A question bank that grounds its explanations in the actual guidance is enormously better than one that asserts an answer and expects you to believe it, because you can check, because you learn where the rule comes from, and because you build the habit of verification that clinical practice demands. But it introduces a specific failure mode, and it catches conscientious candidates hardest. You get a question wrong, you open the cited source, you begin reading the guideline properly, and ninety minutes later you have read a great deal, learned very little that will survive to next week, and answered no further questions.
Key takeaways
- The purpose of opening a source is to extract one rule, not to read the document.
- Almost every question turns on a threshold, an exception, or a sequence: identify which, take it, and leave.
- Do not open the source at all when the explanation has already resolved your uncertainty.
- Write the extracted rule as a single line in your own words, because rewriting is where the learning happens.
- Return the rule through spaced practice, because a rule read once is a rule you will not have in six weeks.
The rabbit hole
The pattern is familiar to anyone who has revised conscientiously. You miss a question about anticoagulation in atrial fibrillation. The explanation cites the guidance. You open it, and it is genuinely interesting: there is a section on risk scores, a section on bleeding risk, a table you had not seen, a discussion of a scenario you have encountered on a ward. An hour passes. You close it feeling that you have done something serious.
You have not. You have done something enjoyable and largely unretained. Reading is receptive, and receptive activity produces the feeling of understanding, which is not the same as the ability to produce an answer under time pressure in six weeks' time. The guideline is not designed as a revision resource. It is designed as a reference for practice, which is a different job, and using it as the former is why so much diligent revision produces so little movement.
What the question was actually testing
Here is the reframe that fixes it.
A question does not test a guideline. It tests one specific thing that the guideline contains, and almost always that thing is one of three types.
A threshold. A number above or below which the action changes. An eGFR at which a dose is adjusted. A score at which you anticoagulate. A duration after which a symptom becomes a red flag.
An exception. The circumstance in which the usual rule does not apply. The drug you cannot use in pregnancy. The patient in whom the standard first-line agent is contraindicated. The presentation where the normal pathway is bypassed.
A sequence. The order in which things must happen. What you do before you investigate. What you confirm before you treat. What you escalate before you manage.
Identify which of the three the question turned on, extract exactly that, and close the document. It takes ninety seconds and it is the whole lesson.
When to open the source at all
Most of the time, do not.
If the explanation resolves your uncertainty, and you now understand why your answer was wrong and what rule applies, you are done. Opening the source adds nothing except the illusion of rigour.
Open it in three circumstances. When the explanation states a rule and you do not believe it, because a moment's verification is worth having and because a bank can be wrong. When the explanation leaves you genuinely unclear about the boundary, and you need to see the actual threshold or the actual exception rather than a summary of it. And when the topic is one you know you are systematically weak in, and you have decided to invest in it properly, which is a deliberate choice rather than a reflex triggered by a single wrong answer.
Otherwise, take the rule from the explanation and move to the next question. Volume of questions, honestly reviewed, beats depth of reading for exam preparation, and it is not close.
Write the rule in one line, in your own words
When you do extract something, write it down, and write it as a single sentence in your own phrasing rather than copying the guideline's.
"Anticoagulate at a score of two or more in men, three or more in women, and reassess bleeding risk rather than withholding on the basis of it."
The rewriting matters more than it looks. Copying is transcription and produces nothing. Rewriting forces you to have understood the rule well enough to state it, and the act of generating that sentence is a retrieval event, which is the thing that actually builds memory. If you cannot write the line without looking, you have not extracted the rule, you have merely read it.
Keep those lines somewhere singular: one running document of extracted rules, not scattered marginalia across five resources. That document, by the end of your preparation, is the most valuable thing you own, and it is small.
Then space it
A rule extracted once and never revisited will be gone. This is not a moral failing, it is how memory works, and the counter is well established: retrieval, spaced over increasing intervals.
So do not let the rule sit in your document as a monument. Feed it back into practice. Meet the same principle again in a week, in a different question, and see whether you produce it. That return is what converts a line you once wrote into a rule you actually hold.
This is precisely the job spaced repetition exists to do, and it is why extracting rules and spacing them beats reading guidelines and hoping.
A note on which source
One point specific to exam preparation rather than practice. The rule you want is the rule your exam is testing, which is the rule from the guidance that governs the jurisdiction the exam is set in.
A candidate revising for a UK exam who resolves an uncertainty by reading American guidance has not extracted the rule they needed, and may have extracted a rule that is confidently wrong for their paper. Check that the source you are opening is the source your exam is grounded in, because "clinically correct somewhere" and "correct for this exam" are not the same thing, and the difference is a whole category of avoidable error.
Where iatroX fits
iatroX's explanations are grounded in NICE, CKS, SIGN and the SmPC and carry the source with them, so the rule and its provenance arrive together and you can verify without leaving the question. The Socratic Tutor asks you to reason before it explains, which surfaces the specific threshold, exception or sequence you actually missed rather than handing you a wall of text to read, and spaced repetition returns that rule at intervals so the extraction survives to exam day rather than expiring quietly in week five. Try it with free sample questions at iatroX. For the review habit that turns a wrong answer into a durable rule, see the distractor audit.
Frequently asked questions
Should I read the full guideline when I get a question wrong? Almost never. Extract the one rule the question turned on, usually a threshold, an exception, or a sequence, and close the document. Reading whole guidelines is receptive, enjoyable and largely unretained.
When is it worth opening the cited source? When you do not believe the explanation, when the boundary genuinely remains unclear, or when you have deliberately decided to invest in a topic you know you are systematically weak in. Not as a reflex after every error.
Why should I rewrite the rule in my own words? Because rewriting is a retrieval event and copying is not. If you cannot state the rule without looking, you have read it rather than learned it, and it will not be there in six weeks.
Does it matter which country's guidance I read? Enormously. Your exam tests the guidance of the jurisdiction it is set in. Resolving an uncertainty from another country's guidance can leave you confidently holding a rule that is wrong for your paper.