REMED: An Essential Method for Improving the Quality and Safety of Medication Care

Patient safety is a top priority in healthcare. At the heart of this concern lies the management of risks associated with medications and related medical devices, an area where errors can have devastating consequences. Faced with this challenge, the Medication and Associated Device Error Review (REMED) is emerging as a fundamental method. Developed by the French Society of Clinical Pharmacy (SFPC), REMED is much more than a simple procedure; it is a systemic and educational approach aimed at analyzing errors after the fact to draw collective lessons and implement concrete improvement actions. This article explores REMED in depth, from its definition to its practical application, including its history, scope, and the tools it provides healthcare professionals to build a sustainable safety culture. What is REMED and its importance? REMED (Review of Errors Related to Medications and Associated Devices) is a continuous quality improvement approach for healthcare and the prevention of iatrogenic drug-related risks. Its English equivalent is "Review of Errors in Medicines." This method helps optimize the organization of therapeutic care for patients where medication and, where applicable, associated medical devices are used. The expression "Medication Error Review" is also commonly used. REMED is a structured approach to the post-hoc analysis of anonymized medication error cases. It is distinguished by its collective, multi-professional, and systemic approach. Its main objective is to design and implement actions to reduce risks related to patient medication management. A medication error is defined as the omission or unintentional act occurring during the care process involving a medication, which may lead to a risk or adverse event for the patient (according to AFSSAPS 2010). The SFPC (2006) adds that a medication error is a deviation from what should have been done and is therefore, by definition, avoidable. REMED is crucial because awareness of the need to manage health risks, particularly those related to errors, is gradually growing among healthcare professionals, despite the difficulty of addressing it publicly. The World Health Organization (WHO) reiterated in March 2002 that more than 75% of the causes of healthcare errors are systemic in origin. In France, the Public Health Law of August 9, 2004, established serious adverse events, particularly those related to health products, as a public health issue. Recent reports (Costagliola & Bégaud 2013) estimate that between 10,000 and 30,000 deaths annually in France are attributable to medications. A very large proportion of this medication-related iatrogenesis is, however, avoidable because it is linked to an error. The French National Authority for Health (HAS) requires healthcare facilities to implement a health risk management policy from the initial certification process. In this context, REMED, as a validated offshoot of morbidity and mortality reviews by the HAS, becomes an indispensable tool. The systemic analysis conducted during REMED takes into account all the organizational, technical, and human elements of a patient's medication management. It allows us to move beyond a purely individual-centered approach and to draw lessons about the strengths and vulnerabilities of the healthcare system. By collectively reflecting on the process retrospectively, REMED makes it possible to characterize errors, identify shortcomings, analyze these situations of "non-quality" in a transparent and non-blaming manner, investigate the root causes, and propose improvement actions. The aim is to describe the facts and analyze situations to understand, learn, and act collectively, thereby strengthening the quality and safety of care. Considering the individuals involved as "resource persons" for improvement, without value judgment, is essential to the success and sustainability of the approach.

The History and Recognition of REMED: A Tool Validated by Experts

The French Society of Clinical Pharmacy's (SFPC) commitment to combating medication-induced iatrogenesis dates back more than ten years. It was within this framework that a specific working group, led by Dr. Edith Dufay, was established. As early as 2008, these expert professionals in the field proposed an initial version of the REMED method.

Recognition of REMED accelerated thanks to strategic collaborations. In 2007, REMED was selected as a working theme in partnership with two other learned societies: the French Society of Geriatrics and Gerontology (SFGG) and the French Society for Risk Management in Healthcare Facilities (SOFGRES). A Delphi method was used to define and formalize the method and its tools. A major turning point was the multicenter study MERVEIL (Method for Evaluating the Review of Errors and their Iatrogenesis related to Medications), conducted under the auspices of the SFPC. This study, conducted from 2009 to 2011, involved professionals from 77 healthcare facilities in France, Belgium, and Luxembourg. Its objective was to evaluate and validate REMED as a method for assessing professional practices. The results of the MERVEIL study not only validated REMED, but also positioned it as the reference method for the post-hoc analysis of medication errors. The lessons learned from the MERVEIL study led to the development, in 2013, of a second version of REMED, easier to apply and consistent with other risk management methods. The manual for this version 2.0 was published in 2014. Institutional recognition followed quickly: in 2012, REMED was recognized by the French National Authority for Health (HAS) as one of the methods listed within the framework of Continuing Professional Development (CPD). The HAS (French National Authority for Health) has also recognized it as a "validated daughter method" for morbidity and mortality reviews. Today, REMED is considered an indispensable tool for evaluating and improving the safe management of patients treated with health products. Its multidisciplinary approach, developed with expert professionals and field practitioners, takes into account all the organizational, technical, and human aspects of medication management. It proposes the collective implementation of actions to improve the quality and safety of care, thus fully integrating into the organization of healthcare facilities regarding medication risk management. The success of REMED is the result of momentum and collective effort, as highlighted by the presidents of the SFPC (French Society of Clinical Pharmacy), Professor Marie Claude Saux and Professor Rémi Varin. Numerous contributors, including Edith Dufay, as well as members of the working groups, writers, and reviewers, participated in its design and development. The ongoing support of the HAS and various collaborating bodies and societies was essential. The SFPC has made this work freely available to the profession, demonstrating collective pride in this project.

The scope of REMED investigation: which events are targeted?

REMED focuses specifically on the preventable adverse events. These are events that would not have occurred if care had been provided according to standards of care considered satisfactory at the time of their occurrence. The scope of REMED covers medication errors and/or harm observed in patients as a consequence of these errors.

More specifically, medication errors are dreaded events that disrupt care activities, reduce the level of safety, and can harm the patient because they deviate from recommendations or procedures. They encompass active errors by professionals, latent errors in organization and teamwork, technical failures, systemic malfunctions, and defects in health products, often in the presence of contributing or triggering factors. These elements constitute adverse events in themselves. The damages are the consequences that these errors generate or are likely to generate in the patient. They can take several clinical forms:

• Worsening of an existing condition or lack of expected improvement in health.
• Onset of a new condition, for example, a nosocomial infection following poorly administered antibiotic prophylaxis, or heart failure due to forgetting to take a diuretic.
• Impairment of a bodily function, or an adverse effect/harmful reaction caused by one or more medications taken by the patient.

Some of these consequences are considered serious when they result in incapacity or disability (temporary or permanent), physical, moral, or psychological harm, a life-threatening situation, death, or an anomaly/malformation congenital.

According to REMED, it is essential to clearly distinguish, when using the term "adverse event," between the facts (the medication error itself, or medication error and patient safety incident) and their consequences (the harm occurring to the patient, or adverse drug event and adverse drug reaction).

Beyond the harm to the patient, a medication error can also have organizational and/or institutional consequences. For the patient, this may mean the implementation of corrective treatment, increased monitoring, transfer to an intensive care unit, resuscitation unit, dialysis center, or another facility, a prolonged hospital stay, or the hospitalization of an outpatient or resident patient. For the healthcare system, the consequences can be economic, financial, social (related to healthcare staff), legal, or media-related. Sources provide concrete examples, such as a methotrexate overdose resulting in a patient's death (Example 1), or an omission of thromboembolic prophylaxis leading to phlebitis (Example 2). Cases without serious clinical consequences for the patient, such as the IV injection of a crushed tablet intercepted before harm (Example 3), or a dosage error intercepted by the pharmacist (Example 4), are also relevant events for REMED. REMED does not cover certain types of events. Adverse events related to the spontaneous progression of the patient's condition and adverse drug effects related to the intrinsic pharmacological properties of the medication are excluded. Similarly, events related to addictive behavior (e.g., drug use by drug addicts) or to the misuse/abuse of a health product (e.g., erythropoietin use by athletes) are excluded due to their intentional nature. It is important to remember that adverse events related to health products must be reported to the organizations managing regulatory health surveillance (pharmacovigilance, materiovigilance) because of their potential impact on marketing. Specific medication errors must be reported to the Medication Error Reporting Office of the French National Agency for Medicines and Health Products Safety (ANSM).

Medication Management: A Complex Process at the Heart of REMED

Patient medication management is a highly complex process, often one of the least well-organized in healthcare facilities, as demonstrated by the results of the first accreditation and certification processes conducted by the French National Authority for Health (HAS). This is why the collective interdisciplinary analysis carried out during a REMED adopts a problem-based approach that addresses the impact on the quality of care outcomes. It targets the process and interfaces related to the medication management of patients (hospitalized, resident, or outpatient) where an error has occurred.

This process generally breaks down into five major steps:

1. The diagnosis of the pathology, with the definition of therapeutic objectives, the choice of a therapeutic strategy, and then the prescription of medications and other care.
2. The dispensing of the medication(s) and the delivery, if necessary, of the associated device.
3. The administration and the patient's intake of the medication.
4. Therapeutic monitoring of the patient, including the reassessment of the benefit/risk balance.
5. The

This complexity is due to several factors:

• The diversity of pathologies, their degree of urgency and severity.
• The diversity of available therapies.
• The diversity of patient profiles.
• The involvement of different healthcare professionals whose roles are complementary.
• The sequence of numerous steps that take place in geographically different locations.
• The superimposition of an informal communication network onto the regular and formalized information flow to facilitate task coordination.

The document details the sequence of behaviors following an adverse event in six phases:

1. Interpretation: The patient or professional is alerted by an unusual fact or anomaly.
2. Questioning: The professional questions themselves, tries to reconstruct the chronology, and analyzes the events.
3. Detection: The cause is identified, allowing for an appropriate response.
4. Interception: The professional intervenes to interrupt the erroneous process, before or after harm occurs.
5. Recovery: An attempt to rectify the situation for the patient, the success of which depends on the speed of detection, the Drug hazard, severity of the event, professional experience, risk awareness, quality of teamwork, and organizational performance.
6. Analysis: The professional lists the immediate causes. This analysis is rarely formalized or systematic and often stops at the medication error itself, without seeking the root causes.

REMED is the answer to this last step, offering a method for a thorough and collective analysis, essential for identifying system vulnerabilities and implementing sustainable improvement actions.

Practical organization of REMED: implementing an effective approach

Implementing REMED is a structured approach that is fully integrated into the continuous quality improvement policy of any structure, whether it is a healthcare facility, a social care facility, a department, a service, a care unit, or a primary care structure. It concerns all sectors of activity involved in the therapeutic care of patients using a health product.

A fundamental aspect of REMED's organization is the support of healthcare professionals. The management of the institution must implement a real change in behaviors and organizations to promote a safety culture. To facilitate the involvement of professionals in reporting events, it is crucial to professionalize risk management, make the reporting system visible, and disseminate a charter defining management's position regarding staff involved in or who have reported an error. Psychological support must be offered to staff involved in a serious adverse event, the objective being to transform feelings of guilt into a moral strength, leading them to see themselves as resources for preventing recurrence.

To begin, the organization of REMED is entrusted to a volunteer healthcare professional, familiar with the method and recognized by their peers. Internal regulations, adapted to the institution and describing the organization and operation of the MEDIC (Medication Incident Response) system, must be developed and distributed to all relevant professionals. Case identification uses standard risk management and epidemiology search and detection methods. This includes reporting medication errors within the institution, direct observation of care practices, scheduled review of patient records, continuous recording of indicators related to errors or iatrogenesis, and monitoring of sentinel events (such as transfer to intensive care, unplanned readmission, or death). Case selection prioritizes "high-risk" medication errors, whether intercepted or not, that have had or could have had significant clinical consequences for the patient. Eligible cases may include "never events" listed by the ANSM (French National Agency for Medicines and Health Products Safety), errors associated with a type of care (anesthesia), a therapeutic class (anticoagulants), a route of administration (intrathecal), a population (elderly, child), a step in the product circuit (dispensing, administration), or a sector of activity (surgery, pharmacy). REMED is a multidisciplinary team effort. Typical participants are prescribers, pharmacists, pharmacy technicians, nurses, and healthcare managers. Other professionals may be invited depending on the cases studied, and professionals directly involved in the error may participate if invited. A multidisciplinary approach is essential, requiring the presence of at least one representative from the medical, pharmaceutical, nursing, and pharmacy technician professions. Confidentiality is strictly enforced for all participants.

A facilitator is appointed by the working group to lead the REMED process. Their responsibilities include organizing sessions, writing reports, facilitating discussions, formalizing the case analysis, and liaising with the relevant departments within the institution to monitor improvement actions. It is recommended that this role be assigned to a senior practitioner who volunteers, has training in case analysis, and possesses strong interpersonal skills (listening, empathy, rigor, and impartiality).

Traceability and archiving are essential. A standardized and anonymized report, along with an improvement action plan, is written and shared with participants. An annual activity report is prepared, and all documents are kept in the institution's quality records. The dissemination of information must be targeted: comprehensive information for staff directly involved in the event, and more strategic information, focused on improvement actions, for institutional recipients. A structured feedback process is organized, based on the annual review of reported errors and the medication management improvement program. Additional tools are available, such as the medication error tracking chart and the Med'Seth program for monitoring improvement actions. The 5C scenario, which diagrams the chain of causes and consequences of a medication error, is also a key educational tool.

Conducting a REMED session: a structured process with key steps

Conducting a REMED session is part of W. Edwards Deming's continuous improvement model PDCA (Plan, Do, Check, Act). The collective analysis of an error (or series of similar errors) relies on the REMED binder, whose central tool is the REMED notebook.

Here are the seven structured steps for effectively conducting a REMED session:

Step 0: Before the meeting (PLAN) This is the meticulous preparation phase. It includes the logistical planning of the meeting and the creation of an anonymized agenda. A case-specific REMED binder must be created, information recording must begin, relevant scientific documentation must be gathered, and, if necessary, interviews with the individuals involved must be conducted. Supporting those involved and offering them the opportunity to participate in the REMED process are essential for a non-judgmental approach. Finally, all documents, materials, and health products related to the case must be collected, and the drafting of the facts and the characterization of the error must begin. Step 1: Introducing the Meeting (PLAN) The facilitator begins by designating the meeting secretary and distributing the document "REMED in Summary." It is crucial to reiterate the fundamentals of the process: ethical rules, confidentiality, and anonymization. The objectives of REMED are specified: to improve practices and teamwork, to communicate about difficulties and organizational shortcomings, and to correct flawed processes. Emphasis is placed on avoiding value judgments to reassure participants and encourage open expression. Data recording in the REMED notebook continues throughout the session. Step 2: Presenting the Case and Identifying Problems Encountered (DO) This step aims to answer the questions "What happened?" and "How did it happen?". After a chronological description of the facts, the 5W2H (Who? What? Where? When? How? How many? Why?) tool is used to clarify and supplement the facts in a non-interpretive manner with the participants. This also involves identifying the safety barriers that failed, those that worked, and the methods for detecting and recovering from errors.

Step 3: Characterize the medication error (MO) This involves determining the type of error. The characterization of the medication error is validated by the participants according to six significant elements (Appendix I):

• The health products involved (medicine, treatment, protocol, medical device).
• The level of realization of the error (potential, actual and intercepted, actual and identified after reaching the patient).
• The nature of the medication error (patient, omission, medication, dose, method, timing, duration of administration).
• The observed severity of the consequences (minor, significant, major, critical, catastrophic, according to the HAS scale).
• The risk associated with the error (risk-carrying or not).
• The initial stage of occurrence in the patient care process (identification, prescription, dispensing, administration, follow-up, information, logistics). Consideration is also given to the secondary stages where the error continued.

Step 4: Investigate the causes and contributing factors (CHECK) This fundamental step answers the question "Why did this happen?". The tool used for root cause analysis synthesizes recognized frameworks (REEM, Savall, ISMP, ALARM, HAS, MERVEIL). The facilitator guides the exploration of causes by area of ​​investigation. For each problem, the causes and contributing factors are investigated, identifying why the barriers failed (Appendix I). The analysis focuses on eight areas of causes: health products, patients, healthcare professionals, operational tasks and procedures, team functioning, work environment, organization and management, and institutional context. More than 250 possible causes are listed. If necessary, the causes are prioritized using the cause weighting matrix (Appendix L).

Step 5: Synthesize the Action Plan (ACT) This second key step of the REMED process aims to answer the questions "What have we learned?" and "What changes should be implemented?". A climate of trust facilitates everyone's participation. Additional information (internal feedback, literature review) is shared. The next step is to identify the prevention, interception, or recovery actions to be implemented. The action plan is developed to track the progress of the selected measures (Appendix J). If necessary, improvement actions are prioritized by assessing their feasibility and their link to the risk of recurrence of the error (Appendix M). Communication and information procedures for teams and recipients are also established.

Step 6: Closing the Meeting (ACT)At the end of the session, the summary report is read and validated, ensuring its complete anonymization (Appendix K). Reports to the relevant vigilance systems are recorded if necessary. The action plan decided upon during previous meetings is reviewed. Medication error cases for the next meeting are selected. Finally, participants are registered on an attendance sheet.

Step 7: After the Meeting (ACT)This post-meeting step is crucial for the sustainability of the process. Feedback is provided by sending the report to each participant and guest. Reports to institutions are made if necessary. Monitoring and evaluation of the action plan are being implemented. La REMED est enregistrée dans le bilan annuel, et le cas d'erreur médicamenteuse ainsi que les actions d'amélioration sont consignés dans les tableaux de suivi correspondants.

Les outils indispensables à la REMED : du classeur aux matrices d'analyse

La méthode REMED est accompagnée d'une panoplie d'outils spécifiquement conçus pour faciliter son organisation, sa conduite et l'exploitation des données qu'elle génère. L'outil central est le Classeur de la REMED, disponible en version électronique et téléchargeable sur le site de la SFPC (www.sfpc.eu).

Le Classeur de la REMED est un support interactif composé de plusieurs onglets :

• Onglet 1 - Présentation du classeur de la REMED : Fournit une introduction à la REMED, son histoire, ses auteurs et des définitions clés.
• Onglet 2 - Mode d'emploi du classeur de la REMED : Décrit les consignes d'utilisation des différents onglets, notamment les codes couleur pour la saisie des informations.
• Onglet 3 - Cahier de la REMED : C'est l'outil pivot. Il guide le déroulement de la REMED, de la préparation du cas à la séance pluri-professionnelle. Il sert de compte rendu détaillé, renseigné au fur et à mesure de l'étude du cas.
• Onglet 4 – Liste de questions : Aide à structurer les entretiens avec les professionnels et les discussions lors de l'analyse des causes et de la recherche d'actions d'amélioration, en s'appuyant notamment sur la méthode QQOQCCP (Qui ? Quoi ? Où ? Quand ? Comment ? Combien ? Pourquoi ?).
• Onglet 5 – Liste des documents utiles : Un mémento pour identifier les documents pertinents nécessaires à l'objectivation des constats.
• Onglet 6 - Caractérisation de l'EM : Permet de déterminer les six éléments signifiants d'une erreur médicamenteuse pour sa classification.
• Onglet 7 – Liste des causes : Propose plus de 250 causes possibles, classées en 8 domaines d'investigation (produits de santé, patient, professionnels de santé, tâches et procédures opérationnelles, fonctionnement de l’équipe, environnement de travail, organisation et management, contexte institutionnel). Cette liste synthétise diverses méthodes d'analyse des risques.
• Onglet 8 – Tableau des actions d'amélioration : Permet de suivre la mise en œuvre des actions d'amélioration décidées. Les informations saisies dans le Cahier de la REMED se transfèrent automatiquement dans cet onglet.
• Onglet 9 - Compte rendu synthétique : Génère un compte rendu anonymisé et synthétique du cas analysé, support de communication.

En complément de ces outils de base, le classeur intègre également des outils optionnels, dits "experts" :

• Onglet 10 – Matrice de pondération des causes (Annexe L) : Permet de hiérarchiser les causes identifiées en fonction de l'intensité de leur lien avec l'erreur produite, de manière objective et sans jugement de valeur.
• Onglet 11 – Matrice de priorisation des actions d'amélioration (Annexe M) : Aide à prioriser les actions d'amélioration en évaluant leur faisabilité et l'intensité de leur lien avec le risque de reproduction de l'erreur.

Ces outils sont conçus pour être évolutifs et ne pas contraindre les professionnels de santé à un formalisme excessif, tout en assurant l'harmonisation de la démarche REMED.

En plus du classeur, d'autres documents complémentaires sont proposés pour renforcer la gestion des risques associés aux produits de santé :

• La liste des "événements qui ne devraient jamais arriver" (never events), fournie par l'ANSM.
• La liste des événements sentinelles associés aux produits de santé, établie par la SFPC.
• Le scénario 5C d'une erreur médicamenteuse (Causes-Conséquences, Corrections, Chronologie, Conséquence Résiduelle). Cet outil pédagogique schématise l'enchaînement des faits, soulignant la complexité et la temporalité des événements.
• La charte d'encouragement au signalement des événements indésirables, proposée par le Centre hospitalier de Lunéville.
• Un tableau de suivi des erreurs médicamenteuses signalées dans l'établissement.
• Le programme Médicament & Sécurité Thérapeutique (Med'Seth), qui facilite le recensement et le suivi des actions d'amélioration de la prise en charge médicamenteuse.

L'ensemble de ces ressources est téléchargeable sur le site internet de la SFPC, offrant aux professionnels les moyens nécessaires pour organiser, conduire et exploiter les analyses des erreurs médicamenteuses de manière structurée et efficace.

La REMED au service de la culture de sécurité et du développement professionnel continu

La REMED ne se limite pas à être une simple méthode d'analyse ; elle est un puissant levier pour transformer la culture de sécurité au sein des établissements de santé et contribuer activement au développement professionnel continu des soignants. La prise de conscience de la gestion des risques sanitaires s'affirme progressivement, et la REMED s'inscrit dans cette dynamique en fournissant un cadre pour aborder les erreurs de manière constructive.

Historiquement, affronter les erreurs était difficile pour les professionnels, souvent perçues avec un caractère culpabilisant. L'OMS a pourtant souligné dès 2002 que les causes des erreurs sont, dans plus de 75% des cas, d'origine systémique. La REMED adopte précisément cette approche systémique et non culpabilisante. Elle encourage à considérer les personnes impliquées dans la survenue d'une erreur comme des personnes ressources pour l'amélioration de la qualité et de la sécurité des soins, un facteur indispensable au succès et à la pérennité de la démarche. Cette vision permet de dépasser la seule réflexion centrée sur les individus pour tirer des enseignements sur les forces et les vulnérabilités du système de soins.

La REMED favorise une démarche pédagogique essentielle pour le renforcement d'une culture de sécurité :

• Elle permet une meilleure compréhension de la complexité du processus de prise en charge médicamenteuse des patients.
• Elle contribue à l'amélioration des pratiques professionnelles liées aux soins médicamenteux.
• Elle assure le perfectionnement des connaissances grâce au retour d'expérience.
• Elle renforce la communication entre professionnels de santé, un élément clé dans la prévention des erreurs.
• Elle s'inscrit dans l'amélioration continue de la qualité et de la sécurité de la prise en charge médicamenteuse, ainsi que dans la maîtrise et la gestion des risques liés aux produits de santé.

En France, la loi de Santé Publique de 2004 a érigé les événements indésirables graves en problème de santé publique, soulignant l'importance de la diminution des événements indésirables liés aux médicaments. La REMED s'aligne parfaitement avec ces objectifs. Elle est également citée dans le Guide méthodologique de la HAS sur la sécurité des patients.

La participation active aux REMED permet aux professionnels de valider en partie leurs obligations individuelles dans le cadre du Développement Professionnel Continu (DPC). En effet, la REMED est l'une des méthodes spécifiquement listées par la Haute Autorité de Santé (HAS) dans son document "Développement professionnel continu – Méthodes et modalités de DPC" de décembre 2012. Elle participe également aux démarches de certification des établissements de santé et au management de la qualité et des risques, ainsi qu'à leurs dispositifs contractuels (contrats pluriannuels d'objectifs, contrats de bon usage).

La REMED s'inscrit dans les dispositions relatives aux droits des patients et à la qualité et sécurité des soins, que le patient soit hospitalisé, résident ou ambulatoire. C'est un outil qui appuie la mise en œuvre des exigences de l'arrêté du 6 avril 2011 relatif au management de la sécurité des soins médicamenteux du patient hospitalisé.

En conclusion, la REMED est un "remède à l'erreur" en transformant l'erreur, inhérente à toute activité humaine organisée, en un outil pédagogique. Le soin médicamenteux étant un processus complexe, la REMED offre une méthode pluriprofessionnelle pour l'améliorer, en se basant sur la pédagogie par l'erreur et en prenant en compte l'ensemble du processus de prise en charge du patient. C'est un travail collectif dont la réussite, mesurée par la satisfaction des utilisateurs et la pertinence des réponses et pistes fournies, bénéficiera directement aux patients. La REMED est donc un pilier pour une gestion des risques médicamenteux proactive et une culture de sécurité en constante évolution.

Source

https://sfpc.eu/wp-content/uploads/2019/11/Manuel_de_la_remed-_jan_2013_-_version_dfinitive_300114-copie.pdf