Competency-based medical education makes a straightforward claim: a graduate should be certified on what they can do, not on how long they attended.
The claim is widely accepted and unevenly implemented, and the reason is practical rather than philosophical. Certifying ability requires evidence of ability — repeated, across contexts, observed. Most institutions have assessment instruments that produce evidence of knowledge, and the gap between the two is where the model tends to stall.
Time and competence are different measures
The traditional model uses time as a proxy: complete the rotations, sit the examinations, graduate. It has real advantages — it is administratively simple, it is fair in an obvious way, and everyone knows what is required.
Its weakness is that it assumes exposure produces competence at a uniform rate. Two students completing the same rotation have not had the same experience, because what they encountered depended on which patients were admitted while they were there. Clinical exposure is largely a matter of chance, and time-based certification treats that variance as though it did not exist.
Competency-based education replaces the proxy with the thing itself. Which is correct, and considerably harder to administer.
What has to be true for it to work
Three requirements, and institutions usually have the first and struggle with the others.
Competencies defined specifically enough to assess. Not "communicates effectively" but something observable, in a described situation, at a stated standard. Vague competencies produce assessment that is really global impression with a rubric attached.
Repeated observation across contexts. A single observation says little. Competence is consistency — the same standard in different situations, under different pressures, on different days. This is where the model becomes expensive, because observation requires an observer.
Evidence that accumulates into a record. Assessments spread across supervisors, rotations and forms, held in different places, cannot answer whether a student is competent. They can only answer whether a form was completed.
That third requirement is usually the binding constraint, and it is an infrastructure problem rather than an educational one.
The observation bottleneck
The honest difficulty with competency-based education is arithmetic.
Meaningful assessment requires a competent observer watching a student manage a real situation, then giving structured feedback. Consultants have clinics, wards, theatre lists, administration and research. Providing that for every student, repeatedly, across every competency, is not achievable at scale — not because of unwillingness but because the hours do not exist.
Institutions respond in predictable ways: assessment becomes a form completed retrospectively, or the number of assessed competencies shrinks to what is manageable, or observation becomes cursory. Each preserves the appearance of the model while removing what made it worthwhile.
This is where simulation has a specific and limited role. It does not replace observation by a clinician. It changes the ratio — a student can accumulate many assessed encounters without consuming supervisor hours for each, so that scarce supervision is spent where it adds most: on the difficult cases and on the students who need it.
What accumulating evidence looks like
For competency-based education to function, the record has to answer institutional questions, not just produce individual scores:
- Which competencies has this student demonstrated, how many times, in which contexts?
- Where is the evidence thin?
- Which students are consistently weak in the same dimension?
- Is a particular competency weak across the whole cohort — which is a curriculum finding, not a student one?
That last question is the one institutions most want answered and least often can. If a cohort is uniformly weak at investigation strategy, that is a teaching problem, and it is invisible when assessment lives on paper in separate rotations.
SYNTAX assesses across nine dimensions — history taking, communication, physical examination, investigation strategy, differential diagnosis, clinical reasoning, management planning, patient safety and professionalism — precisely because a single mark cannot support any of those questions. The reasoning is in how clinical reasoning is actually assessed.
Where it goes wrong in implementation
Competencies multiply. A framework with two hundred items is unassessable, so it becomes a tick-box exercise. Fewer, well-defined competencies assessed properly beat comprehensive ones assessed nominally.
Assessment becomes documentation. Forms completed after the fact to satisfy a requirement. This is the most common failure and the hardest to detect, because the paperwork looks correct.
The remediation path is missing. Identifying that a student is not yet competent is only useful if something follows. Without a route to deliberate practice and reassessment, identification is just a slower way of passing them.
Time reappears through the back door. Students must still finish on schedule, so competence gets assumed at the deadline regardless of evidence.
What we would say honestly
Truffaire builds SYNTAX, so our interest here is disclosed. Two things we do not claim.
Simulation does not replace clinical rotations. It prepares students to use real exposure better and does not substitute for patients. Any claim otherwise would be a claim about medicine we are not entitled to make.
Software does not decide competence. In SYNTAX, scores are computed by logic against defined criteria and a model writes the narrative — the separation exists so assessment is auditable. The judgement about whether a student is ready remains with faculty, and should.
What the technology genuinely changes is volume and record. A student can practise the same decision repeatedly and safely, and the evidence accumulates in one place rather than across forms — which is what makes the institutional questions above answerable.
Frequently asked questions
Does competency-based education mean students finish at different times?
In principle. In practice most institutions run a hybrid — fixed duration with competency requirements inside it. That is a reasonable compromise, provided competence is genuinely evidenced rather than assumed at the deadline.
How many observations are enough?
Enough to establish consistency, which depends on the competency. A single observation of anything establishes very little; the standard should be defined per competency rather than as a global number.
Is this more work for faculty?
Initially yes, and the goal is to change where the effort goes — less time on routine assessment, more on the students and cases that need judgement. Implementations that only add work do not survive.
Can simulation-based assessment count formally?
That is a regulatory and institutional decision rather than ours to assert. What it can do regardless is provide evidence of practice and progression that supports formal assessment.
How do we identify struggling students earlier?
Patterns across encounters — repeated omissions, consistently low scores in one dimension, safety issues recurring. These appear long before examination results, provided the record aggregates across encounters rather than sitting in separate rotations.
Where to start
Take one competency in your curriculum. Ask three questions: is it defined specifically enough that two assessors would agree, how many times is it observed per student, and where does that evidence live.
If the answer to the third is "on forms in different departments", that is the constraint — and it is an infrastructure problem, not a pedagogical one.
If you are evaluating this for a college or teaching hospital, get in touch, or see where SYNTAX sits in Truffaire's systems.