Free preview · no signup
MRCP Part 1 candidates lose marks by pattern-matching the noise. Find the Pivot breaks each stem into its findings and asks a single question: which one changes what you do next. It drills the step before the answer.
Runs in your browser on a phone · nothing to install · your progress stays on your device
Tap the finding that most changes immediate management.
A real case, shown solved — every batch works exactly like this.
Clinical reasoning is not binary, and marking it that way teaches a distorted model of it. A candidate who taps a raised JVP in a right-ventricular infarct is thinking correctly — so the case says so, rather than marking them wrong and moving on.
The finding that commits you to a different action. One per case, and the case only opens up once you find it.
The right line of reasoning, not decisive on its own. You are told why it is close and what it is still missing.
True, often alarming, and it changes nothing you do next. Learning to leave these alone is most of the skill.
Cases ship in topic batches so a sitting is one coherent piece of the blueprint rather than a shuffle. Three are open. Batch 01 is still the place to start, and Batch 03 is the one just published.
Batch 01 · Live
Nine cases on the decisions that actually get tested: who needs imaging, which test, and when the answer is to change the drug rather than add one.
Batch 02 · Live
Eleven cases built on NICE NG28, on the decisions the guideline actually turns on: which regimen a comorbidity commits you to, and when the right answer is to change nothing.
Batch 03 · Live New
Fourteen cases built on NICE NG115 and NG114, on the COPD decisions that turn on a single criterion: when to step inhaled treatment up or back, who qualifies for oxygen, and when to refer.
In preparation
Every case in the bank, shuffled across specialties the way the paper actually comes at you — cardiology, endocrine, tox, neurology, haematology and the rest.
Not open yet
It is written as prose, not a list. A stacked list reads as multiple choice, which is the habit this is meant to break.
Every finding is tappable and the prompt never names the answer. Marks stay on screen, so the trail of what you tried is still there when you find it.
Finding the pivot opens a short resolution: why that finding and not the others, and what it changes about the plan.
Sometimes nothing in the stem changes the plan already on the table, because the management is right as it stands. None, under the case, is how you say so — a real answer, not a trap.
Three batches are finished and out for criticism before the next one is written. If a case is wrong, ambiguous, or just annoying, say so in the thread — that is worth more than a compliment.