Establishing recommendations for physical medicine and rehabilitation: the SOFMER methodology
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Annales de réadaptation et de médecine physique 50 (2007) 106–110
http://france.elsevier.com/direct/ANNRMP/
Professional practices and recommendations
Establishing recommendations for physical medicine and rehabilitation:
the SOFMER methodology
F. Rannoua,*, E. Coudeyreb, P. Ribinikc, Y. Macéa, S. Poiraudeaua, M. Revela
a
Service de rééducation, APHP, universite René-Descartes, groupe hospitalier Cochin, 27, rue du Faubourg-Saint-Jacques, 75014 Paris, France
b
Centre de MPR Notre Dame, BP 86, 4, avenue Joseph-Claussat, 63404 Chamalières, France
c
Service de MPR, centre hospitalier de Gonesse, 25, rue Pierre-de-Theilley, BP 71, 95503 Gonesse, France
Received 3 January 2007; accepted 10 January 2007
Abstract
Objective. – The development of a rigorous methodology based on published results of clinical trials, evaluation of daily practice in France
and multidisciplinary expert opinion to elaborate recommendations for rehabilitation interventions.
Methods and results. – The following describes the methodology of SOFMER (Société Française de Médecine Physique et de Réadaptation
[French Society of Physical Medicine and Rehabilitation]) for developing recommendations for rehabilitation interventions. The test case was
developing recommendations for rehabilitation in hip or knee osteoarthritis (OA) and hip or knee arthroplasty. Physicians in rehabilitation,
orthopedic surgery and rheumatology identified, synthesized, and analyzed data from the literature by use of the usual French system of grading
trials (the French Agency for Accreditation and Evaluation in Healthcare [ANAES] scale). The data were published results of comparative
controlled studies such as randomized controlled trials, controlled clinical trials, cohort studies, case control studies, reviews of clinical trials,
and case series, as well as uncontrolled cohort studies. The resulting recommendations were presented to the three annual French national con-
gresses of rehabilitation, rheumatology, and orthopedic surgery for comment and for adapting to French professional practice. Finally, a panel of
multidisciplinary experts (physicians in physical medicine and rehabilitation, rheumatologists, orthopedic surgeons, general practitioners, physical
therapists, social workers, podologists, occupational therapists, nurses, and patients) validated the recommendations.
Conclusion. – The SOFMER methodology could be an interesting tool for use in developing recommendations elaborated by all the con-
cerned medical and surgical specialists in the wide domain of rehabilitation.
© 2007 Elsevier Masson SAS. All rights reserved.
Keywords: Clinical practice guidelines; Evidence based practice; Practitioner feed back survey; Rehabilitation; Systematic review; Methodology
1. Introduction tism (EULAR) recommendations to use paracetamol (up to 4 g/
day) to manage hip OA [9].
In 1990 and 1995, the US Institute of Medicine described Besides pharmacological recommendations, nonpharmaco-
clinical practice guidelines as ‘systematically defined state- logical recommendations, mainly rehabilitation, are increas-
ments to assist practitioner and patient decisions about appro- ingly being indicated as routine treatment. For example, the
priate health care for specific clinical circumstances’ [4,5]. In first EULAR recommendations for optimal management of
musculoskeletal diseases, clinical practice guidelines have been hip or knee OA suggest a combination of nonpharmacological
published for osteoarthritis (OA), rheumatoid arthritis, ankylos- and pharmacological treatment [6,9]. Unfortunately, for daily
ing spondilitis, and gout. Precise recommendations for pharma- practice, only one item of the recommendations describes non-
cological treatment can be easily described with the drug name pharmacological treatment—education, exercise, use of appli-
and exact dosage, as in the European League Against Rheuma- ances, and weight reduction—thus leading to under-utilization
of nonpharmacological treatments [2,6,9]. Nonpharmacological
recommendations are less precise than pharmacological recom-
* Correspondingauthor. mendations to ‘assist practitioner and patient decisions about
E-mail address: francois.rannou@cch.aphp.fr (F. Rannou). appropriate health care for specific clinical circumstances.’
0168-6054/$ - see front matter © 2007 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.annrmp.2007.01.004F. Rannou et al. / Annales de réadaptation et de médecine physique 50 (2007) 106–110 107
In France, in a large prospective, observational, multicenter Table 1
Members of the steering committee
survey conducted among 1030 general practitioners (GPs),
only 48.7% of the GPs prescribed physical activity and less Calmels Paul, MD, Ph.D., service de MPR, hôpital Bellevue, 42055 Saint-
Étienne, France
than 6% the use of insoles for knee OA as compared with
Coudeyre Emmanuel, MD, centre de MPR Notre-Dame, 63404 Chamalières,
95.8% of GPs prescribing paracetamol [3]. However, 99% of France
the GPs agreed with the nonpharmacological recommendation, Lecocq Jehan, MD, Ph.D., service de MPR, hôpitaux universitaires–hôpital de
thus showing a great discrepancy between following a recom- Hautepierre, 67098 Strasbourg, France
mendation for rehabilitation in daily practice (48.7%) and Lefevre-Colau Marie-Martine, MD, Ph.D., service de MPR, hôpital Corentin-
Celton, 92133 Issy-les-Moulineaux, France
agreement about this recommendation (99%). This phenom- Poiraudeau Serge, MD, Ph.D., service de MPR, hôpital Cochin, université
enon was not observed for the pharmacological recommenda- Paris-V, 75014 Paris, France
tion, as demonstrated by the very similar percentages for fol- Rannou François, MD, Ph.D., service de MPR, hôpital Cochin, université
lowing the recommendation (95.8%) and agreement about the Paris-V, 75014 Paris, France
recommendation (97%). Revel Michel, MD, service de MPR, hôpital Cochin, université Paris-V, 75014
Paris, France
Our aim was to develop a specific methodology to elaborate Thevenon André, MD, service de MPR, CHRU, 59037 Lille, France
recommendations for rehabilitation interventions, which should Thoumie Philippe, MD, Ph.D., service de MPR, hôpital Rothschild, 75012
therefore be adapted more to the daily practice of GPs to facil- Paris, France
itate the correct prescription of a clear intervention and thus a Tiffreau Vincent, MD, service de MPR, CHRU, 59037 Lille, France
Vautravers Philippe, MD, service de MPR, hôpitaux universitaires–hôpital de
better clinical improvement for the patient. We devised a meth- Hautepierre, 67098 Strasbourg, France
odology that combines analysis of published results of clinical MPR: médecine physique et de réadaptation (Physical Medicine and Rehabili-
trials by all the medical and surgical specialists involved in the tation).
area of interest, evaluation of daily practice in France and mul-
4. Agreement from the French societies of rehabilitation,
tidisciplinary expert opinion. This methodology was used for
orthopedic surgery, rheumatology, and the bone and joint
the first time in France in year 2006 to establish recommenda-
decade
tions for rehabilitation in knee or hip OA and hip or knee
arthroplasty.
Agreement and support for the study were obtained from the
presidents of SOFMER, the French Society of Rheumatology
2. Formation of the steering committee (SFR), and the French Society of Traumatic and Orthopedic
Surgery (SOFCOT). Agreement from the Bone and Joint Dec-
The following methodology was developed and financed by ade was obtained from its French representative Pr Liana
SOFMER (Société Française de Médecine Physique et de Euler–Ziegler.
Réadaptation [French Society of Physical Medicine and Reha-
bilitation]). SOFMER defined the clinical area for the recom- 5. Choice of the scientific committee and experts
mendations—rehabilitation in knee or hip OA and knee or hip
arthroplasty—and assigned three of its members (FR, MR, and The scientific committee members who directed the analysis
PR) to create a steering committee representing all the practices of publications and development of recommendations were
(academic, public and private) of physical medicine and reha- nominated by the SOFMER, SFR, and SOFCOT and repre-
bilitation in France (Table 1). sented the different modalities (academic, public or private)
of medical and surgical expertise in France (Table 2).
All experts who analyzed the publication results were nomi-
3. Choice of questions relevant to rehabilitation in knee
nated by SOFMER, SFR, and SOFCOT and were practitioners
or hip OA and knee or hip arthroplasty
with academic, public or private professional positions. Two
experts each (one each from SFR and SOFMER) were
Each steering committee member was asked to prepare sev- assigned to the medical questions—questions 1 and 2—and
eral questions related to rehabilitation in hip or knee OA and two each (one each from SOFCOT and SOFMER) to the ques-
hip or knee arthroplasty. These questions had to relate to high- tions related to surgery—questions 3–5 and all the subques-
volume activities of rehabilitation in France that currently had tions. During all procedures, each expert was blinded to the
no recommendations. The questions were collected and pre- identity of the other expert for a given question or subquestion
sented at a national meeting of the steering committee funded for internal control. A total of 32 experts were selected, 26 in
by SOFMER in Paris at Cochin Hospital. At the meeting, the charge of 1 question or subquestion and 6 in charge of 2 ques-
questions were discussed, and if two-thirds of the committee tions or subquestions.
members agreed on the question, it was retained. Finally, five
questions were selected: two related to rehabilitation in knee or 6. Literature search strategy
hip OA and three to rehabilitation in hip or knee arthroplasty.
For each question, subquestions were added to elaborate the Literature search professionals systematically searched the
recommendations, for a final total of 19 subquestions. PubMed, Pascal Biomed, and Cochrane databases for articles108 F. Rannou et al. / Annales de réadaptation et de médecine physique 50 (2007) 106–110
Table 2 8. Analysis of manuscripts by experts, categories
Members of the scientific committee of evidence and strength of recommendation
Coudeyre Emmanuel, MD (SOFMER), centre de MPR Notre-Dame, 63404
Chamalières, France
Genet François, MD (SOFMER), service de MPR, groupe hospitalier
Two experts from two different medical specialties were
Raymond-Poincaré, 92380 Garches, France assigned to each question as described previously. They were
Genty Marc, MD (SOFMER), service de MPR, clinique Valmont, GLION S- blinded from each other during the analysis stage, for internal
1823, Switzerland control. Experts received electronic or paper versions of each
Ribinik Patricia, MD (SOFMER), service de MPR, centre hospitalier de manuscript. Using the full-length versions, they identified all
Gonesse, BP 71, 95503 Gonesse, France
comparative controlled studies such as randomized controlled
Rannou François, MD, Ph.D. (SOFMER), service de MPR, hôpital Cochin,
université Paris-V, 75014 Paris, France trials, controlled clinical trials, cohort studies, case control stu-
Revel Michel, MD, (SOFMER), service de MPR, hôpital Cochin, université dies, reviews of clinical trials, and case series and all uncon-
Paris-V, 75014 Paris, France trolled cohort studies related to their question. Studies were
Beaudreuil Johann, MD, Ph.D. (SFR), Fédération de rhumatologie, hôpital selected if they measured concordant outcomes related to the
lariboisière, 75475 Paris, France
question. For each question, the scientific committee and
Lefevre-Colau Marie-Martine, MD, Ph.D. (SFR), service de MPR, hôpital
Corentin-Celton, 92133 Issy-les-Moulineaux, France experts selected outcomes of interest. The main outcomes
Biau David, MD (SOFCOT), service de chirurgie orthopédique et used were impairment, disability, and quality of life.
traumatologie B, hôpital Cochin, université Paris-V, 75014 Paris, France The quality of each manuscript was assessed according to
Anract Philippe, MD (SOFCOT), service de chirurgie orthopédique et the grading scale of the French Agency for Accreditation and
traumatologie B, hôpital Cochin, Université Paris-V, 75014 Paris, France
Evaluation in Healthcare (ANAES) (Table 3). The quality of
SFR: Société française de rhumatologie (French Society of Rheumatology) ;
SOFMER: Société française de médecine physique et de réadaptation (French results could be analyzed according to three levels of recom-
Society of Physical Medicine and Rehabilitation) ; SOFCOT: Société française mendation: A, established scientific evidence; B, scientific pre-
de chirurgie orthopédique et traumatologique (French Society of Traumatolo- sumption; and C, low level of evidence. At the end of the ana-
gic and Orthopedic Surgery). lysis, each expert had to write a manuscript on the model of
scientific publication.
published from January 1966 to January 2006 using search
terms defined by the scientific committee. Selected were 9. Meeting of experts to compare categories of evidence
abstracts of studies of all design that included an abstract, and strength of conclusions from literature analysis
were published in English or French, and investigated adult
human patients. For question 1, 401 references were found in A 2-day meeting organized in Paris at Cochin Hospital and
PubMed, and 190 in Pascal Biomed. For question 2, 92 refer- financed by SOFMER was devoted to the experts’ results for
ences were found in PubMed, and 92 in Pascal Biomed. For the five questions and subquestions. The first day was devoted
question 3, 4, and 5, 273 references related to the knee and 307 to the medical questions 1 and 2. A discussion of the findings
references related to the hip were found in PubMed, and 197 of the two experts for each question was moderated by a mem-
references related to the knee and 153 references related to the ber of the scientific committee (FR or PR). Each expert’s
hip were found in Pascal Biomed. FR, for the scientific com- manuscript was collected by the scientific committee. Differ-
mittee, searched the Cochrane database for articles: six articles ences in assessment of a report’s evidence and final recommen-
were selected for question 1, 4 for question 2, and 10 for ques-
Table 3
tions 3, 4, and 5. Grading system for evidence from published trials of the French Agency for
Accreditation and Evaluation in Healthcare (ANAES)
7. Selection of manuscripts Grading of Published Trial Results
1 Randomized controlled trials of high power
1 Meta-analysis of randomized controlled trials
The literature search professionals sent abstracts to FR, MR, 1 Decision based on well designed trials
and PR, for the scientific committee, who then narrowed the
selection of abstracts to analyze by ensuring that ‘rehabilitation 2 Randomized controlled trials of low power
2 Comparative non randomized trials well designed
intervention’ was present in the abstract, then requested the
2 Cohort studies
full-length articles of the selection from professional literature
searchers. For question 1, 172 manuscripts were selected from 3 Case control studies
PubMed, 74 from Pascal Biomed, and six from the Cochrane
database. For question 2, 92 manuscripts were selected from 4 Comparative studies with major bias
PubMed, 11 from Pascal Biomed, and four from the Cochrane 4 Retrospective studies
4 Cases series
database. For questions 3, 4, and 5, 156 manuscripts were
selected from PubMed, 69 from Pascal Biomed, and 10 from Grading of Recommendations
the Cochrane database. A total of 274 manuscripts were A Established scientific evidence (level 1 of evidence)
selected from PubMed and Pascal Biomed and 16 from the B Scientific presumption (level 2 of evidence)
Cochrane database. C Low level of evidence (levels 3 and 4 of evidence)F. Rannou et al. / Annales de réadaptation et de médecine physique 50 (2007) 106–110 109
dations for each question were resolved by the representative Table 4
Members of the reading committee
member of the scientific committee. At the end of the meeting,
conclusions from the literature analysis were obtained for each Kemoun Gilles (PMR)
Dauty Marc (PMR)
question by consensus of the two experts. The same validation
Nys Alain (PMR)
process was used the second day for questions 3, 4, and 5. Tavernier Christian (rheumatologist)
Forestier Romain (rheumatologist)
10. Elaboration of questions to evaluate French physicians’ Reboux Jean François (rheumatologist)
prescription and daily practice for rehabilitation in knee Hamadouche Moussa (orthopedic surgeon)
or hip OA and knee or hip arthrplasty Cottias Pascal (orthopedic surgeon)
Caton Jacques (orthopedic surgeon)
Lorenzo Alain (general practitioner)
A 2-day meeting of the scientific committee organized in Favre Madeleine (general practitioner)
Paris at Cochin Hospital and financed by SOFMER aimed to Fabry Stephane (Physical therapist)
elaborate questions related to each of the 19 questions that Marc Thierry (physical therapist)
could be used to evaluate French physicians’ prescription for Pillu Michel (physical therapist)
rehabilitation in knee or hip OA and knee or hip arthroplasty. Jouhaneau Sylvie (social worker)
Massaro Raymond (podologist)
FR asked each committee member to elaborate questions on
Hynaux Isabelle (occupational therapist)
the area related to their competence (rehabilitation, rheumatol- Matter Claire (nurse)
ogy, and orthopedic surgery). During the meeting, all questions Guillemin Dominique (patient)
were collected. When two-thirds of the scientific committee
members agreed on a question, it was accepted. A list of 37 The scientific committee formed a multidisciplinary reading
questions was defined. committee (Table 4) to comment on the recommendations
developed from analysis of published clinical trials and evalua-
11. Presentation of conclusions from experts’ analysis tion of daily practice in France. This multidisciplinary commit-
of the literature to the 2006 national congresses tee consisted of physicians in physical medicine and rehabilita-
of rehabilitation, orthopedic surgery, and rheumatology tion, orthopedic surgery, rheumatology, and family practice;
and evaluation of prescription and daily practice physical therapists; occupational therapists; podologists; social
for rehabilitation workers; nurses; and patients. Members were nominated to this
reading committee by their representative national organiza-
The conclusions from the experts’ analysis of the literature tions. Each member received a recommendation one at a time
were presented at the national congresses of rehabilitation and had 1 week to comment. In the absence of a response, the
(SOFMER Congress, Rouen, France, October 18, 2006), ortho- scientific committee considered that the reading committee
pedic surgery (SOFCOT National Congress, Paris, November member accepted the recommendation. Feedback by these
7, 2006), and rheumatology (SFR National Congress, Paris, practitioners and patients was discussed by the scientific com-
December 4–5, 2006). Before the presentation of the conclu- mittee, and the final recommendations were revised accord-
sions of each literature review, FR asked the physicians attend- ingly.
ing the meetings the 37 questions evaluating their prescription
and daily practice for rehabilitation. Electronic vote was used 13. Final publication of recommendations
for an exact recording of responses. One hundred and twenty
(120) physical medicine and rehabilitation physicians were pre- The final recommendations will be published in referenced
sent for the recommendation session at the SOFMER congress, medical journals in French and English and will be available to
120 orthopedic surgeons at the SOFCOT congress, and 100 the public on the web sites of SOFMER, SFR and SOFCOT.
rheumatologists at the SFR congress. After the literature review
presented by the experts responsible for each question, the ses- 14. Discussion
sion was open for questions and comments. A medical secre-
tary took notes during the question-and-comment period. Even if rehabilitation for several diseases is widely accepted
in general practice, the actuality of prescription and application
12. Elaboration of recommendations in terms of literature, are still debated. Perhaps the process of developing recommen-
transdisciplinary French prescription and daily practice dations has an influence.
and external review by a reading committee The SOFMER methodology has been developed from a
combination of analysis of results of published clinical trials,
Experts added comments and notes gained from the three evaluation of prescription and daily practice in France, and
national congresses to their own literature review and opinion multidisciplinary expert opinion to define recommendations
to propose recommendations concerning French prescription for rehabilitation, in this case rehabilitation in knee or hip OA
and daily practice for rehabilitation. These recommendations and knee or hip arthroplasty. To date, the development of
were reviewed by the scientific committee before validation recommendations has resulted from the opinion of experts, lit-
by a reading committee. erature search, or both. EULAR used this hybrid method to110 F. Rannou et al. / Annales de réadaptation et de médecine physique 50 (2007) 106–110
develop recommendations for musculoskeletal diseases [9]. By studying and learning from the EULAR methodology,
The SOFMER methodology involves a modified hybrid the MOVE consensus, and the Philadelphia panel, we have
method of literature search and expert opinion added to adap- elaborated a new method to develop recommendations for
tation to French prescription and daily practice. In contrast to rehabilitation, one that is based on literature search, multidisci-
the EULAR use of expert opinion, the SOFMER methodology plinary expert opinion and application to daily primary health
requires a multidisciplinary final reading committee of not only care. By its very nature, this method may help physicians and
physicians but also health practitioners and patients. As seen in patients accept recommendations for a wide variety of areas of
Table 4, the reading committee represented 10 different rehabilitation, including neurological rehabilitation.
domains of health care. The EULAR methodology involves a
predominance of rheumatologists (18/23) and orthopedic sur- Acknowledgements
geons (4/23) involved in secondary or tertiary care of OA
patients [9]. The SOFMER multidisciplinary method for devel- The authors wish to thank N. Thibiéroz, F. Wergens and
oping recommendations might help increase the acceptability J. Kalfon from the Paris-V University Library for help in the
of the recommendations in real-life daily activity in primary literature search process.
care.
To our knowledge, the SOFMER methodology is the first to References
take into account patient opinion, which could be a first step in
reaching high acceptance of a nonpharmacological treatment [1] Cluzeau F. Appraisal Instrument of Guidelines for Research and Evalua-
by patients. Even if recommendations are scientifically perti- tion (AGREE) Instrument. Available at: htpp://www.agreecollaboration.
nent, if they are not used by GPs and accepted by patients, org.
their usefulness is questionable. In addition, the financial inde- [2] Chevalier X, Marre JP, de Butler J, Hercek A. Questionnaire survey of
pendence of the SOFMER methodology could help with management and prescription of general practitioners in knee osteoarthri-
tis: a comparison with 2000 EULAR recommendations. Clin Exp Rheu-
acceptance of recommendations. For the EULAR recommen- matol 2004;22:205–12.
dations, the pharmaceutical industry was involved in their [3] Denoeud L, Mazieres B, Payen-Champenois C, Ravaud P. First line
development. However, all the costs for the development of treatment of knee osteoarthritis in outpatients in France: adherence to
SOMER recommendations were absorbed by SOFMER, a non- the EULAR 2000 recommendations and factors influencing adherence.
profit professional organization. Ann Rheum Dis 2005;64:70–4.
[4] Fiel MJ, Lohr KN. Clinical practice guidelines: directions for a new pro-
Roddy et al. [8], for the MOVE consensus, were the first to
gram. Washington, DC: The National Academies Press; 1990.
try to describe precisely the practical delivery of exercise ther- [5] Fiel MJ. Setting priorities for clinical practice guidelines. Washington,
apy in OA. The authors first used a round of expert opinion DC: The National Academies Press; 1995.
based on the Delphi method and then conducted a literature [6] Jordan KM, Arden NK, Doherty M, Bannwarth B, Bijlsma JW, Dieppe
search. In our methodology, expert opinion follows the litera- P, et al. EULAR Recommendations 2003: an evidence based approach to
ture search, for an unbiased search of the literature. As well, the management of knee osteoarthritis: report of a Task Force of the
Standing Committee for International Clinical Studies Including Thera-
the MOVE authors did not collect the daily practice of physi- peutic Trials (ESCISIT). Ann Rheum Dis 2003;62:1145–55.
cians implicated in rehabilitation nor patient opinion. Lastly, [7] Philadelphia Panel evidence-based clinical practice guidelines on selected
the Philadelphia panel developed a methodology of expert opi- rehabilitation interventions: overview and methodology. Phys Ther 2001;
nion, literature search and external practitioner review for reha- 81:1629–40.
bilitation interventions [7]. In this study, the final evaluation [8] Roddy E, Zhang W, Doherty M, Arden NK, Barlow J, Birrell F, et al.
Evidence-based recommendations for the role of exercise in the manage-
involving the AGREE scale gave too much importance to
ment of osteoarthritis of the hip or knee—the MOVE consensus. Rheu-
expert opinion and thus to subjective opinion [10]. Zhang and matology (Oxford) 2005;44:67–73.
Doherty, in a recent study critiquing the EULAR methodology, [9] Zhang W, Doherty M, Arden N, Bannwarth B, Bijlsma J, Gunther KP,
point out the subjectivity of the AGREE scale [1] leading to an et al. EULAR evidence based recommendations for the management of
emphasis on subjective opinion of experts and largely neglect- hip osteoarthritis: report of a task force of the EULAR Standing Commit-
tee for International Clinical Studies Including Therapeutics (ESCISIT).
ing research evidence [9]. In the field of rehabilitation, we need
Ann Rheum Dis 2005;64:669–81.
scientific objectivity, not subjective opinion based on non- [10] Zhang W, Doherty M. EULAR recommendations for knee and hip
scientific and medical evidence, in the development of recom- osteoarthritis: a critique of the methodology. Br J Sports Med 2006;40:
mendations. 664–9.You can also read
