Use your FRCEM SBA question bank to identify a gap, work out why it occurred, and test whether you can handle a different version of the problem. Choose the practice mode for that purpose. A completed set is useful when it changes your next revision decision, not simply when it adds to your total question count.
FRCEM SBA is the Fellowship of the Royal College of Emergency Medicine single best answer examination. This guide is written from StudyMedical’s perspective as a question-bank provider. The workflow and numerical examples are our editorial suggestions, not an official RCEM study programme or a validated prediction of passing.
Start with the job you want the session to do
Before opening a quiz, finish this sentence: “At the end of this session, I want to know whether I can…” You might choose to interpret a type of data, distinguish two similar presentations, retrieve a guideline recommendation or maintain accuracy within a time limit.
Keep that aim narrow enough to assess. “Get better at emergency medicine” gives you little basis for choosing a mode or deciding what to review. “Explain why I keep choosing a diagnosis when the question asks for the next investigation” gives you a specific behaviour to examine.
Practice testing, feedback and spacing learning over time have support in Dunlosky and colleagues’ review of learning techniques. This is broader educational evidence, not evidence that a particular FRCEM question quota or interval guarantees success. Our practical application is to answer actively, investigate the feedback and return to the concept after a gap.
Choose new, incorrect, topic or timed practice
These choices answer different questions. “New” describes prior exposure, “incorrect” describes previous performance, “topic” describes content and “timed” describes conditions. They are not necessarily mutually exclusive. Check which combinations your platform supports.
Practice choice | Useful purpose | What to watch for |
|---|---|---|
New questions | Sample what you can do without remembering the item | A new item may still test a familiar concept; new does not mean harder |
Previously incorrect questions | Check whether you have corrected an identified error | Recognising the answer is weaker evidence than explaining the reasoning |
Topic-focused questions | Build or repair understanding in a specific area | Knowing the topic in advance may provide a cue |
Mixed questions | Practise identifying the relevant concept without a topic label | A small mixed set may not cover the curriculum evenly |
Timed questions | Explore pacing and decisions under a time constraint | Review remains necessary after the clock stops |
StudyMedical’s FRCEM SBA question-bank options list custom topic quizzes, timed quizzes and modes for new or previously incorrect questions. Use those controls to answer a learning question. For example, revisit an error after reviewing the concept, then seek unfamiliar practice to check whether the improvement transfers.
Keep curriculum coverage visible
Maintain a separate coverage checklist so that repeated work on weak areas does not become permanent avoidance of everything else. RCEM maps the FRCEM SBA to the full Emergency Medicine curriculum, including research, management and quality improvement. Refer to RCEM’s examination information and linked blueprint when planning coverage.
A useful weekly check has two parts: what remains untouched, and what remains uncertain despite previous study. Those lists require different actions. Untouched material needs an initial learning plan; a persistent uncertainty needs closer examination of the reason for the error.
Use a repeatable answer–review–retest routine
During a learning session, commit to an answer before reading the explanation. Note your confidence briefly: secure, uncertain or guessed. This helps you find fragile correct answers as well as marked errors. You do not need a complex scoring system.
During review, explain the decisive detail in the stem and why your chosen alternative was less appropriate. “I did not know it” is sometimes accurate, but it can hide a reading error, a calculation mistake or a failure to apply knowledge you already had.
Worked example: one session, several different problems
Imagine a candidate answers 18 of 30 unfamiliar questions correctly. Of the 18 correct answers, four were guesses. The review list therefore contains 16 items: 12 wrong answers plus four correct guesses. These are fictional numbers for demonstrating a workflow, not a recommended target score.
What the review reveals | Useful next step | Evidence to look for later |
|---|---|---|
Five errors involve missing knowledge | Read a focused source; write a brief explanation in your own words | Retrieve the principle without the original options |
Three errors involve choosing the wrong task | Identify whether the question asks for diagnosis, investigation or management | Fewer task-selection errors in another set |
Two errors involve numerical working | Write the formula, units and denominator before substituting values | Complete a different example correctly |
Two errors occur after time runs short | Inspect where time was spent earlier in the block | Finish a comparable timed block without the same end-of-set rush |
Four correct guesses | Review the reasoning as if the answer had been wrong | Explain why alternatives fail in a fresh scenario |
The candidate should not automatically repeat all 30 questions immediately. That would mix several problems into one action. A better next session might repair the main knowledge gap, while a later short block tests pacing. The action should match the cause you found.
Keep the error log small enough to use
For each recurring problem, record the concept, the reason for the error, the corrective action and the planned check. One entry can cover several questions. Avoid copying entire explanations into a document that becomes too large to revisit.
Concept | Reason for error | Correction | Retest |
|---|---|---|---|
Interpreting a study result | Confused relative and absolute measures | Work through definitions and one fresh example | Explain the result using different numbers |
Selecting the next step | Answered the likely diagnosis instead of the requested action | Underline the task and state it before comparing options | Use an unfamiliar vignette |
Reading an image | Recognised a pattern but ignored the clinical context | Describe the finding, then relate it to the question | Review a different image with its accompanying history |
When the same error recurs, update the entry rather than adding another identical note. Ask whether the previous correction was too vague. “Read more” can become “compare the two concepts and explain the feature that changes the answer”.
Separate first-attempt scores from repeat scores
Keep unfamiliar attempts and repeated attempts visible separately. Also record whether the set was timed, topic-focused and completed without notes. A percentage without those conditions can be difficult to interpret.
Worked score example
A candidate completes 20 unfamiliar questions and gets 12 correct, then revisits 20 previous questions and gets 18 correct. The combined score is 30/40, or 75%. However, first-attempt accuracy is 12/20, or 60%, and repeat accuracy is 18/20, or 90%.
The combined number is mathematically correct but conceals an important difference. It does not show 75% accuracy on unfamiliar material. Nor does the 90% repeat result prove that the original concepts can now be applied in new situations. The useful next step is to explain the corrected reasoning and test related unfamiliar material.
These small sets are also too limited to establish a stable estimate of overall ability. Their topics and difficulty may differ. Use them as prompts for investigation, not as a diagnosis of readiness or a reason to abandon a resource after one disappointing block.
RCEM sets the FRCEM SBA pass mark using an Angoff-derived cut score plus one standard error of measurement. A commercial bank’s percentage is not that standard. See RCEM’s marking information; do not turn a suggested practice target into a claim that you will pass.
Add timing without losing the review stage
The FRCEM SBA contains 180 questions across two two-hour papers, separated by a one-hour break. This gives an overall average of 80 seconds per question during the four hours of answering time. Check the current format with RCEM before planning a full simulation.
For a shorter pacing exercise, 30 questions in 40 minutes uses the same average. It is still a short practice block, not an equivalent test of full-exam endurance or content coverage. If you cannot yet explain the topic, start with learning-focused work before making speed the main goal.
After a timed block, inspect where time went. Did you repeatedly reread stems, spend too long calculating, or change answers without a clear reason? Pick one behaviour to test in the next block. A single average time cannot tell you which of those problems occurred.
Reserve time after the block for review. If your available study period is only 40 minutes, a 40-minute test leaves no space for immediate analysis. Either shorten the block or schedule a separate review appointment that you can realistically keep.
Check explanations against the right sources
If an explanation conflicts with your understanding, identify the exact disputed point before opening more resources. Is the question asking about a different population, a different stage of care or a different outcome? A difference in context may explain what initially looks like a contradiction.
For a clinical recommendation, check the relevant current authoritative guidance and its scope. StudyMedical’s exam revision resources links to organisations including RCEM, NICE and specialty bodies. A resource directory helps you find a source; it does not replace reading the applicable recommendation.
If you still cannot reconcile the answer, record the uncertainty and report it through the provider’s available support route. Do not convert an unresolved explanation into a memorised rule. In particular, preserve the conditions and exceptions that make a clinical recommendation appropriate.
Build a week that fits your rota
Plan sessions by function rather than imposing the same quota every day. One week might include a short unfamiliar set, a focused correction session, a revisit of earlier errors and a timed block. This is an example structure, not a proven optimal schedule.
During a busy week, reduce the size of the set before deleting its review stage. Keep one small follow-up task available—for example, explaining a concept from your error log without notes. Protect rest and schedule substantial practice for a time when you can engage with it.
At the end of the week, ask three questions: which gaps have become clearer, which errors still recur, and which curriculum areas have received no attention? Use the answers to choose the next sessions. A rising attempt count alone cannot answer those questions.
What to do when you remember the answers
Recognising an old question is not a failure; repeated practice has a role. The problem is interpreting recognition as evidence of broader competence. Before viewing the options, explain the likely answer and the feature that makes it appropriate.
Then test transfer with a different item or a self-generated variation. Change one feature and ask whether the answer would change, checking the reasoning against a reliable source. Do not invent a clinical rule simply to make the variation work.
If you have exhausted unfamiliar material in one area, consider appropriate official samples or another reliable source. You do not need to buy another subscription automatically. First establish whether the limitation is material availability, conceptual understanding or the way you are using the existing resource.
Start your next session with a clear question
Choose one purpose, answer actively, review the uncertain decisions and plan a check of what changed. That gives a question-bank session a useful endpoint: a clearer explanation and a specific next action.
Try the free FRCEM SBA demo or explore our FRCEM SBA question bank, which lists over 3,000 questions alongside topic-based custom quizzes, timed practice and new or previously incorrect question modes. Choose the mode that fits the learning task you have identified.
