This plan is for candidates preparing for USMLE Step 2 CK, whether US students after clinical rotations or international graduates on the licensing pathway. The established resources dominate, and nothing here argues against them: UWorld and the NBME materials, with AMBOSS for many candidates, are the recognised core, and your preparation should be built around them. iatroX is not a replacement for these resources; it is the adaptive remediation and retention layer on top, which converts the misses they surface into reasoning you retain — closing the gap where good questions do not stick.
What you're up against
You may be balancing preparation with clinical rotations, a job, or the wider demands of the licensing pathway, so time is finite. Step 2 CK is management-focused, with longer clinical vignettes that test the best next step rather than recall, and it samples a broad curriculum. The risk is not the resources but the loop: doing UWorld blocks, reading the excellent explanations, feeling prepared, and failing to retrieve the reasoning on fresh questions later. The plan has to keep you active and to retain a broad curriculum to test day.
What to revise from
Anchor on UWorld as the core question bank, the NBME self-assessments and clinical-science materials for calibration to the real standard, and AMBOSS for additional questions and its integrated library if you use it. These define your coverage and your question volume. Use iatroX as the adaptive remediation and retention layer alongside these: its engine re-sequences your missed concepts and spaces them so they survive the broad curriculum, and its Socratic Tutor unpacks the reasoning behind a board-style miss rather than restating the explanation.
The plan, week by week
Plan across the months before your sitting, building an active loop on top of the dominant banks. Work UWorld blocks systematically, but before reading any explanation, commit to why you chose your answer; then read it, name the precise misconception, and re-derive rather than re-read. Take recurring misses into adaptive remediation that re-presents the underlying concept at spaced intervals. Use NBME self-assessments at intervals to calibrate. As the exam nears, shift to timed, mixed practice that matches the format and stamina. The weekly minimum is a daily block of questions properly reviewed plus spaced re-testing of your weak concepts, with periodic self-assessments. The discipline is active review, because recognition of an explanation is not retrieval.
A representative week
To put it concretely, picture a week around rotations or work. On most days you do a UWorld block, predicting your reasoning before reading and debriefing each miss into the precise misconception rather than passively reading the explanation, with the remediation layer scheduling your weak concepts to return. You hold a focus across several days so it consolidates. Once every week or two you sit an NBME self-assessment to calibrate against the real standard and to surface weaknesses under timed conditions. On busy clinical stretches you protect a smaller daily block rather than skipping, since consistency over months is what builds retention across a broad curriculum. As the exam nears, you add full-length timed practice for stamina. Across the week the dominant banks generate the practice and the misses, and the remediation loop stops those misses recurring on the day.
The remediation gap excellent banks leave
It is worth being precise about the gap, because UWorld and the NBME materials are genuinely excellent and the temptation is to assume they are sufficient alone. Their explanations are detailed and well written, and that is where the risk hides: reading a strong explanation produces a confident sense of understanding that is not the same as retrieving the reasoning, cold, on a different vignette weeks later. Step 2 CK is a retention problem as much as a knowledge one, because it samples a broad curriculum long after you first studied each piece. The banks generate excellent practice and pinpoint your misses, but the loop that converts a miss into durable, retrievable reasoning — active interrogation rather than re-reading, and spaced re-testing rather than a single pass — is something you have to build. The evidence supports this: a randomised study by Bastani and colleagues found an assistant that gave answers left learners worse on a later unaided test, while one that gave hints did not. An adaptive layer that schedules your weak concepts to return, and a tutor that makes you reason rather than re-read, close that gap deliberately.
Where iatroX comes in
Think of iatroX, specifically, as the adaptive remediation and retention layer beside UWorld, AMBOSS and the NBME materials, not a competitor to them. Its engine targets the related weaknesses a miss reveals and spaces them for retention across a broad curriculum, and its Socratic Tutor unpacks the reasoning behind a miss — behaving as the coach rather than the crutch — on top of the resources that remain your foundation.
Knowing when to deviate
If you are early in preparation and still building coverage, prioritise working through UWorld; the remediation loop earns its value in the consolidation phase, when retention and reasoning are the constraint. If your misses are not recurring and your self-assessments are tracking well, you may not need more. If time is short, protect the NBME-calibrated essentials and active review over raw volume. The danger sign is a high question count with stagnant scores — usually a sign of passive review that active remediation fixes.
Questions worth answering
Does iatroX replace UWorld or the NBME materials? No — those remain the dominant, foundational resources; iatroX is the adaptive remediation and retention layer on top.
Why add a remediation layer? Because Step 2 CK tests a broad curriculum months later, and excellent explanations produce recognition rather than the durable retrieval a remediation loop builds.
When should I start using it? In the consolidation phase, once you have coverage and the problem becomes retention and reasoning.
How do I know passive review is the issue? A high question count with flat self-assessment scores is the classic sign.