Doctors learn in different modes — deliberate exam preparation, just-in-time clinical search, reflective CPD, case-based learning, guideline updates, and repeated retrieval practice. Different tools serve different modes. The question is not which format is best — each serves a different purpose. The question is whether these modes can be connected, so that a real clinical question becomes not just an answered query but a retained learning point that the clinician can recall when the question recurs.
What Microlearning Does Well
Microlearning — short, focused learning episodes of 2-10 minutes — suits the fragmented reality of clinical schedules. It is low friction (no sustained time commitment), habit-building (fits into daily routines), mobile-accessible (usable during commutes and breaks), effective for updates (new guidelines, new drugs, new pathways), and easy to log as CPD (short, dateable, reflectable).
Platforms like BMJ Learning, RCGP Learning, FourteenFish, and Praktiki occupy this space — providing short modules, clinical updates, and CPD-logging tools for clinicians who cannot commit to hour-long study sessions.
The limitation is depth and retention. A 5-minute module provides a useful update but does not test whether the clinician can apply it under exam conditions, retrieve it three months later, or integrate it with related clinical knowledge. Comprehension — understanding something when it is presented — is not the same as retention — being able to recall it unprompted when needed. The recognition-recall gap applies: recognising information ("yes, I remember reading that") is easier than recalling it actively ("what was the recommended threshold?").
What Q-Banks Do Well
Q-banks apply the two most evidence-based learning principles: active recall (testing yourself rather than passively reviewing) and spaced repetition (scheduling review at expanding intervals). Karpicke and Roediger's 2008 study demonstrated approximately 50% better retention from retrieval practice compared to passive study — one of the most robust findings in learning science.
Q-banks are strongest for: exam preparation (directly testing examinable material), pattern recognition (repeated clinical scenarios build diagnostic skills), measurable progress (performance analytics), curriculum mapping (ensuring coverage), and long-term retention (spaced repetition algorithms).
iatroX Q-banks cover 15+ exams with AI-adaptive question selection, spaced repetition, and mock exam modes.
The limitation: Q-banks test pre-written questions — they do not answer the clinician's own real-time questions. A Q-bank tests "what is the first-line treatment for gout?" It cannot answer "this patient has gout AND CKD stage 4 AND is on warfarin — what should I do?"
What Clinical Search Does Well
Clinical search answers real questions in real time — at the point of care, relevant to the specific situation, source-grounded with citations, highly contextual, and immediately actionable.
Ask iatroX provides cited clinical answers oriented around UK practice. The strength is immediacy and specificity — the clinician asks their exact question and gets a cited response.
The limitation: clinical search solves the immediate question without automatically ensuring retention. A clinician who searches the same guideline every month has a search habit, not a learning outcome. The knowledge disappears after each consultation unless deliberately captured.
The Missing Connection
The most effective learning happens when modes connect. A real clinical question prompts a search. The search produces a cited answer. The answer addresses a genuine learning need. The need is captured as reflection. The reflection informs future retrieval. Future retrieval is reinforced by a Q-bank question on the same topic.
Most medical education tools miss this flywheel because they serve one mode in isolation. The microlearning platform provides updates but does not test whether the clinician can recall the content next month. The Q-bank tests knowledge but cannot answer the specific clinical question that arose during Tuesday's clinic. The clinical search tool answers the question but does not capture the learning for future retention. The CPD tool logs activity but does not connect the reflection to the original clinical question or reinforce the learning through future testing.
A worked example of the connected flywheel. A GP sees a patient with suspected inflammatory back pain. They are uncertain about the referral criteria. They search: "When should I refer suspected axial spondyloarthritis?" — and receive a cited answer from NICE NG65. They learn the key distinguishing features (morning stiffness >30 minutes, improvement with exercise, onset before age 45). They save this as a CPD reflection: "Updated my understanding of inflammatory vs mechanical back pain features and NICE referral threshold." Two weeks later, they encounter an MRCP-style question on the same topic in their Q-bank — and answer correctly because the learning was real, recent, and clinically grounded. The clinical question became search, became learning, became CPD, became exam performance.
This is the flywheel that transforms clinical practice into structured professional development — and it only works when the tools are connected.
Which Format Should Clinicians Use?
| Need | Best format |
|---|---|
| Passing a specific exam | Q-bank with adaptive learning |
| Checking a patient-specific question | Clinical search with citations |
| Keeping up to date | Microlearning modules |
| Building appraisal evidence | CPD reflection and logging |
| Retaining knowledge long-term | Retrieval practice + spaced repetition |
| Changing clinical practice | Reflection + real-world application |
How iatroX Combines the Formats
iatroX connects clinical search, exam practice, and CPD so the questions doctors ask in real life become learning they retain. Ask iatroX for search. Q-banks for retrieval practice. Calculators for applied tools. CPD for reflection and logging. Each component feeds the others — a clinical question revealing a gap becomes a study target; a Q-bank error prompts a search; a search answer saves as CPD.