Sign in to continue
Step 2 and Shelf preparation use exam analytics and blocks. Ortho audition preparation uses the longitudinal course, fracture conference workflow, imaging, anatomy, presentation practice, and linked Anki.
Open the Ortho Course Home for the longitudinal plan, dated calendar, daily workflow, fracture conference packet, image occlusion lab, skeleton classification atlas, presentation gym, Friday journal lab, and Marty McFlyin Anki linker.
Take authentic NBME blocks, let the engine track your weak areas, then close the gaps with Focus Session. Everything else follows from that loop.
Repeat this cycle. Every other feature supports it.
Open the Library, choose a form, and take it timed and unspoiled — same as test day.
Open Past Blocks, filter to wrong answers, then use Expert Review to understand each one — not just the right answer, but why you were wrong.
Run a Focus Session from Home or Library. It reads your Weakness Map and serves questions in your worst concept areas first.
Use Score Forecast to see if your score is moving. Use Weak Areas to find the patterns that keep costing you points.
The most important tool after you have taken at least one block.
Rahvan scores every concept area you have answered. Focus Session reads that map and sends you questions from the weakest areas first — not randomly.
If you have fewer than 40 answered questions, run the 20-question baseline diagnostic first. It seeds the Weakness Map so the adaptive engine has something real to work with.
By default Focus Session pulls from authentic NBME questions only. You can switch to Practice Bank (4,000+ questions) or mix both in the setup screen.
When Focus Session finds a gap with no available question, the AI generates a targeted item on the spot so you never run out of material in a weak area.
What each page does and when to use it.
Choose from NBME forms, Free 120, CMS blocks, or Practice Bank. Start a scored block, resume a section, or launch Focus Session from the nav.
Return to any completed block, filter to wrong answers, mark questions reviewed, and open source pages. Always start here after finishing a block.
Load any question and ask the AI anything — why not B, what the key clue was, how two diagnoses differ, or what the next step rule is. Deep reasoning, not just definitions.
Find the concept areas, question types, and traps that keep costing you points. Use it to decide which area to target next, then launch targeted practice.
See your predicted score range, score trend over time, and how close you are to your target. Updates after every block.
Accuracy by system, timing breakdown, review debt, and readiness signal — all in one place.
Save algorithms, comparisons, and question-linked explanations you want to see again. Spaced repetition keeps them surfacing at the right intervals.
Practice reading stems, comparing choices, and applying elimination strategy before committing to an answer.
Talk through clinical reasoning out loud — useful for building the internal voice you need on test day.
COMLEX-style boards covering mechanics, diagnosis, autonomics, and treatment. Separate from Step 2 prep.
Use this only when the form is not already in the Library.
Upload the question PDF. Readable PDFs process fastest; scanned PDFs may need OCR.
Upload the answer PDF or paste a simple key like 1. C, 2. A, 3. D.
Rahvan keeps original pages available so images, ECGs, tables, and wording can be checked.
Review problem questions before starting. If the count is wrong, fix it before trusting the block.
If you are new, start with the Library. Pick one form, take a block, review misses, then look at your stats.
Open About & Support and include a screenshot plus what you clicked. Recent browser errors can be attached automatically.
If OCR is messy, this is the prompt to use before pasting cleaned text back into Rahvan.
Practice the habit: read the ask first, glance at choices, categorize the task, hunt abnormal positives, rebuild the timeline, judge severity, cross out deliberately, then compare finalists evenly.
Rahvan's Strategy Guide is built from a question-level internal review of Step 2 NBME and CMS material. The point is not to promise a magic score; it is to expose the repeatable places where strong students leak points and turn each leak into a specific review action.
These percentages come from Rahvan's local Step 2/CMS tagging work, not from USMLE or NBME. Official score prediction should stay tied to score-valid exams; Rahvan-generated practice is for repair, not inflation.
| Task | Step 2 NBME | CMS | Meaning |
|---|---|---|---|
| Diagnosis | 847 / 1,600 (52.9%) | 1,748 / 2,800 (62.4%) | CMS is especially diagnosis-heavy. |
| Next best step | 296 (18.5%) | 434 (15.5%) | Both require sequence discipline. |
| Test selection | 179 (11.2%) | 177 (6.3%) | Step 2 has more readiness-style test choice. |
| Initial management | 70 (4.4%) | 145 (5.2%) | Stability and timing still matter. |
| Treatment/pharm | 78 (4.9%) | 131 (4.7%) | Therapy questions often hide eligibility traps. |
CMS should repair recognition and subject-level misses. Full Step 2 NBMEs should protect the readiness signal. The practice bank should target weak rules without being treated as a score predictor.
| Skill | Step 2 NBME | CMS | Repair move |
|---|---|---|---|
| Illness-script recognition | 885 (55.3%) | 1,753 (62.6%) | Name the syndrome before looking for the answer. |
| Management sequencing | 440 (27.5%) | 670 (23.9%) | Decide stable vs unstable before test vs treat. |
| Lab/imaging interpretation | 74 (4.6%) | 58 (2.1%) | Name the finding before the diagnosis. |
| Localization | 49 (3.1%) | 93 (3.3%) | Translate anatomy into the next step. |
| Screening guidelines | 41 (2.6%) | 67 (2.4%) | Check age, risk, interval, and contraindications. |
| Trap | Step 2 NBME | CMS |
|---|---|---|
| Ignoring key negative | 774 (48.4%) | 1,348 (48.1%) |
| Treating before eligibility | 423 (26.4%) | 626 (22.4%) |
| Testing when treatment needed | 163 (10.2%) | 200 (7.1%) |
| Premature closure | 88 (5.5%) | 269 (9.6%) |
| Ignoring timeline | 63 (3.9%) | 68 (2.4%) |
The useful unit is not "review cardiology." The useful unit is the pivot clue that changed the answer and the trap that almost pulled you away from it.
176-item cluster. Repeated pivots included bilateral papilledema with hyperreflexia after trauma, early awakening with memory difficulty, lateral foot sensory deficits, and gait plus urinary changes.
83-item cluster. Repeated pivots included hallucinations during the episode, headache plus confusion plus seizure, pain relieved by leaning forward, and cognitive decline plus gait change.
26-item cluster. Pivots included Rh-negative status with prior RhIG, gestational age, transverse lie, fetal distress or no distress, and gush of fluid.
29-item cluster. Pivots included low-output symptoms, orthopnea and edema, acute distress without fever or chest pain, and feeding difficulty with tachypnea.
Uses score-valid material for readiness. Practice-bank performance can explain weak rules, but should not inflate the forecast.
Turns misses into leak categories: missed clue, wrong task, timing error, eligibility error, or premature closure.
Allocates realistic hours to repair, readiness blocks, and spaced review instead of generic daily advice.
Builds focused reps from weak rules. It should train reasoning without pretending to be an official predictor.
Internal Rahvan evidence: the audit numbers above come from Rahvan's tagged local Step 2/CMS material. Official exam facts: the 2026 USMLE Step 2 CK delivery update describes a one-day exam using shorter blocks, with up to 316 items and sixteen 18-to-20-item blocks. Rahvan should use official facts for exam-format simulation, and internal findings for study strategy.
This page turns Rahvan's Step 2/CMS tagging into today's practice stack. Use it when you want the next high-yield decision skill to train, not another strategy essay.
Start with Illness scripts: it is the biggest reasoning bucket and it feeds every other decision.
These are internal findings from our tagged Step 2 CK + CMS library. They are useful for study direction, but they are not an official USMLE or NBME blueprint.
Rahvan evidence: our Step 2/CMS tag report found diagnosis was the largest task type, while management sequencing was the second major reasoning pattern. Official framing: USMLE describes Step 2 CK as applying clinical science to patient care and includes diagnosis, management, diagnostic studies, prevention, ethics, and patient safety in its task framework. The percentages on this page come from Rahvan's local dataset, not from USMLE.
Why this matters: Most wrong answers start before management: the syndrome is misread, or one disqualifying clue is ignored.
Why this matters: The same diagnosis can have a different answer depending on how sick the patient is right now.
Why this matters: Step 2 loves answers that are medically true but wrong for this exact moment.
Why this matters: OB/GYN questions often turn on timing, trimester, stability, and contraindications.
Why this matters: Medicine emergencies are where diagnosis, triage, and timing collide under pressure.
Why this matters: Age, time course, and localization do a lot of the work in pediatrics and neurology.
Why this matters: Ethics questions are usually testing priority, tone, consent, capacity, confidentiality, and safety.
Why this matters: Prevention questions are efficient points if you know age, risk, pregnancy, immunization, and contraindication triggers.
Do two mixed timed blocks, then review only misses and guessed-correct answers. Label each miss as diagnosis, stability, test/treat timing, contraindication, ignored negative, ethics tone, or screening fact. Then open Approach Coach for one or two representative misses and rehearse the method out loud.
These sources support the structure of the guide. Rahvan's local analysis explains what appeared most often in our Step 2/CMS library; the external sources explain the public exam/task framing and prevention references.
Estimate your current range, understand what is holding you back, and see what moves the score next.
Misses are signals. Build a new targeted set from the pattern first; replay saved questions only when you deliberately want exact recall.
Your attempt log: reopen full blocks, resume unfinished sessions, include or exclude scores, and audit what happened on a specific test day.
Cross-exam pattern analysis. Every missed question classified by error type so you know exactly what to fix — not just what you got wrong.
Browse every assessment, practice tool, and uploaded exam without pre-exam spoilers.
Search FA algorithms, First Aid Step 2 CK, NBME nuance, Mehlman HY, UWorld tables, Inner Circle notes, and Sketchy Micro/Pharm from one place while reviewing misses.
High-yield antineoplastic drug toxicities for rapid recall during review.
The old foundation workspace has been superseded by the dedicated OMM practice engine. It now opens private scored practice, review targets, explanations, and the remaining locked review queues in one place.
Paste OMM notes, tables, diagrams, transcripts, faculty corrections, or learning signals. Each becomes a traceable intake object.
Candidate facts become draft rules, diagnostic decision trees, traps, and contraindication checks. Nothing is approved automatically.
Future question writing follows blueprint objective, rule selection, clinical scenario, validation, and human review.
Clinical presentations, physician tasks, competencies, OMM concepts, rules, treatments, contraindications, traps, questions, and validators are represented as JSON nodes now, with a database-ready shape later.
Manage saved attempts, backups, and local browser data from one place. Export a backup before switching computers, clearing browser data, or deleting old uploads.
Create personal invite codes and see how many times each one has been used.
Rahvan keeps the data needed to reopen blocks, review attempts, and protect your progress.
Use this before clearing browser data, switching devices, or making major changes.
Use this checklist before clearing old browser data, switching devices, or deleting local files. This is the migration home for students who still need to bring old local progress into Rahvan.
These tools protect analytics while reducing storage use. Export a backup before deleting anything.
Rahvan was built for the long, often uncertain stretch of medical studying: the part where you need realistic practice, honest feedback, and a calm system that helps you keep moving.
Rahvan is inspired by rahwar, a word used to describe a swift, smooth, ambling gait in horses. That image matters to us: progress that is steady, controlled, and sustainable, even over a difficult distance.
The name also carries the spirit of the raven: intelligence, transformation, prophecy, and the ability to move between what is known and what is still hidden. Ravens are often symbols of seeing clearly in dark places, facing the unknown, and learning to trust what careful attention reveals.
Studying for boards can make capable people feel lost. Rahvan is meant to be a guide through that space: helping you practice under realistic conditions, review your decisions honestly, find the patterns behind your misses, and turn uncertainty into a next step you can actually take.
The goal is not just to help you answer more questions. It is to help you become the test taker you are working toward: calmer, sharper, more self-aware, and more willing to keep going when the path is not fully lit yet.
Rahvan is still growing, and thoughtful feedback helps shape what it becomes. If something feels confusing, broken, useful, or worth improving, reach out anytime.
You can email us directly at teamrahvan@gmail.com. For optional donations or development support, use the support page.
Rahvan is an independent study platform. It is not affiliated with NBME, USMLE, COMLEX, or any exam publisher, and it is not medical advice. It is a private workspace for practice, review, and studying with more direction.
Rahvan could not prepare this exam automatically.
Rahvan will keep the technical checks quiet unless something needs attention.
Rahvan already has the questions and scoring for this material. Choose a section if shown, then begin.
This session will use the loaded question set.
The timer is paused. Your progress is saved automatically.
Voice starts only when you press Start reasoning. AI coaching requires one-time consent.
Paste any USMLE or COMLEX question. Get writer intent, decisive clues, trap analysis, pathophysiology, and a durable review save.
Paste a question on the left and click Analyze →
Your attempt is summarized below. Save is automatic when the block ends, and you can export your full history from the dashboard.
Select a topic from the list to view the reference
For this block, I intend to:
What pulled you away?
Your progress is saved automatically.
Send a bug, content problem, or feature request into the Rahvan fix queue. Please avoid pasting full question stems unless the issue requires it.
A recall deck for algorithms, comparisons, and near-miss rules. It is separate from the mistake list: save what you need to remember, even when you answered correctly.
Not every rule worth keeping comes from getting a question wrong.
Rahvan could not launch similar practice from this card yet. The saved concept is still structured for later practice:
Try opening the original question first, then use Similar Qs again.
When you speak your reasoning, Rahvan can send it to an AI coach for richer, more natural feedback — not just keyword matching.
