This article is adapted from a post in Dr Koshy Jacob's "Building Revise Radiology in Public" series on LinkedIn.

A little while ago we sat down and seriously considered building our own library of recorded lectures, a full series covering the radiology curriculum from first principles, taught by our own faculty and badged as ours. It is the obvious thing to build, and it would have looked impressive on a slide. We decided against it, and the reasoning behind that matters rather more than the decision itself, because it goes to the heart of what a learning radiologist actually lacks.

The temptation

The appeal of a lecture series is easy to understand. It is visible, it is countable, and it feels like substance. One hundred lectures sounds like a serious educational offering in a way that harder-to-photograph work never quite does. It also flatters everyone involved. Consultants enjoy teaching, cameras are gratifying, and there is a quiet satisfaction in having said your piece on a subject you have spent twenty years thinking about.

The trouble is that none of those are reasons a trainee should pay for something. They are reasons we would have enjoyed making it.

Excellent teaching already exists

The first honest observation is that the world is not short of radiology lectures. There is a great deal of excellent recorded teaching available, some of it free, much of it produced by societies, departments and individual radiologists who have given their time generously over many years. A trainee with a laptop and a search box has access to more high-quality explanation than any generation of radiologists before them.

If we had built our own series, the realistic outcome is that we would have produced something slightly worse than material that already exists, and then charged for it. We would have been duplicating good work with our logo on the front. We would rather point our members towards the material that is already out there than pretend we are the only source of it.

And watching is not learning

The second observation is less comfortable, because it questions something the whole medical education market is built on. Passive video consumption is a weak way to learn, and it is worth setting out what the evidence does and does not support here.

What is well established is the value of the opposite approach. Retrieval practice, the effort of pulling an answer out of your own head rather than reviewing it on a page, has been tested repeatedly across education and consistently outperforms passive review, with meta-analyses putting the effect at around a half a standard deviation. The benefit grows when corrective feedback is built in, and self-testing gives the learner something else besides retention, which is calibration, a sense of which topics genuinely need more attention. Within radiology specifically the literature is thinner than you might hope. A recent systematic review in the Journal of the American College of Radiology screened over thirteen hundred records and found only eight eligible trials, five of which showed a benefit.

The evidence on video is more contested, and I want to be careful with it. One study of two medical student cohorts found that the number of videos watched was a negative predictor of licensing examination performance while the number of practice questions worked was a positive one. An earlier study of lecture recordings found the same direction of association, but its authors concluded that heavy video use was largely a marker of students who were already struggling with the material rather than a cause of poor performance. Several other studies have found no association at all. The honest reading is that video use may often be a symptom rather than a cause. That still tells us something worth knowing. If watching is what a struggling learner reaches for instead of the harder thing, building more of it does not help them.

The mechanism is not mysterious either way. Watching an expert reason through a case feels like understanding. It produces the pleasant sensation of comprehension without the effortful retrieval that lays down durable memory.

We should be careful here. Video is not useless, and a well-made explanation of a difficult concept can save a trainee hours of confusion. What matters is what surrounds it. A lecture interrupted by questions the learner has to answer, or followed by cases where the idea has to be applied, does real teaching work. Left on its own, the same recording gives you the pleasant feeling of an evening well spent and very little that survives the week.

What is actually missing

If material is abundant and passive watching is weak, the question becomes what a learning radiologist genuinely lacks. From everything we see across our members, the answer is consistent, and it has almost nothing to do with volume.

The first thing missing is sequence. Most trainees preparing for an examination are not short of resources. They are sitting in front of sixteen open browser tabs with no clear idea what they should do on Tuesday evening. They do not know whether they have covered enough gastrointestinal imaging, or whether the gaps they can feel are real gaps or ordinary anxiety. They have no reliable way to tell whether they are on track.

The second is feedback. A trainee can look at three thousand cases and get slowly better at looking at cases, but without knowing whether their report was right, whether their reasoning was sound, and where precisely it went wrong, the improvement is slow and accidental. This is why the day-to-day work of learning radiology is perceptual and practical rather than purely informational. Knowledge is necessary, but the skill is in seeing, deciding and saying.

The third is simply someone to ask. Every trainee has questions that are too small to email a consultant about and too persistent to ignore. In a well-staffed department with a generous reading-room culture, those questions get answered in passing. Many trainees, in many places, do not have that.

So this is what we build instead

Our work is organised around those three absences rather than around content production. In practice that means putting cases first, with knowledge arriving as feedback on the case rather than as a lecture delivered in advance of it. It means mapping a very large case library against the curriculum so that coverage can actually be trusted, which is unglamorous, painstaking work and the single most valuable thing we do. Structure and sequencing follow from that, so a candidate with an examination date knows what this week is for. And we run live viva practice with real examiners, because the part of the FRCR that most often defeats good candidates is not knowledge but the experience of presenting under pressure, and that cannot be delivered by any recording ever made.

None of that photographs well. It is much harder to summarise on a slide than a lecture count. We think it is the part that changes outcomes, and we would rather be judged on it.

The limits of this, stated plainly

It would be dishonest to present a platform as a complete radiology education, and we have no intention of doing so. The evidence points consistently towards blended learning outperforming both traditional teaching alone and online teaching alone, which is precisely why we work with clinical partners rather than pretending we can replace them.

There are things only a department can give. Supervised reporting with a consultant looking over your shoulder. Consent and procedural work on real patients. The unspoken pointing, the raised eyebrow, the rhythm of question and answer that makes reading-room culture such a powerful teacher and travels so badly through a screen. Contribution to a real multidisciplinary meeting, where the stakes are actual. Junior trainees in particular need volume of real clinical work, and a platform that positions itself as a substitute for that is doing them harm. Our job is the structure, the sequencing, the feedback and the examination preparation. The clinical formation belongs to the department, and it always will.

Three questions worth asking anyone

If you are a trainee choosing where to spend limited money and much more limited time, we would suggest three questions, and they apply to us as much as to anyone else. First, does this tell me what to do next, or does it simply give me more things I could do. Second, does it tell me whether I was right, and why. Third, can I ask a human being a question and get an answer from someone who has actually sat the examination.

A resource that answers all three is worth paying for. A resource that answers none of them is a library, and libraries are already free.

Where this leaves us

The abundance problem in medical education has been solved. There is more material than anyone can consume and it keeps arriving. The scarcity has moved, and it now sits in structure, in feedback, and in access to someone who knows. Those are harder to build and much harder to market, which is probably why so much of the sector is still competing on volume.

We decided not to build our own lecture series because the trainees we serve do not need another one. They need to know what to do on Tuesday, whether they got it right, and who to ask when they did not. Getting that right is a long job, we are some way from finished, and there are weeks when the curriculum mapping feels endless. It is still the job our members keep asking us to do.

A note on the evidence

On retrieval practice: Thompson and Hughes, The Effectiveness of Spaced Learning, Interleaving, and Retrieval Practice in Radiology Education: A Systematic Review, Journal of the American College of Radiology, 2023. The meta-analytic effect sizes cited come from Rowland, Psychological Bulletin, 2014, and Yang and colleagues, Psychological Bulletin, 2021.

On video and examination performance: Clemmons, Vuk and Jarrett, Educational Videos Versus Question Banks: Maximizing Medical Student Performance on the United States Medical Licensing Examination Step 1 Exam, Cureus, 2023, a correlational study of two cohorts at a single institution. The earlier study finding the same direction while attributing it to student difficulty rather than to the videos themselves is McNulty and colleagues, An analysis of lecture video utilization in undergraduate medical education, BMC Medical Education, 2009. Other studies have found no significant association between lecture attendance or recording use and licensing examination performance, including a multi-institution survey of medical students published in 2021 and a survey of 560 students across 102 schools published in 2024

Originally shared by Dr Koshy Jacob on LinkedIn. Follow the series there