CICM examiner reports · fix it Not specific enough Naming the right idea without the dose, interval, threshold or number that earns the mark.
What it looks like: "Start an antibiotic" or "monitor closely" is the idea, not the answer. Marks go to the exact drug, dose, route, interval or threshold, and the options often differ only in those details.
How to fix it: When options differ only in a number, recall the number before you read them. Learn the doses, intervals and cut-offs from the local guidelines, not the general principle.
The drill: 15 questions Questions from the Intensive care bank picked because they test this habit. You answer one at a time and see the explanation straight away.
The first 3 are free, below and in the drill with a free account. Any Intensive care plan opens all 15.
Try 3 now. No account needed. Question 1 of 3 Neurology and neurocritical care
A 38-year-old man is on day 4 in ICU after a severe traumatic brain injury with contusions and an intracranial pressure monitor. He is haemodynamically stable with no active bleeding, and his haemoglobin has drifted to 82 g/L. The bedside nurse asks whether to transfuse. Which approach is best supported by the most recent large randomised trial in acute brain injury? A Withhold transfusion unless haemoglobin falls below 70 g/L, as a liberal strategy gave no benefit B Transfuse one unit now, as a threshold of 90 g/L reduced unfavourable neurological outcome C Transfuse to a target of 100 g/L, as only a target that high lowered mortality at 6 months D Withhold transfusion unless brain tissue oxygen falls, as only physiology-guided transfusion helped
Pick an answer to see the explanation. Next question →
Question 1 of 3 · Neurology and neurocritical care
A 38-year-old man is on day 4 in ICU after a severe traumatic brain injury with contusions and an intracranial pressure monitor. He is haemodynamically stable with no active bleeding, and his haemoglobin has drifted to 82 g/L. The bedside nurse asks whether to transfuse. Which approach is best supported by the most recent large randomised trial in acute brain injury?
Withhold transfusion unless haemoglobin falls below 70 g/L, as a liberal strategy gave no benefit Transfuse one unit now, as a threshold of 90 g/L reduced unfavourable neurological outcome Transfuse to a target of 100 g/L, as only a target that high lowered mortality at 6 months Withhold transfusion unless brain tissue oxygen falls, as only physiology-guided transfusion helped Show the answer B. Transfuse one unit now, as a threshold of 90 g/L reduced unfavourable neurological outcome
TRAIN randomised 850 patients with acute brain injury (about 60% traumatic brain injury, with subarachnoid and intracerebral haemorrhage making up the rest) and haemoglobin below 90 g/L to a liberal strategy (transfuse one unit when haemoglobin fell below 90 g/L) or a restrictive strategy (below 70 g/L). Unfavourable neurological outcome at 180 days (GOS-E 1 to 5) was 62.6% with the liberal strategy versus 72.6% with the restrictive strategy (relative risk about 0.86), and cerebral ischaemic events were also less frequent. The trial was unblinded, but it shifted many units towards a 90 g/L threshold in traumatic brain injury. In traumatic brain injury alone, HEMOTION (100 versus 70 g/L) did not show a significant difference in unfavourable outcome, so TRAIN is the stronger signal for a 90 g/L threshold. A 100 g/L target and brain tissue oxygen guided transfusion were not tested in TRAIN. References: Taccone FS et al. Restrictive vs liberal transfusion strategy in patients with acute brain injury (TRAIN). JAMA 2024.
Why not A: The liberal 90 g/L strategy reduced unfavourable neurological outcome in this trial Why not C: A 100 g/L target was not tested and mortality was not the primary outcome Why not D: Brain tissue oxygen guided transfusion was not the strategy tested Question 2 of 3 · Neurology and neurocritical care
A 52-year-old woman is on day 6 after coiling of a ruptured anterior communicating artery aneurysm, with a modified Fisher grade 3 haemorrhage. She is alert, has no new deficit, and her haemoglobin is 88 g/L. A colleague recommends transfusion to keep haemoglobin above 100 g/L to prevent delayed cerebral ischaemia. Which statement best reflects the evidence specific to this condition?
A 100 g/L threshold gave no significant fall in poor 12-month outcome versus a threshold of 80 g/L A 100 g/L threshold lowered poor outcome only in those with radiological vasospasm A 100 g/L threshold significantly lowered the rate of poor 12-month functional outcome A 100 g/L threshold increased poor outcome because of more thrombotic complications Show the answer A. A 100 g/L threshold gave no significant fall in poor 12-month outcome versus a threshold of 80 g/L
SAHaRA randomised 742 adults with aneurysmal subarachnoid haemorrhage and anaemia to a liberal strategy (transfusion mandated below 100 g/L) or a restrictive strategy (transfusion permitted below 80 g/L). Unfavourable neurological outcome at 12 months (modified Rankin 4 to 6) was 33.5% versus 37.7%, risk ratio about 0.88 with a confidence interval crossing 1, so the liberal strategy did not significantly improve outcome. There was no signal of net harm and no proven benefit in a vasospasm subgroup. This contrasts with TRAIN, where most patients had traumatic brain injury. Many units therefore individualise transfusion in subarachnoid haemorrhage, accepting a threshold between 80 and 100 g/L. References: English SW et al. Liberal or restrictive transfusion strategy in aneurysmal subarachnoid hemorrhage (SAHaRA). N Engl J Med 2025.
Why not B: No vasospasm subgroup benefit was a finding of the trial Why not C: The difference in poor outcome was not statistically significant Why not D: There was no significant excess of poor outcome with the liberal strategy Question 3 of 3 · Haematology and transfusion
A 76-year-old man in ICU with an inferior myocardial infarction treated with PCI 2 days ago now has a haemoglobin of 84 g/L after femoral access site bleeding, which has stopped. He has no angina and is haemodynamically stable. The cardiologist asks whether to transfuse to above 100 g/L. Which statement best reflects the largest randomised trial in this population?
The liberal strategy significantly reduced recurrent myocardial infarction or death at 30 days The restrictive strategy was non-inferior and reduced heart failure, supporting a 70 g/L threshold Death or reinfarction at 30 days was not significantly lower but favoured liberal transfusion The liberal strategy increased death or reinfarction because of volume overload and thrombosis Show the answer C. Death or reinfarction at 30 days was not significantly lower but favoured liberal transfusion
MINT randomised 3504 patients with acute myocardial infarction and haemoglobin below 100 g/L to a restrictive strategy (transfusion permitted below 80 g/L, recommended below 70 g/L) or a liberal strategy (transfusion to keep haemoglobin at or above 100 g/L). Death or recurrent myocardial infarction at 30 days was 16.9% with the restrictive strategy versus 14.5% with the liberal strategy (risk ratio about 1.15, 95% CI 0.99 to 1.34), which was not statistically significant, and death alone was 9.9% versus 8.3%. The direction favoured liberal transfusion, and the trial was not designed as a non-inferiority study. Many clinicians therefore use a higher threshold, around 80 to 100 g/L, in acute coronary syndromes than in general ICU patients. References: Carson JL et al. Restrictive or liberal transfusion strategy in myocardial infarction and anemia (MINT). N Engl J Med 2023.
Why not A: The difference in the primary outcome did not reach statistical significance Why not B: The trial was not a non-inferiority design, and the result did not favour restriction Why not D: Outcomes trended in favour of, not against, the liberal strategy Create a free account to save your results and keep going →
Where the CICM examiners flagged it Second Part Examination (written SAQs, hot cases, vivas) 2026.1 (March/May 2026) : Generic template answersSecond Part Examination (written SAQs, hot cases, vivas) 2026.1 (March/May 2026) : Imprecise data interpretationSecond Part Examination (written SAQs, hot cases, vivas) 2024.1 (March/May 2024) : Missing specificsGeneric "proforma" SAQ answers (ABC, lines, cultures, antibiotics) that are not tailored to the patient in front of you, and missed parts of multi-part questions. All CICM examiner reports →
Fellowship Ready is not affiliated with or endorsed by the CICM. Not affiliated with the CICM.