EDN and ECOS: how France tests clinical reasoning

Table of contents
  1. What the EDN and the ECOS are
  2. How the EDN work
  3. How the ECOS work
  4. How scoring and thresholds work
  5. How clinical reasoning shows up
  6. What the studies say
  7. How to train reasoning for the EDN and the ECOS
  8. How NEXO helps you train
  9. Frequently asked questions
  10. References

An intern enters a room with eight minutes on the clock. The voiceover, read from outside, says he is in the emergency department and must announce a serious diagnosis to a family member. The encounter is with a standardized participant and two examiners who do not respond to requests for help. The next day, the same candidate opens a tablet and faces a progressive dossier with pointing-zone questions, QROC and critical reading of an article. The two scenes belong to the same gateway to the third cycle in France: the Épreuves Dématérialisées Nationales, the EDN, and the Examens Cliniques Objectifs Structurés, the ECOS.

NEXO is a daily medical diagnosis game: one clinical case a day, six progressive clues, and ICD-10 proximity feedback across five levels.

This combination sums up what the French reform wants to measure: knowledge on the tablet and competence in the station. This text explains what the EDN and the ECOS are, how each exam works, how the thresholds are set, what the studies say and how to train. At the end there is a table, frequently asked questions and a checklist.

Key points

  • The EDN and the ECOS replaced the former épreuves classantes nationales as the national assessment for access to the third cycle of the études de médecine (MESR; arrêté of 21 December 2021).
  • In the first session, the EDN bring together four three-hour exams; only those who reach at least fourteen out of twenty in the connaissances de rang A move on to the ECOS (CNG EDN).
  • The national ECOS have ten stations over two days, with eight minutes of épreuve and two minutes of movement; the threshold is ten out of twenty in the sum of the scores of the ten stations (CNG ECOS; arrêté de abertura 2026).
  • The rang A and rang B scores of the EDN account for sixty percent of the note globale of the dossier for the appariement; the scores of the second EDN session do not enter that dossier (arrêté of 21 December 2021).

What the EDN and the ECOS are

The Épreuves Nationales for access to the third cycle of the études de médecine in France have two sides. The épreuves dématérialisées nationales (EDN) assess knowledge in digital format. The examens cliniques objectifs structurés (ECOS) assess competences in standardized clinical stations. The code de l'éducation describes this pair: national assessment of knowledge in dématérialisée form and national assessment of competences in the form of ECOS (articles R. 632-2 and following).

The Ministry of Higher Education presented the reform as replacing the ECN with EDN, ECOS and greater value for the parcours de formation (MESR). The detailed organization is in the arrêté of 21 December 2021. The CNG oversees the calendar, the convocations and the Even platform (CNG campagne).

Where each piece fits in Miller's pyramid

In 1990, George Miller described four steps: know, know how, show how and do (Miller, 1990). The EDN mostly ask for the second step; the ECOS rise to the third. It is the logic that separates a written exam such as the USMLE Step 2 CK or the Examen MIR from a clinical exam such as the AMC Clinical Exam or the second stage of Revalida. France requires both in the same path.

How the EDN work

According to the CNG, the first session of the EDN brings together four three-hour exams (CNG EDN). The arrêté of 21 December 2021 details the structure.

Four plages and question formats

Three plages horaires of three hours mix questions isolées and dossiers progressifs with problèmes a elemento-chave. Each dossier progressif has three to eight questions. Each of these plages counts between 90 and 110 questions and forms an independent unité de composition, with the same valeur as the others (arrêté of 21 December 2021).

The fourth plage is the lecture critique d'articles, the LCA: two unités of one and a half hours, each with a scientific article (orientation clinique and orientation physiopathologique). Each lecture critique has 13 to 17 progressive questions. The notação of the LCA questions has pondération double face às questions of the three other unités (arrêté of 21 December 2021).

The planned categories include QRM or QRU with four to five propositions, a drop-down menu, zones to point to, QROC (five words maximum) and script concordance tests (arrêté de 21 de dezembro de 2021). The conseil scientifique published scoring rules (discordance in QRM, binary in QRU, among others) (docimologia CNG).

Second session

If the candidate does not reach the rang A threshold in the first session, there is a second session in the same academic year, before the ECOS, organized by the universities. The CNG describes two demi-journées of three hours with 180 to 220 questions, without an épreuve de LCA (CNG EDN). Only those who reach fourteen out of twenty in rang A in one of the sessions take part in the ECOS.

Syllabus and annales

The syllabus is that of the deuxième cycle des études de médecine (CNG EDN). The conseil scientifique prepares and checks the questions. The CNG makes annales from the first sessions available from October 2023, without answer keys, as practice and not as a corpus to memorize (CNG EDN).

Old hand and forearm prosthesis, with a wooden or composite hand in a resting position and a dark leather arm with straps and metal rings, placed diagonally on a reflective black background
European hand and forearm prosthesis. In ECOS stations with a procedure or examination, the candidate must show the gesture and narrate the reasoning, not just name the diagnosis. Credit: Wellcome Collection, CC BY 4.0 (licença), via Wikimedia Commons (página do arquivo).

How the ECOS work

The national ECOS are a session unique with ten stations split into deux épreuves of five consecutive stations per day (CNG ECOS). The arrêté de abertura para 2026 sets the durée unitaire at ten minutes: eight minutes of épreuve including reading the vignette and two minutes to move between stations. The guide du candidat confirms: eight minutes per station, a sound signal at the start and end, two minutes until the next one (guide ECOS 2026).

Domains and station types

Each station covers a situation de départ and a domaine. The ECOS-N cover ten of eleven domains (annonce, communication interprofessionnelle, éducation / prévention, entretien, examen clinique, iconographie, procédure, stratégie diagnostique, prise en charge, synthèse paraclinique, urgence) (guide ECOS 2026).

The vignette states whether the station is AVEC PS (standardized patient, parent or proche), AVEC PSS (personnel de santé standardisé) or SANS PS ni PSS. With PS or PSS there is no physical contact; without them, the candidate is asked to think aloud or describe the examination and iconography. In simulation, precautions and gestures are described aloud, without speaking to the mannequin as if to a person (guide ECOS 2026). The scénarios start from the situations cliniques de départ of the régime des études and are set by the conseil scientifique (CNG ECOS).

Examiners and anonymity

There are examiners from outside the host university (arrêté de 21 de dezembro de 2021). The guide asks for anonymous presentation (prénom = day; nom = month). There is no interaction with the examiners (guide ECOS 2026).

Who can take the ECOS

Only students who reached the threshold of fourteen out of twenty in connaissances de rang A in the EDN enter the ECOS (CNG EDN). The centre de épreuves is the same one assigned for the EDN of the same campagne (CNG ECOS).

How scoring and thresholds work

EDN threshold (rang A)

The note minimale de rang A is fourteen out of twenty (arrêté de 21 de dezembro de 2021; CNG EDN). The rang A / rang B split follows the annexe of the régime des études. For the note validante, the docimologia of the conseil scientifique considers rang A questions in QRU, QRP, zone or QROC (docimologia CNG).

Weight in the appariement dossier

The rang A and rang B notes make up sixty percent of the note globale of the dossier for the appariement. The notes from the second session do not count: only those from the first do (arrêté de 21 de dezembro de 2021). The pondération double of the LCA changes the time strategy in the fourth plage.

ECOS threshold

The sum of the notes uniques non pondérées of the ten stations must be equal to or greater than ten out of twenty (arrêté de 21 de dezembro de 2021; CNG ECOS). Those who fall below have no second session in the same year; they can return the following year under the conditions of the arrêté. This note is one of the elements of the appariement dossier.

Item What the rule sets Official threshold or weight
EDN, first session Four plages of three hours (three of dossiers/QI + LCA) Rang A: fourteen out of twenty to go to the ECOS
EDN, second session Two demi-journées; no LCA (CNG) Same rang A threshold; notes do not count for the appariement
LCA Pondération double of the questions Part of the EDN note
ECOS Ten stations, eight + two minutes Ten out of twenty in the sum of the ten stations
Dossier appariement Rang A + rang B of the EDN (first session) Sixty percent of the note globale
Copy of an Etruscan dental bridge in wavy gold with natural teeth attached, lying diagonally on a white background
Copy of an Etruscan gold dental prosthesis. In questions of stratégie diagnostique or procédure, the ECOS vignette and the EDN dossier ask for the justified next step, not just the disease label. Credit: Wellcome Collection, CC BY 4.0 (license), via Wikimedia Commons (file page).

How clinical reasoning shows up

On the EDN tablet

A dossier progressif releases information little by little. Zona a apontar and QROC force precision. The LCA asks you to read the article and answer about method and interpretation. Eva described reasoning as gathering information, generating hypotheses and testing them (Eva, 2005). The EDN measures this digitally: the task is in the prompt and the distractor is usually plausible.

In the dossier progressif, the order of the questions matters. An early answer can open or close later information; the arrêté documentation describes dossiers with three to eight questions (arrêté of December 21, 2021). Training as if each item were isolated loses the habit of carrying hypotheses through the case. Zona a apontar and QROC require naming the finding or the step in few words: the mistake here is usually knowing the idea and failing at the wording.

The LCA changes the type of reasoning. Instead of a fictional patient, the candidate reads an article and answers about design, biases and interpretation. The pondération double makes this plage strategic: time badly spent on the first three unités leaves critical reading under pressure (arrêté of December 21, 2021). Anyone who only memorizes conclusions from reviews and never opens the article's method feels the gap in this session.

At the ECOS station

The vignette says what you should and should not do. With PS, reasoning goes through communication; without PS, it must be said out loud; with simulation, gesture and narration count together (ECOS 2026 guide). Harden described the OSCE to make observable what used to stay inside a long case (Harden et al., 1975; Harden and Gleeson, 1979).

The candidate guide makes the register explicit: with PS or PSS there is no physical contact; without them, reasoning must be said out loud; in simulation, precautions and gestures are described without treating the mannequin as an interlocutor (ECOS 2026 guide). Mixing these registers is a frequent mistake in local practice sessions. The vignette also defines the domaine (annonce, urgence, procédure, stratégie diagnostique, among others): reading only the expected diagnosis and ignoring the requested domain changes the examiner's grid.

The two minutes of movement are not a study break. They are time to leave the previous role and enter the next without carrying the approach from the annonce station into the procédure one. The sound signal marks start and end; anyone who keeps talking after the end loses control of the next station (ECOS 2026 guide).

Time and discipline

In the EDN, three hours per unité and the pondération double of the LCA demand pace and attention to format. In the ECOS, eight minutes include the reading (ECOS 2026 guide). Anyone who spends too long on the vignette loses the dialogue; anyone who does not read the last line gets the language register wrong.

What the studies say

The OSCE began in Dundee as several short stations with combined criteria (Harden et al., 1975; Khan et al., 2013). In France, papers discuss performance factors in ECOS and the use of the format in specialties (2022; ophthalmology; pathology).

Miller places multiple choice at knowing how and the OSCE at showing how (Miller, 1990). A consensus on diagnostic reasoning insists on aligning the instrument with what is measured (Ilgen et al., 2012). In other written exams, performance was associated with measures in residency (Sharma et al., 2019; Morrison et al., 2014).

The practical reading: the EDN measure application of knowledge; the ECOS measure what is shown at the station. Training only on summaries leaves out gesture, voice and time.

French studies on OSCE performance and on the use of the format in specialties reinforce the operational obvious: the station has observable criteria and the candidate must show the item, not just know its name (PubMed 35292159; ophthalmology; pathology). The consensus by Ilgen et al. on assessing diagnostic reasoning calls for aligning the instrument with what is being measured (Ilgen et al., 2012): that is why France does not treat the tablet and the station as the same exam in different clothes.

For those who also train for written exams from other countries, the useful transfer is the habit of the next step and of telling distractors apart, not the scoring scale. The MIR punishes error explicitly; Step 2 CK and the MCCQE have their own logics. Copying the guess strategy of one jurisdiction to the French discordance or to the ECOS grid is an expensive shortcut.

How to train reasoning for the EDN and the ECOS

Training for French access to the third cycle means training decision-making on the tablet and performance in the station. Official references: arrêté of 21 December 2021, CNG, past papers, docimology and the ECOS guide.

EDN: formats before volume

Work through past papers paying attention to the question type. QRM with discordance is not trained like a binary QRU. Zona and QROC demand precision. The LCA deserves timed sessions. Classify the error: rang A, reading, format or haste.

ECOS: out loud and vignette

Rehearse eight-minute stations with the vignette outside and read to the last line. In communication, train reformulation; without a PS, narrate out loud; in simulation, precautions and the gesture. See also the AMC Clinical Exam, the second stage of Revalida, PLAB and UKMLA and the MCCQE.

Set up a circuit of five stations on the same day, with two minutes between them, to approximate the load of the deux épreuves de cinq stations of the national format (CNG ECOS). Record audio of your narration: what seems clear in your head usually comes out incomplete on first listening. Ask a colleague to mark only what the vignette asked for, not what you know beyond it.

Typical errors to classify after a block

After each timed session, label the error before going back to the summary. Without that label, the volume of past papers does not correct the pattern.

  • Incomplete reading of the vignette (ignored AVEC PS, AVEC PSS or SANS PS, or the last line of the task).
  • EDN format: treated QRM with discordance as if it were a binary QRU, or exceeded the five words of the QROC (CNG docimology).
  • Early closure in the dossier: fixed a diagnosis at question 1 and did not update it when question 4 brought a contrary finding.
  • Wrong register in the ECOS: spoke to the mannequin, asked the examiner for help or omitted precautions in the procédure (ECOS guide 2026).
  • Time management: spent too long on the LCA or on reading the vignette and lost the final questions or the dialogue with the PS.
  • Confusion between threshold and classement: trained only to pass at rang A and forgot that rang A and rang B of the first session weigh on the appariement dossier (arrêté of 21 December 2021).

Do not mix strategies from other countries

The MIR punishes error; the EDN use discordance, binary and QROC. The ECOS have their own grid and rules. When using foreign question banks, adjust to the French rule. Compare with the Examen MIR and the USMLE Step 2 CK without copying the scoring.

Old Anel lacrimal syringe with an open lined case and interchangeable cannulas, in a black and white photograph
Dominique Anel lacrimal syringe and case. In procédure stations, the vignette usually asks for the sequence and the precautions, not just the name of the instrument. Credit: Wellcome Collection, CC BY 4.0 (license), via Wikimedia Commons (file page).

Checklist for the final stretch

  • I have read the arrêté of 21 December 2021 (or the consolidated version in force) and the page for my campagne on the CNG.
  • I know that the first session of the EDN has four épreuves of three hours and that the LCA has pondération double.
  • I know the rang A threshold: fourteen out of twenty to access the ECOS.
  • I know that the grades of the second EDN session do not count for the appariement.
  • I have read the guide du candidat aux ECOS: eight minutes, two for moving, ten stations, threshold ten out of twenty.
  • I have trained reading the vignette up to AVEC PS / AVEC PSS / SANS PS and introducing myself anonymously.
  • I have done timed sessions of dossier progressif and of LCA.
  • I have an error notebook by format (QRM, QRU, zona, QROC, TCS) and by type of failure in the station.
  • I know that the ECOS center is the same as the EDN of the campagne.
  • I have already looked at how the dossier (EDN + ECOS + parcours) feeds the appariement, without confusing threshold with classification.

Want to train differential diagnosis between a block of EDN and a station test? Play the case of the day on NEXO.

How NEXO helps you train

NEXO offers one clinical case per day, with fictional cases with a real clinical presentation, in up to six clues. It is a short differential diagnosis workout between a dossier and a station of stratégie diagnostique.

It does not replace annales or the ten-station circuit, but it exercises putting data together and not closing too early. See the NEXO Economy, the NEXO Closer, the NEXO Researcher, the NEXO Curator and the NEXO Creators.

Frequently asked questions

What are the EDN?

They are the épreuves dématérialisées nationales that assess knowledge for entry to the third cycle of études de médecine in France (code de l'éducation; CNG EDN).

What are the ECOS?

They are the examens cliniques objectifs structurés nationaux: ten stations in session unique that assess standardized clinical skills (CNG ECOS).

What is the threshold to move from the EDN to the ECOS?

Fourteen out of twenty in the connaissances dites de rang A (arrêté de 21 de dezembro de 2021; CNG EDN).

What is the ECOS threshold?

Ten out of twenty in the sum of the notes uniques non pondérées of the ten stations (CNG ECOS).

Does the second EDN session count for the appariement?

No. Only the scores from the first session enter the processus d'appariement (arrêté de 21 de dezembro de 2021).

References

  1. Centre national de gestion. Les Epreuves Dématérialisées Nationales (EDN) campagne 2026. [https://www.cng.sante.fr/epreuves-dematerialisees-nationales-edn-campagne-2026](https://www.cng.sante.fr/epreuves-dematerialisees-nationales-edn-campagne-2026)
  2. Centre national de gestion. Les Examens Cliniques Objectifs Structurés (ECOS) campagne 2026. [https://www.cng.sante.fr/examens-cliniques-objectifs-structures-ecos-campagne-2026](https://www.cng.sante.fr/examens-cliniques-objectifs-structures-ecos-campagne-2026)
  3. Ministère de l'Enseignement supérieur. Arrêté du 21 décembre 2021 relatif à l'organisation des épreuves nationales donnant accès au troisième cycle des études de médecine. JO 28 décembre 2021. [PDF](https://ufr3s.univ-lille.fr/fileufr3s/user_upload/ufr3s-formations/medecine-scolarite/R2C/textes-reglementaires/JO-2021-12-28_EDN_Organisation.pdf)
  4. Arrêté du 13 novembre 2025 portant ouverture des ECOS au titre de l'année universitaire 2026-2027. JO 19 novembre 2025. [PDF](https://ufr3s.univ-lille.fr/fileufr3s/user_upload/ufr3s-formations/medecine-scolarite/R2C/joe_20251119_ECOS2026.pdf)
  5. CNG. Guide du candidat aux ECOS nationaux 2026. [PDF](https://www.cng.sante.fr/sites/default/files/media/2026-04/guide-du-candidat-aux-ecos-nationaux_2026.pdf)
  6. Conseil scientifique en médecine / COSUI R2C. Règles générales de notation (docimologie ED). [PDF](https://www.cng.sante.fr/sites/default/files/media/2024-10/r2c-docimologie-et-correction-valide-cosui-24-25.pdf)
  7. Ministère de l'Enseignement supérieur. Rentrée 2021: poursuite de la réforme des études de santé. [https://www.enseignementsup-recherche.gouv.fr/fr/rentree-2021-poursuite-de-la-reforme-des-etudes-de-sante-50984](https://www.enseignementsup-recherche.gouv.fr/fr/rentree-2021-poursuite-de-la-reforme-des-etudes-de-sante-50984)
  8. Code de l'éducation, accès au troisième cycle des études de médecine (R. 632-2 et s.). [Legifrance](https://www.legifrance.gouv.fr/codes/section_lc/LEGITEXT000006071191/LEGISCTA000027864857)
  9. Miller GE. The assessment of clinical skills/competence/performance. Acad Med. 1990. [https://pubmed.ncbi.nlm.nih.gov/2400509/](https://pubmed.ncbi.nlm.nih.gov/2400509/)
  10. Eva KW. What every teacher needs to know about clinical reasoning. Med Educ. 2005. [https://pubmed.ncbi.nlm.nih.gov/15612906/](https://pubmed.ncbi.nlm.nih.gov/15612906/)
  11. Harden RM, Stevenson M, Downie WW, Wilson GM. Assessment of clinical competence using objective structured examination. Br Med J. 1975. [https://pubmed.ncbi.nlm.nih.gov/1115966/](https://pubmed.ncbi.nlm.nih.gov/1115966/)
  12. Harden RM, Gleeson FA. Assessment of clinical competence using an objective structured clinical examination (OSCE). Med Educ. 1979. [https://pubmed.ncbi.nlm.nih.gov/763183/](https://pubmed.ncbi.nlm.nih.gov/763183/)
  13. Khan KZ, et al. The Objective Structured Clinical Examination (OSCE): AMEE Guide No. 81. Part I. Med Teach. 2013. [https://pubmed.ncbi.nlm.nih.gov/23968323/](https://pubmed.ncbi.nlm.nih.gov/23968323/)
  14. Ilgen JS, et al. Assessing diagnostic reasoning: a consensus statement. Acad Emerg Med. 2012. [https://pubmed.ncbi.nlm.nih.gov/23279251/](https://pubmed.ncbi.nlm.nih.gov/23279251/)
  15. Identification of factors associated with performance in OSCE. Rev Med Interne. 2022. [https://pubmed.ncbi.nlm.nih.gov/35292159/](https://pubmed.ncbi.nlm.nih.gov/35292159/)
  16. Setting up OSCE in ophthalmology. J Fr Ophtalmol. 2022. [https://pubmed.ncbi.nlm.nih.gov/35738961/](https://pubmed.ncbi.nlm.nih.gov/35738961/)
  17. Promoting pathology and cytology in OSCE. Ann Pathol. 2025. [https://pubmed.ncbi.nlm.nih.gov/40360355/](https://pubmed.ncbi.nlm.nih.gov/40360355/)
  18. Sharma A, et al. USMLE Step 2 CK and residency performance. J Grad Med Educ. 2019. [https://pubmed.ncbi.nlm.nih.gov/31440335/](https://pubmed.ncbi.nlm.nih.gov/31440335/)
  19. Morrison CA, et al. NBME CCSSA and USMLE Step 2 CK. Teach Learn Med. 2014. [https://pubmed.ncbi.nlm.nih.gov/25318033/](https://pubmed.ncbi.nlm.nih.gov/25318033/)

Back to the blog