Common FRCEM SBA mistakes that cost marks

TL;DR — the five mistakes that quietly cost the most marks in the FRCEM SBA:

1. Answering before reading the lead-in — picking the right treatment for the wrong question (“initial” vs “definitive”, “most appropriate next step” vs “definitive management”).
2. Burning four minutes on one toxicology curveball and then rushing through ten easy paediatric stems you would have got right cold.
3. Skipping the “dry” SLOs — governance, statistics, safeguarding, EBM — so a fixable 15-20 marks vanish before you sit down.
4. Over-thinking common presentations into zebras the stem never mentioned.
5. Doing question banks passively — reading explanations without testing yourself the next day, so nothing transfers under exam pressure.

Facts last verified 30 May 2026.

If you have just walked out of a mock — or worse, a real sitting — feeling like you knew the medicine but somehow the marks did not land, you are in good company. Most FRCEM SBA candidates who fail do not fail because they do not know enough emergency medicine. They fail because of a small handful of repeatable, observable behaviour mistakes during the paper. These are the mistakes that turn a 58% into a 64%, and a borderline fail into a comfortable pass.

This article is the behaviour-mistake companion to our piece on cognitive errors in exam questions. That one covers what is happening inside your head — anchoring, premature closure, availability bias. This one is about what you are doing with your hands, your eyes, and your clock. Both matter. The behaviour fixes are usually quicker wins.

For more on this, see our guide to why skipping mock exams is the biggest FRCEM SBA mistake.

Why do most FRCEM SBA candidates lose marks they should have got?

The FRCEM SBA is 180 single-best-answer questions over four hours. That is 80 seconds per question on average. RCEM blueprints the paper to the 2021 Emergency Medicine Curriculum — which means the spread is fixed, predictable, and not weighted towards the topics you find most exciting. Resus and trauma will absolutely come up. So will clinical governance, audit cycles, sensitivity and specificity, safeguarding pathways, and end-of-life decisions. None of that is hidden — it is in the published blueprint.

Failure analysis from candidates who have sat the paper more than once almost always lands in the same place: knowledge was adequate, exam behaviour was not. The fix is rarely “read another textbook.” It is usually “change three habits.”

Mistake 1: Answering before you have actually read the lead-in

This is the single biggest mark-leak in the SBA. You see a stem about a septic patient, you skim, you spot “IV antibiotics within the hour” in the options, you click. The lead-in, which you did not read carefully, was “What is the most appropriate immediate step?” — and the correct answer was high-flow oxygen, or airway management, or a fluid bolus, depending on the physiology.

RCEM SBAs are written with deliberate distinction between:

• Most appropriate initial / immediate step — what happens in the first 60 seconds. Think ABCDE, think safety, think the action you would do before you had time to think.
• Most appropriate next step — what comes after you have stabilised the obvious thing. Often a specific investigation or escalation.
• Most appropriate definitive management — what fixes the underlying pathology. Often the “correct” pharmacology or procedure.
• Most likely diagnosis — pattern recognition, not action.

All four can appear in the same options list. The right treatment for the wrong question scores zero.

For more on this, see our guide to why qbank pass does not predict FRCEM SBA pass.

The fix: read the lead-in twice. Cover the options with your hand or with the on-screen masking tool until you have answered the question in your own head. Only then look at the five options and pick the closest match. This single change typically adds 5-8 marks across a paper.

Mistake 2: Spending four minutes on one question while ten easy ones quietly slip away

The maths is unforgiving. At 80 seconds per question you have no slack. Every minute you spend wrestling with a niche toxicology stem is a minute you are not spending on the paediatric fluid calculation or the straightforward ECG further down the paper.

Candidates describe a recognisable “spiral of doom”: one hard question burns three minutes, you panic, you rush the next ten, you make sloppy errors on stems you would have answered correctly with thirty seconds of calm reading.

The fix: use a hard 90-second cap. If you have not landed on an answer by 90 seconds, do all three of these:

1. Eliminate the two options you know are wrong.
2. Pick the most likely of the remaining three.
3. Flag it and move on. You can return if there is time at the end.

You will never lose marks faster than by sacrificing five easy questions to chase one hard one.

Mistake 3: Skipping the “dry” parts of the blueprint

Emergency physicians love resus, trauma, paediatrics, and ECGs. We tolerate toxicology. We tend to avoid clinical governance, statistics, safeguarding pathways, end-of-life law, and EBM.

The exam blueprint does not share your preferences. Candidates who skip the “dry” SLOs — particularly research methodology, audit, clinical governance, and the legal-ethical content — routinely lose 15-20 marks that were genuinely retrievable with a weekend of targeted reading. One Reddit comment from a candidate who failed by a single mark put it bluntly: “Skipped EBM entirely and only revised core subjects like anatomy, physiology, and pharmacology. Missed by just one mark. That stings.”

For more on this, see our guide to how to bounce back after failing FRCEM SBA.

The fix: do the dry topics first, when your brain is fresh. Block them at the start of each study week, not the end. If sensitivity, specificity, positive predictive value, likelihood ratios, and the audit cycle do not feel automatic by the month before your exam, you have unfinished homework.

Mistake 4: Over-thinking common presentations into zebras

This one disproportionately affects experienced registrars and second-attempt candidates. You have seen unusual cases in real life. You read “central crushing chest pain in a 58-year-old smoker” and start hunting for aortic dissection, pulmonary embolism, or oesophageal rupture because you are wary of the obvious answer.

RCEM SBAs are written to test safe practice on the commonest presentations. Unless the stem explicitly gives you a discriminator — tearing pain radiating to the back, unequal upper-limb pulses, sudden onset on exertion in a thin tall young man, recent endoscopy — assume the common diagnosis. The stem will tell you when it wants you to think of a zebra.

The fix: when you find yourself reaching for a rare diagnosis, ask: “Did the stem give me a specific clue, or am I inventing one?” If you are inventing, go back to the common answer.

Mistake 5: Passive question-banking instead of active recall

This is the most expensive long-term mistake, because it makes every hour of revision less efficient than it should be. Doing 50 questions, reading the explanations, ticking “reviewed” and moving on is recognition, not recall. Under exam pressure, recognition collapses.

Candidates who pass on first or second attempt almost universally describe the same workflow: do a block of questions under timed conditions, mark them, then re-test the same content cold the next day from memory before reading the explanation again. Spaced repetition on your wrong answers, not bulk re-reading, is what makes the knowledge land.

The fix: set a rule that any question you get wrong gets re-tested 24 hours later, then a week later, then a month later. A small spaced-repetition deck of your own mistakes is worth more than 1,000 fresh questions you skim once.

Mistake 6: Mock exams in pyjamas, on the sofa, with the dog

Comfortable mocks lie to you. They tell you your score is X, when your real-conditions score is X minus 5 to 10%. The FRCEM SBA at a Pearson VUE centre is four hours of sustained concentration, a stranger’s keyboard, ambient noise, no snacks, no music, and a clock you cannot pause. If your mocks do not reproduce that, you are not training the skill that gets tested.

The fix: at least two of your final mocks should be done full-length, timed, in one sitting, with no phone and no interruptions. Ideally at a desk and chair you do not normally use. The first time you discover that four hours of focused reading hurts your eyes and back should not be exam day.

Mistake 7: Ignoring the question feedback after each mock

This is the mock-analysis sister of Mistake 5. Sitting a mock and only looking at the headline percentage is the academic equivalent of weighing yourself without changing your diet. The mark itself is almost useless. What matters is the category breakdown — where did the marks fall off — and the mistake type for each wrong answer:

• Knowledge gap — you did not know the fact. Fix: targeted reading.
• Misread the question — you read “initial” as “definitive”, or missed a key negative. Fix: lead-in discipline (Mistake 1).
• Rushed — you picked the first plausible option. Fix: 90-second cap (Mistake 2).
• Talked yourself out of the right answer — classic over-thinking. Fix: Mistake 4.

If you do not know which bucket each wrong answer falls into, you cannot fix the right thing.

Mistake 8: Believing the exam is trying to trick you

It is not. RCEM SBAs are written to a published style guide and peer-reviewed. The question writers are trying to test whether you would be safe on the shop floor — not whether you can spot a hidden semantic trapdoor. If an answer looks straightforwardly correct, it usually is. The hardest thing in the SBA is choosing between two reasonable answers where one is fractionally more appropriate for the patient in front of you. That is the test of registrar-level judgement, not a trick.

Candidates who go in believing the paper is adversarial tend to second-guess correct answers, change them, and lose marks they had. Believe the question. Pick the most appropriate answer for this specific patient and this specific lead-in, and trust your reading.

The mistake-to-marks map

A rough guide to where the marks tend to leak and how to plug them. Mark estimates are typical ranges across a 180-question paper; your own audit will give you a more accurate number.

How is this different from cognitive errors?

Cognitive errors — anchoring, availability bias, premature closure, confirmation bias — are internal. They are what your brain does to the question while you are reading it. We cover those in detail in Common Cognitive Errors in Exam Questions.

The mistakes in this article are external and observable. They are about your eyes, your clock, your study schedule, and your mock-exam habits. They are easier to fix because you can audit them objectively: did you read the lead-in twice, yes or no? Did you spend more than 90 seconds on any single question, yes or no? Did you do your last mock under exam conditions, yes or no?

Most candidates need to address both. The behaviour fixes usually land first and free up the cognitive capacity to defuse bias mid-stem.

What should you actually do in the four weeks before your exam?

If you have one month left, ruthlessly prioritise. In rough order:

1. Audit your last three mocks. Bucket every wrong answer into the four mistake types above. Whichever bucket is biggest is your next two weeks of work.
2. Block the dry SLOs. Statistics, governance, EBM, safeguarding, end-of-life law. A weekend each is usually enough to take you from “avoidant” to “competent.”
3. Two full-length timed mocks, minimum. One around three weeks out, one around one week out. Same time of day as your real exam, no interruptions.
4. Spaced repetition on your wrong-answer deck. Daily, short sessions, only the questions you have got wrong before.
5. Stop hoovering up new content in the final week. Consolidation beats novelty.

Frequently asked questions

How many marks do most candidates lose to exam technique rather than knowledge?

There is no published RCEM figure, but failure analyses from candidates who pass on a second attempt typically describe a 5-10% swing on the same baseline knowledge. That is the difference between a marginal fail and a comfortable pass on most papers.

Is the FRCEM SBA pass mark fixed?

No. RCEM sets the pass mark for each sitting using a modified Angoff method against the standard expected of a day-one consultant. It typically lands in the high 50s to low 60s percent, but it varies sitting to sitting. Aim higher than the historic average — do not aim at it.

How long should I actually spend per question?

The maths says 80 seconds. In practice, easy questions take 30-40 seconds and hard ones can take 90-120. Aim to bank time on the easy ones so you have a small buffer for the hard ones. If you have not answered after 90 seconds, guess, flag, and move on.

Are the questions in the exam similar to RCEM Learning practice questions?

Stylistically, yes. RCEM Learning content is mapped to the same blueprint and written in the same house style. The real exam has slightly more clinical nuance and tighter discriminators in the options, but if you can comfortably handle RCEM Learning at speed you are in the right ballpark.

I failed by a small margin. Should I change my whole study method?

Probably not. If you failed by less than 5 marks, you almost certainly do not have a knowledge problem — you have a technique problem. Audit your mock-exam habits and your lead-in reading discipline before you go back and re-read the curriculum.

How many practice questions do I need to do?

Volume matters less than depth. Most candidates who pass first time describe doing 1,500-3,000 questions with full active-recall review rather than 6,000 done passively. Quality beats quantity. A reliable, blueprinted bank used properly is enough.

Is the SBA harder than the MRCEM Intermediate SBA?

The format is similar but the FRCEM SBA expects registrar-level judgement and a deeper grasp of governance, EBM, and end-of-career-stage decision-making. Candidates often describe it as broader rather than harder per question.

Should I trust my gut and change answers, or stick with my first instinct?

Stick with your first answer unless you have a specific reason to change it — a misread, a re-read of the lead-in, or a piece of information you missed. Changing answers based on vague doubt costs more marks than it saves.

How do I stop panicking when I hit a question I do not know?

Have a pre-decided protocol. The moment you realise you do not know: eliminate two wrong options, pick the most likely of the three, flag, move on. The decision is already made before exam day, so there is nothing to panic about.

I keep failing the dry SLOs. What is the minimum I need to know?

For statistics: sensitivity, specificity, positive and negative predictive values, likelihood ratios, p-values, confidence intervals, study designs and their levels of evidence. For governance: the audit cycle, incident reporting, root cause analysis, duty of candour. For ethics and law: capacity, consent, safeguarding referral routes, deprivation of liberty. That is roughly a focused weekend of work.

Are second-attempt candidates penalised by examiners?

No. RCEM exams are marked blind to attempt number. If you have failed once, the marks are sitting there for you to take with a sharper technique. Most second-attempt passes describe smaller behaviour changes, not a knowledge overhaul.

Where should I go from here?

Start with the cognitive-error companion piece, then build a spaced-repetition deck of your own wrong answers, then schedule two full-length timed mocks. The full FRCEM SBA preparation library is at emfinalexams.com.

Facts last verified 30 May 2026. RCEM blueprints and curriculum content are kept under review by the College — check the RCEM website for the current published version before your sitting.

Ready to build your plan? EMF Premium gives you all 40,000 questions and 20 mocks for £59 — one payment, six months' access.