FRCEM Final SBA stems read like the potted histories your junior colleagues bring you in the resus bay. A few clinical facts, a bit of context, a question at the end. What's being tested is your response as a consultant — combining the clinical context in front of you with the evidence base to land on the right call for this patient.
That's why more than one option is usually clinically reasonable. Both might be things you'd genuinely do in real life. The exam is asking you to make a decision between options that are all reasonable — and choose the one that best fits your patient, in this condition, at this time.
What separates the right answer from the trap isn't knowledge of the disease. It's reading the stem and the lead-in precisely — the clinical context, the patient's current state, and what the question is actually asking for.
The lead-in is the last sentence of the question. It's the bit that actually asks something. It's also the part candidates skim, because they're focused on the stem.
Don't skim it. The lead-in is the clue. Each phrase points to a different type of answer:
The shift from ‘first step’ to ‘next step in disposition’ isn't a stylistic choice — it's a different question with a different right answer.
There's a specific reason these questions feel harder for ED clinicians than for other specialties. In the real ED, you don't have a single ‘next step’. You have a team. While you're calling for blood products, the nurse is putting in a second line, the F2 is doing the secondary survey, the consultant has activated the trauma call. Three things happen in parallel.
The exam can't test that. To make a question reproducible and defensible — to give it one right answer that holds across every sitting and stands up to challenge — the exam has to mark against a fixed standard. Algorithm steps. Protocol thresholds. Guideline-defined criteria.
So when an FRCEM Final stem asks for the ‘next step’, imagine you don't have your team around you. Or imagine that each person can only do one thing at a time. Strip the parallel reality out. What's the single most important next action right now?
That's a different question from the one your ED brain wants to answer. Your ED brain runs three things in parallel — call for blood, second line, secondary survey. The exam can't mark that. It wants the single thing that, on its own, matters most at this moment.
Slow down. Sequence what would normally run in parallel. Then pick.
A 32-year-old man undergoes procedural sedation in the ED for reduction of a posterior shoulder dislocation. He has been pre-oxygenated and is monitored with continuous capnography via nasal cannula. He receives ketamine 1 mg/kg IV.
Within 30 seconds of the consultant beginning the reduction, you note inspiratory stridor, paradoxical chest wall movement, and SpO2 falling from 99% to 88%. The patient is unresponsive to verbal stimulation. Capnography shows a flattened trace.
You recognise this as laryngospasm.
What is the first step in management?
Here are the options:
Why B looks right
It's exactly what your hands would reach for. The patient is hypoxic; you open the airway and give them O2. It's the first physiological intervention — the thing that addresses the falling sats. In the resus bay, this is what you'd do.
Why B is the trap
In real life, you'd do A and B at the same time. You'd stop the consultant AND open the airway AND apply O2. Your hands are on the airway; your voice is on the team. Both happen in parallel.
The exam can't test that. It has to sequence the parallel reality — pick one step that comes first against a fixed standard. Per the algorithm, step 1 is stop the procedure. Step 2 is 100% O2 + jaw thrust + PEEP. They feel simultaneous, but the algorithm sequences them.
Why A is right
It's step 1 of the algorithm. Laryngospasm in this setting is classically triggered by a noxious stimulus on a lightly sedated patient — shoulder reduction qualifies — or sometimes by ketamine itself. Stopping the procedure removes one of the drivers, frees the consultant's hands, and lets the team focus on the airway. Everything else — including the airway intervention in B — flows from it. The exam needs a defensible first step, and the algorithm gives one: this.
Therein lies the difficulty: B isn't wrong clinically. It's just not what comes first in a sequenced reading of the same reality.
Algorithm reference: standard UK ED/anaesthesia teaching (LITFL Laryngospasm; BJA Education, Gavel & Walker 2014).
The patient's current state is the second clue. The same condition with a different presentation has a different right answer.
Unstable / actively deteriorating
The answer is the immediate evidence-based measure. Resus, source control, A–E priorities. The question wants whatever stops the patient getting worse.
Stable / resus already done
The answer is the definitive management. Theatre, cath lab, ITU, IR, transfer. The acute phase is settled — what's the rate-limiting next step?
Read the stem for cues about timing and stability: vitals, what's already been given, whether the question describes ongoing deterioration or a resolved acute phase. The stem tells you which side of the line you're on.
I had an aortic dissection question in one of my attempts. I was torn between two answers. One was an investigation to confirm the diagnosis. One was BP control. I tied myself in knots: do I need the diagnosis first? Do I control the BP before sending him for a CT?
I went back and forth, lost five minutes, and picked the wrong one.
“The answer is in the exact stem wording — and the condition of the patient.”
— my mentor, after that attempt
That changed how I approached the rest of the exam. The condition was the same. The two options were both medically reasonable. What decided the answer wasn't the disease — it was the exact phrasing of the question and what state the patient was in.
If the stem describes a haemodynamically unstable patient with raging hypertension, BP control comes first. If the stem describes a stable patient and the lead-in asks how you'd confirm the diagnosis, it's the CT. Same condition. Same options. Different right answers — because the wording was different.
It forced me to slow down. To really look at the question. And it made the difference on my next attempt.
When you're down to two options and they both look right, don't guess. Run this check:
- Re-read the lead-in. The last sentence. Word for word.
- Ask what type of answer it wants — first step, most appropriate, next step, most useful, most likely.
- Pick the option that matches the lead-in, not just the one that's clinically right.
Ten seconds. That's the difference between matching the medicine and matching the question. On a 180-question paper, it adds up.
This isn't a knowledge gap. It's a technique gap.
If you've sat FRCEM Final and felt like the questions were harder than your knowledge should allow — this is usually why. Smart, well-prepared candidates fail because they pick the clinically correct answer instead of the question-correct answer.
I failed four times before someone showed me how to read the question properly. If you're revising for the next sitting, work this in early. Practise SBAs where every question forces you to decode the lead-in. Build the habit before the exam tests it.
More on what I changed: how I passed after 4 fails and the broader SBA technique guide.
Practise on questions designed for this trap
1000+ FRCEM Final SBAs with instant feedback — trained on exactly this kind of lead-in distinction.