EPAs: entrustable professional activities in Medicine

Monday, seven in the morning, the intern's first day on the ward. The preceptor hands over the patient list and says: "today you take the history of bed twelve and present it to me before rounds". On Wednesday, the sentence changes: "follow up the three patients in the room and call me if anything goes off script". On Friday, another change: "hand over bed twelve to the night team, I'll just listen". Nobody wrote a form, nobody gave a grade, but an important decision was made three times in the same week. Each day, someone decided how much of that task could be entrusted to the student.

This decision, which has always existed silently, gained a name and structure over the last two decades: entrustable professional activities, known by the acronym EPAs (entrustable professional activities). This guide explains what they are, where they came from, how the levels of supervision work, what weighs in the decision to trust, what the large American pilot showed, what the most recent research adds and, above all, how you, a student or intern, can use the logic of EPAs to learn more and faster during your rotation.

Key points

• An EPA is a unit of real clinical work, observable and complete, that can be entrusted to a learner when they show the necessary competence.
• Assessment stops being only "how much you know" and becomes "how much supervision you need to do this safely".
• The levels range from observing only to supervising a junior colleague, passing through direct and indirect supervision.
• Trust depends on competence, but also on honesty, on recognizing your own limits and on asking for help at the right time.
• In the American pilot, the decision to trust was only possible when there were enough observations in the workplace.
• Diagnostic reasoning runs across several EPAs, and training it outside the shift speeds up trust within it.

Contents

• What EPAs are
• Where the idea came from
• The supervision levels
• What weighs in the decision to trust
• What the American pilot showed
• Limits and criticism
• How the student can prepare
• How NEXO helps train the reasoning behind EPAs
• Frequently asked questions
• Checklist to gain trust on the rotation

What EPAs are

An entrustable professional activity is a clinical practice task that a professional performs from start to finish and that can be handed to someone in training when it is safe to do so. Taking a history and doing a physical exam, building a list of differential diagnoses, ordering and interpreting tests, prescribing, documenting an encounter, presenting a case orally, handing over a shift: each of these tasks is a possible EPA. What unites them is that all can be seen by another person, have a recognizable result and require several competencies at the same time.

That last characteristic is the key. Competencies such as communication, medical knowledge, professionalism and clinical reasoning are qualities of a person. EPAs are descriptions of the work. A well-done shift handover, for example, mixes knowledge about the patient, prioritization, clear communication and awareness of risks. Assessing the shift handover is assessing several integrated competencies, the way they appear in real life.

An EPA is not a procedure checklist

A common mistake is to imagine that an EPA is a list of procedures, such as "placing an IV line" or "suturing". Procedures can be part of an EPA, but the idea is broader. The question is not whether the student can perform an isolated technique, but whether they can be responsible for part of the patient's care with a certain degree of supervision. The AMEE guide on the topic insists that assessment in the workplace translates into trust decisions for defined levels of autonomy, and not into loose grades on skills.

Examples of EPAs in undergraduate training

The best-known set for undergraduate education is the thirteen core EPAs for entry into residency, published by the association of American medical schools in 2014. It includes, among others, taking a history and performing a physical exam, prioritizing differential diagnoses after an encounter, recommending and interpreting tests, prescribing, documenting, presenting cases, formulating clinical questions and seeking evidence, handing over and receiving shifts, collaborating in a team, recognizing a patient who needs urgent attention, obtaining consent and identifying system failures that threaten safety. Other countries created their own lists, adapted to their curricula, but the logic is the same.

Example of an EPA	Competencies involved	How it can be observed
Taking a history and performing a physical exam	Communication, knowledge, clinical reasoning	Direct observation of an encounter
Prioritizing differential diagnoses	Clinical reasoning, knowledge, managing uncertainty	Case presentation with justified hypotheses
Recommending and interpreting tests	Probabilistic reasoning, rational use of resources	Discussion of the diagnostic plan
Documenting the encounter	Written communication, organization of reasoning	Review of the progress note in the medical record
Handing over and receiving shifts	Prioritization, communication, patient safety	Observation of the handover between teams
Recognizing the critically ill patient	Knowledge, judgment, asking for help	Simulation and supervised real situations

Where the idea came from

The concept was proposed by the Dutch educator Olle ten Cate in 2005, in a short article that tried to solve a practical problem of competency-based education. Curricula were full of long lists of competencies, hard to observe in daily practice, and preceptors did not really know how to turn that into concrete decisions. The proposal was to invert the question: instead of assessing abstract competencies, assess professional activities that can be entrusted to someone in training.

Miller's pyramid and the level of "does"

Medical education has used Miller's pyramid for decades, and it separates four levels of assessment: knows, knows how, shows how and does. Written exams assess the first two well, and simulation assesses the third. The top, what the learner actually does at work, has always been the hardest to measure. Ten Cate and colleagues proposed extending the pyramid with the trust decision, arguing that the final question of training is whether that person can be responsible for an activity at defined levels of supervision, such as direct supervision, indirect supervision or none.

From residency to undergraduate education

EPAs were created with postgraduate training in mind, but they soon reached undergraduate education. The transition from the end of the course to residency is a known moment of risk: the recent graduate takes on tasks they often have never performed alone. Defining which activities the student should be able to perform with indirect supervision on the first day of residency was the way found to make this transition safer and more transparent.

The levels of supervision

At the heart of EPAs is a supervision scale. It describes how much of the supervisor's presence the learner needs to perform the activity safely. There are variations between programs, but the most used structure has five levels.

1. Can observe, without performing the activity.
2. Can perform with direct supervision, with the supervisor present in the room.
3. Can perform with indirect supervision, with the supervisor available and reviewing afterwards.
4. Can perform without supervision, as an autonomous professional.
5. Can supervise another, more junior person in the same activity.

This scale changes the conversation between preceptor and student. Instead of "you did well" or "you need to study more", the question becomes concrete: tomorrow, on this task, do I need to be in the room or can I just review afterwards? The answer guides what the student should practice and makes progress visible.

Point-in-time trust and formal trust

Researchers distinguish two types of decision. Point-of-care trust (ad hoc) happens every day, when the preceptor decides, on that shift, to let the intern conduct a consultation. Formal trust (summative) is a recorded decision, made by a group based on several observations, which certifies that the learner can perform that activity with a given level of supervision from then on. Good practice recommends that the formal decision be made by a trained group, based on longitudinal assessments from several evaluators and incorporating day-to-day decisions.

When trust comes too early

Point-of-care trust can also fail. A qualitative study with residents in Switzerland described that, in practice, the decision to trust often happens by default, without explicit conversation. Residents reported situations of excessive trust, when they receive tasks for which they do not feel ready, and of insufficient trust, when they are stuck with tasks they already master. The authors argue that these default decisions should be discussed openly, to avoid early trust and to meet the real need for supervision.

What weighs on the decision to trust

If trust depended only on knowledge, a test would be enough. That is not the case. Studies on trust decisions show that supervisors take into account a set of factors, many of them linked to the learner's behavior at work.

• Competence in the task, observed more than once and in different contexts.
• Integrity, that is, reporting what was done and what was not done without dressing up the data.
• Awareness of one's own limits and willingness to ask for help.
• Reliability: keeping to what was agreed, coming back to see the patient, closing pending items.
• Characteristics of the task and of the patient, such as severity, complexity and time of day.
• Characteristics of the supervisor, such as experience, style and workload.

Knowing one's own limits counts for a lot

The group that guided the American pilot highlighted one point: to trust, the supervisor needs evidence that the student knows their limits and seeks help when necessary. This has a practical consequence that surprises many students. Saying "I don't know" or "I'm not sure, can you look at it with me?" does not reduce trust. On the contrary, it is one of the signals that most increase it, because it shows that the person will not go it alone when outside the safe zone.

Trust in the role, support in the tasks

In internal medicine, where there are fewer procedures, the relationship between trust and supervision is more subtle. A study with supervisors of senior residents on Canadian wards showed that they trust the resident in the team leader role after an initial period of close assessment, followed by a gradual withdrawal of support to promote independence. Trust, therefore, is not a switch. It is built in layers, and the supervisor can trust the overall role while continuing to support specific tasks.

What the American pilot showed

In 2014, the association of American medical schools invited ten schools to test, over several years, whether it was feasible to teach and assess the thirteen core EPAs in undergraduate education. This pilot produced a series of studies that help separate what works from what is still a promise.

The first trust decisions

In 2019, trained groups from four schools reviewed the assessment data of graduating students and made theoretical trust decisions, with no consequence for graduation. Of the specific decisions per EPA, 41.3% were "ready" for indirect supervision, and in 28.4% it was not possible to decide, generally for lack of data. For ten of the thirteen EPAs, having assessments in the workplace available was associated with being able to make the decision. In other words, without real observation, there is no way to trust.

The second year of data

The following year, the same schools improved their systems. The proportion of "ready for indirect supervision" decisions rose from 43.4% in 2019 to 60.1% in 2020, and the proportion of decisions supported by four or more workplace assessments also grew. Even so, five EPAs lagged behind, with few students considered ready and few assessments available: prescribing, handover, urgent care, informed consent and patient safety. These are precisely activities that the student rarely performs alone in undergraduate education.

Revisiting the list

When analyzing the performance of each EPA, the authors of the pilot concluded that some activities worked well as assessment units and others did not. For the ones that had problems, they discussed whether the difficulties were practical, such as a lack of observation opportunities, or theoretical, such as a poorly defined activity. This critical review of the thirteen EPAs shows that the list is not a dogma, but a tool in adjustment.

The schools' experience

Interviews with the teams from nine pilot schools identified what made implementation easier: team commitment, adoption of the EPAs alongside a curriculum reform and the fact that the activities fit naturally into clinical placements. The schools also reported that the process forced them to reflect on their own curriculum and to adjust how they assess. First-year residents trained at these schools were also asked how prepared to perform the EPAs with indirect supervision they felt and how the transition to residency went.

Limits and criticisms

The EPAs spread quickly, and the literature grew along with them. A recent systematic review included hundreds of studies and found great diversity of methods for developing, implementing and evaluating EPAs. Most refer to postgraduate training and to the United States, and the quality of the studies was considered modest, with few papers on cost and effectiveness. This does not invalidate the idea, but it calls for caution about promises.

Bureaucracy and forms

A frequent criticism is that the EPAs can become just another form to fill in. When the preceptor has to record dozens of assessments per student, the temptation is to tick everything automatically, which destroys the value of the information. Successful programs invest in short forms, simple apps and training of assessors.

Unequal opportunities

Another limitation is that not every student has the same opportunities to perform the activities. In hospitals that are very busy or have many residents, the student may go weeks without taking a full history alone. Without observation, there is no data, and without data there is no decision of trust. That is why the EPA proposal needs to come with changes in how placements are organized.

EPAs from the start of the course

There are encouraging signs that the logic works even early on. At the University of Minnesota, first- and second-year students asked residents and teachers for EPA assessments while seeing patients together. There were thousands of assessments, and the growth curves followed the expected learning pattern, with a slow initial phase, rapid growth and then a slowdown. For some EPAs, four assessors on four occasions were enough for a reliable measure.

How the student can prepare

Even if your school does not use EPAs officially, their logic is useful to you. It shows what preceptors are observing and how to turn the placement into a sequence of small gains in autonomy.

Ask to be observed

Many students avoid being observed for fear of making mistakes in front of the preceptor. The logic of the EPAs shows that observation is the raw material of trust. Ask, directly and by arrangement, for the preceptor to watch part of your consultation and say what they would do differently. A sentence like "can I take the next patient's history with you in the room and then you tell me what was missing?" opens doors.

Present with reasoning, not just with data

The oral presentation of the case is one of the clearest windows into your reasoning. Listing findings is not enough. End with a main hypothesis, two alternatives and what you would do to tell them apart. The text on clinical reasoning in the clerkship offers a script for this, and the guide on how to organize reasoning in the medical record helps with the documentation EPA.

Say what you do not know

As we have seen, recognizing limits is one of the factors that most increase trust. When you are in doubt, say so clearly and propose a next step: "I'm not sure whether this test changes management; can I check with you before ordering it?". This kind of question shows judgment, and judgment is exactly what the supervisor wants to see before reducing supervision.

Close the loops

Go back to see the patient, check the result of the test you ordered, let people know when a pending item has been resolved. Reliability shows up in these small gestures. A student who closes the loops is a student who can be given more.

Train your reasoning outside the shift

Several EPAs depend on diagnostic reasoning: prioritizing differentials, choosing tests, recognizing the seriously ill patient. This reasoning can be trained outside the hospital, with written cases and discussions. The guide to differential diagnosis with clues and the text on clinical decision-making show ways to practice.

How NEXO helps train the reasoning behind EPAs

NEXO is a diagnostic reasoning workout, and so it speaks directly to the EPAs that depend on prioritizing hypotheses. In the case of the day, information arrives in up to six clues, and with each clue you decide whether to keep or change your guess. This act of revising the hypothesis in the face of a new piece of data is the same one the preceptor wants to see when asking you to prioritize differentials after a consultation.

With each guess, the proximity in ICD-10 across five levels shows how far your hypothesis was from the diagnosis, which helps you notice whether you were in the right chapter or far from it. The cases are fictional with a real clinical presentation, and the guide what NEXO is explains the game in detail. How the modes are unlocked is on the NEXO Economy page.

For those who coordinate internships, student leagues or residency programs and want to use cases in groups, there is NEXO Closer. Researchers in medical education can propose studies through NEXO Researcher, physicians who review cases take part through NEXO Curator, and those who create health content on social media can take part through NEXO Creators.

Frequently asked questions

Is an EPA the same thing as a competency?

No. A competency is a quality of the person, such as communication or knowledge. An EPA is a task in clinical work that requires several competencies at once. Assessing an EPA means assessing integrated competencies, the way they appear in practice.

Who decides that I am ready?

In day-to-day work, the preceptor decides on the spot. In programs that use EPAs formally, the recorded decision is made by a group, based on several observations by different assessors over time.

Does asking for help hurt my assessment?

In general, it helps. Recognizing your own limits and asking for support at the right time is one of the factors supervisors value most when deciding to trust.

My school does not use EPAs. Is it worth knowing about them?

It is. The logic of EPAs shows what preceptors observe and helps you ask for opportunities in a concrete way, turning the internship into a clear sequence of gains in autonomy.

Do EPAs guarantee that a new graduate is prepared?

There is no guarantee. The American pilot showed advances, but also activities in which few students were considered ready. EPAs help make visible what is missing, which is already an important step.

How many observations are needed?

It depends on the activity and the program. A study with early-year students found reliable measurement with four assessors on four occasions for some EPAs, but that number does not apply to all of them.

Checklist for gaining confidence during the internship

• List the activities you want to be able to do with indirect supervision by the end of the internship.
• Agree with the preceptor on when they will observe you in each of them.
• End every presentation with a main hypothesis, alternatives and a next step.
• Say clearly when you do not know and propose how to find out.
• Close the loops: go back to the patient, check results, report resolved pending items.
• Ask for short feedback after each observation and note what to change.
• Train diagnostic reasoning with cases outside the shift.

Want to train prioritizing hypotheses with each clue? Play the case of the day on NEXO.

References

1. ten Cate O. Entrustability of professional activities and competency-based training. Med Educ. 2005. PubMed 16313574
2. Ten Cate O, Chen HC, Hoff RG, Peters H, Bok H, van der Schaaf M. Curriculum development for the workplace using Entrustable Professional Activities (EPAs): AMEE Guide No. 99. Med Teach. 2015. PubMed 26172347
3. Brown DR, Warren JB, Hyderi A, Drusin RE, Moeller J, Rosenfeld M, et al. Finding a Path to Entrustment in Undergraduate Medical Education: A Progress Report From the AAMC Core Entrustable Professional Activities for Entering Residency Entrustment Concept Group. Acad Med. 2017. PubMed 28557941
4. Ten Cate O, Carraccio C, Damodaran A, Gofton W, Hamstra SJ, Hart DE, et al. Entrustment Decision Making: Extending Miller's Pyramid. Acad Med. 2021. PubMed 33060399
5. Amiel JM, Andriole DA, Biskobing DM, Brown DR, Cutrer WB, Emery MT, et al. Revisiting the Core Entrustable Professional Activities for Entering Residency. Acad Med. 2021. PubMed 34183597
6. Brown DR, Moeller JJ, Grbic D, Andriole DA, Cutrer WB, Obeso VT, et al. Comparing Entrustment Decision-Making Outcomes of the Core Entrustable Professional Activities Pilot, 2019-2020. JAMA Netw Open. 2022. PubMed 36156144
7. Encandela JA, Shaull L, Jayas A, Amiel JM, Brown DR, Obeso VT, et al. Entrustable professional activities as a training and assessment framework in undergraduate medical education: A case study of a multi-institutional pilot. Med Educ Online. 2023. PubMed 36794397
8. Brown DR, Moeller JJ, Grbic D, Biskobing DM, Crowe R, Cutrer WB, et al. Entrustment Decision Making in the Core Entrustable Professional Activities: Results of a Multi-Institutional Study. Acad Med. 2022. PubMed 34261864
9. Obeso V, Grbic D, Emery M, Parekh K, Phillipi C, Swails J, et al. Core Entrustable Professional Activities (EPAs) and the Transition from Medical School to Residency: the Postgraduate Year One Resident Perspective. Med Sci Educ. 2021. PubMed 34956699
10. Lerch SP, Huwendiek S, Nendaz M, Klöppel S, Pinilla S. Perceptions of Ad Hoc Entrustment, the Need for Supervision, and Coping Strategies in Clinical Residents: A Qualitative Study. Acad Med. 2025. PubMed 40048732
11. Hatala R, Ginsburg S, Gauthier S, Melvin L, Taylor D, Gingerich A. Supervising the senior medical resident: Entrusting the role, supporting the tasks. Med Educ. 2022. PubMed 35869566
12. Fernandes N, Neves C, Donato H, Pêgo JM, Ribeiro JC. Entrustable professional activities in medical education: strategies and results - a systematic review. Med Educ Online. 2026. PubMed 42857963
13. Nixon LJ, Poeppelman RS, Traxler M, Violato C. Assessing core entrustable professional activities in the pre-clerkship phase of undergraduate medical education. Acad Med. 2026. PubMed 41812055