that the 3 groups of physicians surveyed are not representative of the whole French physician population. They are probably more representative of French physicians interested in continuing medical education activities. The physicians in group 1 were interviewed before attending a session on guideline development methods, some of the physicians in group 2 participated in the development of RMOs, and the members of the Société de Formation Thérapeutique du Généraliste (group 3) have been involved in medical practice evaluation for years. The response rate of group 3 to our questionnaire was low (55.5%) compared with the response rate of the 2 other groups (92.5 and 78.3%, respectively). The members of this society are probably less enthusiastic about RMOs than the 2 other groups. However, considering the characteristics of the sample of physicians we had chosen, it is probable that the results would have been worse if we had surveyed a less selected population of French family physicians.
Lastly, the fact that French physicians surveyed were unable to correctly identify RMOs does not mean that their actual practice is inappropriate. It is possible that, since the content of most RMOs were not subject to criticism, they were considered by family physicians as too close to their actual practice. The usefulness of some RMOs has also been questioned. In a study on prescription of vasodilatator agents in peripheral occlusive artery diseases (actual topic of RMO correctly identified by 88.5% of our population), it has been shown that 80% of prescriptions of a population of French physicians surveyed was appropriate according to the RMO 1 year before its release.25
Four years after the implementation of the RMO policy, French family physicians are unable to correctly identify topics of RMOs or the RMOs themselves. The number of RMOs and the difficulties of controlling physicians probably explain these results. Thus, it is doubtful that this policy could have a long-term effect on French physicians' behavior.
AUTHOR INFORMATION
Accepted for publication October 4, 1999.
We thank, Françoise Cluzeau, MD, St George's Hospital Medical School, London, England, and Ian Graham, PhD, Loeb Health Research Institute, Ottawa, Ontario, for their comments during the preparation of the manuscript, the Société de Formation Thérapeutique du Généraliste, and all the physicians who participated in the study, and the experts who reviewed our manuscript for their pertinent comments.
Corresponding author: Pierre Durieux, MD, MPH, Santé Publique, Faculté de Médecine Broussais Hôtel Dieu, 15 rue de l'Ecole de Médecine 75006 Paris, France (e-mail: pierre.durieux{at}egp.ap-hop-paris.fr).
From the Santé Publique, Faculté de Médecine Broussais Hôtel Dieu, Paris, France (Dr Durieux); Centre de Recherche Clinique, Faculté de Médecine, Tours, France (Dr Giraudeau); Département de Médecine Générale, Centre Hospitalier Universitaire Bicêtre, Le Kremlin Bicêtre, France (Dr Doumenc); and Unité d'Epidémiologie, Hôpital Bichat, Assistance publique-Hôpitaux de Paris (Dr Ravaud). Dr Gaillac is in private practice in Paris.
REFERENCES
| |
1. Arrêté du 25 Novembre 1993 portant approbation de la convention nationale des médecins. J Off République Française. November 26, 1993: 16297-16311.
2. Ordonnance n° 96-345 du 24 Avril 1996 relative à la maîtrise médicalisée des dépenses de soins J Off République Française. April 25, 1996:6311-6320.
3. Durand-Zaleski I, Colin C, Blum-Boisgard C. Preliminary evaluation of an attempt to merge mandatory practice guidelines and cost containment in France. BMJ. 1997;315:943-946.
FREE FULL TEXT
4. Arrêté du 22 Mars 1994 portant approbation d'un avenant à la convention nationale des médecins. J Off République Française. March 24, 1994:4450-4455.
5. Arrêté du 28 Mars 1997 portant approbation de la Convention nationale des médecins généralistes. J Off République Française. March 29, 1997:4897-4922.
6. Colin C, Geffroy L, Maisonneuve H, et al. Country profile: France. Lancet. 1997;349:791-797.
FULL TEXT
|
ISI
| PUBMED
7. Dorozynski A. French doctors agitated over cash "bonus." BMJ. 1998;316:1037.
FREE FULL TEXT
8. Dorozynski A. France launches plan to control health costs. BMJ. 1998;317:164
FREE FULL TEXT
9. Drummond MF. Financial incentives to change behaviour towards health technology. In: Stocking B, ed. Expensive Health Technologies: Regulatory and Administrative Mechanisms in Europe. Oxford, England: Oxford Medical Publications; 1988:66-77.
10. Greco PJ, Eisenberg JM. Changing physician's practices. N Engl J Med. 1993;329:1271-1273.
FREE FULL TEXT
11. Woolf SH. Practice guidelines: a new reality in medicine, III: impact on patient care. Arch Intern Med. 1993;153:2646-2655.
FREE FULL TEXT
12. Freemantle N, Harvey EL, Grimshaw JM, et al. The effectiveness of printed educational materials in improving the behaviour of health care professionnals and patient outcome [Cochrane Review on CD-ROM]. Oxford, England: Cochrane Library, Update Software; 1997;issue 1.
13. Lomas J, Anderson G, Domnick-Pierre K, Vayda E, Enkin M, Hannah W. Do pratice guidelines guide practice? the effect of a consensus statement on the practice of physicians. N Engl J Med. 1989;321:1306-1311.
ABSTRACT
14. Fowler G, Mant D, Fuller A, Jones L. The "Help your patient stop" initiative: evaluation of smoking prevalence and dissemination of WHO/UICC guidelines in UK general practice. Lancet. 1989;1:1253-1255.
FULL TEXT
|
ISI
| PUBMED
15. Grimshaw J, Russel IT. Effect of clinical guidelines on medical practice: a systematic review of rigourous evaluations. Lancet. 1993;342:1317-1322.
FULL TEXT
|
ISI
| PUBMED
16. Grimshaw JM, Russel IT. Achieving health gain through clinical guidelines, II: ensuring guidelines change medical practice. Qual Health Care. 1994;3:45-52.
FREE FULL TEXT
17. Thomson R, Lavender M, Madhok R. How to ensure that guidelines are effective. BMJ. 1995;311:237-242.
FREE FULL TEXT
18. Solomon DH, Hashimoto H, Daltroy L, Liang MH. Techniques to improve physicians' use of diagnostic tests: a new conceptual framework. JAMA. 1998;280:2020-2027.
FREE FULL TEXT
19. Le Fur P, Sermet C. Les références médicales opposables, impact sur la prescription pharmaceutique. Paris, France: Centre d'Etude et de Documentation en Economie de la Santé (CREDES); 1996. No. 1116.
20. Le Pape A, Sermet C. Les Références Médicales Opposables Sur le Médicament: Bilan de Trois Années d'application. Paris, France: Centre d'Etude et de Documentation en Economie de la Santé (CREDES); 1998. No. 1237.
21. Monrad A. Incentives and financial methods. Health Policy. 1995;34:205-220.
FULL TEXT
|
ISI
| PUBMED
22. Rodwin MA. Medicine, Money and Morals: Physicians' Conflicts of Interest. Oxford, England: Oxford University Press: 1993.
23. Ordre National des Médecins. Maîtrise des dépenses: les médecins en quête de repères. Bull l'Ordre Méd. Mars 1998:10-17.
24. Dixon J. France seeks to curb health costs by fining doctors. BMJ. 1997;315:895-896.
FREE FULL TEXT
25. Vuittenez F, Guignard E, Comte S. Evolution de la prescription des vaso-actifs périphériques et cérébraux avant et après les références médicales opposables. Presse Med. 1999;28:122-126.
RELATED ARTICLE
The Archives of Family Medicine Continuing Medical Education Program
Arch Fam Med. 2000;9(5):463-465.
FULL TEXT
|
