You have not "covered ABEM" because you reached the end of a question bank. You have covered it when you can show, domain by domain, that you have practised the official blueprint at its real weighting, interpreted the image and data formats the Qualifying Examination actually uses, checked your guidance-sensitive topics against a dated source, and scored adequately on unseen, timed, mixed questions. This article is that evidence checklist — the things to verify before you stop doing new questions, not another study timetable.
What "covered" means for the ABEM Qualifying Examination
Completion is a fact about a product; coverage is a fact about you. A bank can be 100% complete on your dashboard while whole EM Model domains sit at low volume, low accuracy or months since last review. The checklist below turns "I finished the bank" into five separate, testable claims: my blueprint spread is proportional, my format practice is deliberate, my interpretation skills are current, my guidance-sensitive topics are dated, and my performance on unseen items is adequate. If you cannot tick all five with data, you are not done — regardless of what percentage the platform shows. This is the same logic set out in the iatroX pillar on why question-bank completion is not coverage.
The current ABEM exam snapshot
Last checked 19 July 2026. The ABEM Qualifying Examination is a computer-based, single-best-answer multiple-choice test of approximately 305 questions, delivered at Pearson VUE test centres and taken after residency as the first step towards certification; passing it makes a candidate eligible for the Oral Certifying Examination, which is a separate, differently formatted assessment. Content is built on The Model of the Clinical Practice of Emergency Medicine (the "EM Model"), which organises the discipline into roughly 20 content areas weighted from about 2% to 10% each — signs, symptoms and presentations, cardiovascular disorders and traumatic disorders each sit near the top of that range. On top of the content spread the exam applies an acuity frame: roughly 30% critical, 40% emergent, 21% lower-acuity and around 9% non-acute presentations, with paediatric content at 8% or more and geriatric content at 6% or more. Questions may carry stimulus images such as ECGs, radiographs or clinical photographs. Treat these as vendor- and board-reported figures and confirm the current EM Model and exam specifications on abem.org before you rely on any single number.
Build a blueprint coverage table
The instrument that makes coverage visible is a table, not a percentage. Build one row per EM Model content area and fill five columns from your own data.
| Content area (EM Model) | Official weight | Questions attempted | First-attempt accuracy | Last reviewed | Confidence (H/M/L) |
|---|---|---|---|---|---|
| Signs, symptoms & presentations | ~10% | ||||
| Cardiovascular disorders | ~10% | ||||
| Traumatic disorders | ~10% | ||||
| Thoracic–respiratory disorders | verify (EM Model) | ||||
| Abdominal & gastrointestinal | verify (EM Model) | ||||
| Nervous system disorders | verify (EM Model) | ||||
| Toxicologic disorders | verify (EM Model) | ||||
| Environmental disorders | verify (EM Model) | ||||
| HEENT & ophthalmology | verify (EM Model) | ||||
| Psychobehavioural disorders | verify (EM Model) | ||||
| Procedures & skills | verify (EM Model) | ||||
| Other core competencies (EMS, admin, ethics) | verify (EM Model) |
Populate the "official weight" column from the current EM Model document rather than from a bank's internal tags, because a bank can label questions by its own taxonomy and quietly under-represent a low-frequency, high-stakes area. A row is only "green" when three things are true at once: attempted volume is proportional to the official weight, first-attempt accuracy (not your accuracy after seeing the answer once) is at or above your target, and the "last reviewed" date is recent enough that the knowledge is still retrievable. Any row that is high weight, low volume and stale is a content gap the checklist exists to catch.
Ten domain-level blind spots self-selection tends to hide
Left to our own devices we practise what we already tolerate. The following ten areas are the ones that most often stay green on a dashboard yet remain genuinely under-practised, because candidates skip them, guess and move on, or never reach them before deciding they are "done". Treat this list as prompts for exam-specific clinician review, not as a substitute for it.
- Toxicology — specific antidotes, less-common ingestions, toxidrome discrimination and decontamination timing, an area easy to under-sample because the individual agents feel rare.
- Environmental disorders — hypothermia and heat illness staging, drowning, high-altitude illness, envenomation, electrical and lightning injury; low base rate, high consequence.
- Ophthalmology and ENT emergencies — acute vision loss, orbital compartment syndrome, dental and airway emergencies, epistaxis management; frequently image-dependent and often skipped.
- Psychobehavioural emergencies — agitation management, capacity and involuntary hold, and the medical clearance of the psychiatric patient, which candidates tend to treat as "not real medicine".
- Obstetric and gynaecological emergencies beyond the obvious ectopic — ovarian torsion, postpartum haemorrhage and emergencies in the pregnant trauma patient.
- Paediatric-specific presentations — neonatal emergencies, weight-based dosing, non-accidental injury and the febrile infant pathway; the blueprint mandates paediatric content at 8% or more, so under-practice here is disproportionately costly.
- Geriatric-specific presentations — atypical sepsis, polypharmacy and adverse drug events, falls and elder abuse; mandated at 6% or more and easy to fold incorrectly into "general adult".
- Procedures and skills — indications, contraindications and complications of core EM procedures and the interpretation of point-of-care ultrasound findings described in stems.
- Other core competencies — EMS and prehospital care, disaster and mass-casualty triage, medicolegal duties, consent, capacity and error disclosure; the "administrative" content that carries real marks.
- Dermatological and rash-based life threats — Stevens–Johnson syndrome and toxic epidermal necrolysis, necrotising soft-tissue infection and meningococcaemia, where the diagnosis often lives in an image.
The format checklist: rapid prioritisation, images, toxicology and resuscitation sequencing
Coverage is not only about topics; it is about the cognitive formats the exam favours. Verify that you have deliberately practised, and not merely encountered, four things. First, rapid prioritisation — "what do you do first" and "what is the most immediate threat" stems, where the wrong-but-reasonable option is the trap. Second, image-embedded items — you should have worked a meaningful number of questions where the answer depends on reading the picture, not on the surrounding text. Third, toxicology decision-making — antidote selection and the sequence of decontamination, supportive care and specific therapy. Fourth, resuscitation sequencing — the ordering of interventions in unstable presentations, where knowing the components is not the same as ordering them correctly under time pressure. If any of these four is something you have read about but not repeatedly practised as questions, it is a format gap.
The interpretation checklist: images, ECGs, radiographs, data and calculations
Break "interpretation" into its parts and confirm each has been trained on unseen stimuli. ECGs — ischaemic patterns, dangerous rhythms, conduction disease and the ECG mimics that change disposition. Radiographs and described imaging — chest, extremity and cervical-spine films, plus CT patterns described in text. Clinical photographs — rashes, wounds, eyes and ENT findings. Laboratory trends — blood-gas interpretation, lactate, anion-gap and osmolar-gap problems, and toxicology screens read in context rather than in isolation. Calculations — weight-based paediatric doses, corrected values and infusion rates. Ethics and evidence items — informed consent, capacity and the light statistical and evidence-appraisal content the EM Model includes. For each, the standard is the same: can you interpret a novel example at exam pace, or have you only memorised the specific images your bank happened to use?
The recency checklist: guidance-sensitive topics with a dated source
Some answers move. Before you stop, list the topics where current guidance could differ from the version you first learned, and record the date and jurisdiction of the source you are trusting. Guidance-sensitive areas for emergency medicine typically include sepsis recognition and bundle timing, advanced cardiac and paediatric life support algorithms, acute stroke thrombolysis and thrombectomy windows, anticoagulant reversal agents, procedural sedation, and community-acquired infection treatment. For each, write down which body you are following — for example the American Heart Association for resuscitation, the relevant specialty society or the Centers for Disease Control and Prevention — and the year of the guidance. A bank question written three years ago is not automatically wrong, but an undated answer you cannot trace is a recency gap. Never rely on a memorised medicines fact without checking it against current prescribing information.
The performance checklist: unseen, timed, mixed and calibrated
The final block is about whether your knowledge survives exam conditions. Verify five signals. Unseen — your recent accuracy is measured on questions you have not seen before, not on a re-run of items whose answers you now remember. Timed — you are working at roughly the exam's pace, around a minute per item, rather than luxuriating over each stem. Mixed — you practise interleaved blocks that span domains, so the surrounding topic does not cue the answer. High-confidence errors — you specifically track the questions you were sure about and got wrong, because those are the ones the exam punishes hardest. Calibrated — you have benchmarked yourself against official-style material, such as the ABEM in-training experience or reputable board-review sets, so your internal sense of "ready" is anchored to something external. Your raw bank percentage is not any of these signals; the reasons are set out in Your Q-Bank Percentage Is Not Your Exam Score.
The stop/continue decision tree
Use the measured gap, not fatigue or novelty, to choose the next action.
- Continue new questions when whole content areas remain low-volume or low-accuracy — you still have breadth to acquire.
- Consolidate (stop adding, re-test your misses on a spaced schedule) when volume is adequate but retention is slipping and old errors are recurring.
- Simulate (full timed, mixed blocks under exam conditions) when coverage and retention are solid but your speed, stamina or high-confidence error rate is untested.
- Seek teaching when the same error type persists across attempts despite review — that is a conceptual gap a person should unpick, not a volume problem.
- Rest when accuracy is falling with more hours rather than fewer; a fatigued brain manufactures false gaps.
A worked example (invented data)
Consider a fictional candidate, "Dr A", four weeks out. Their table shows cardiovascular and traumatic disorders green — high volume, first-attempt accuracy in the high seventies, reviewed this week. But toxicology sits at 40 questions attempted against a target of 120, accuracy 58%, last reviewed six weeks ago; environmental disorders is worse. Paediatric content is at half the volume its 8%-plus weighting demands. Meanwhile the dashboard reads 84% complete. Reading the table, Dr A does not "keep doing random blocks". They continue new questions in toxicology, environmental and paediatric emergencies to close the volume gap; they schedule a consolidation pass on a cluster of high-confidence ECG errors flagged three weeks ago; and they book one full timed mixed simulation for day five to test pace and stamina. The 84% was never the problem — the untested 8%-weighted paediatric row was. That is the difference between a completion number and a coverage decision.
The one-page ABEM content-gap checklist
Copy this and tick only what you can evidence.
- Blueprint table built, one row per EM Model area, weights taken from the current EM Model.
- Every high-weight area is proportional in volume, at target on first-attempt accuracy, and reviewed recently.
- The ten common blind spots are each at adequate volume and accuracy.
- Deliberate practice logged for rapid prioritisation, image items, toxicology and resuscitation sequencing.
- Interpretation trained on unseen ECGs, radiographs, photographs, lab trends and calculations.
- Guidance-sensitive topics listed, each with a dated, jurisdiction-labelled source.
- Recent accuracy measured on unseen, timed, mixed blocks.
- High-confidence errors tracked and re-tested on a spaced schedule.
- Performance calibrated against official-style material.
- Stop/continue decision made from the measured gap, not from the completion percentage.
Frequently asked questions
How do I know whether I have covered the full ABEM blueprint? You know when your blueprint coverage table shows every EM Model content area practised in proportion to its official weight, with first-attempt accuracy at target and a recent review date, and when the mandated cross-cutting requirements — paediatric content at 8% or more, geriatric at 6% or more, and the acuity spread of roughly 30% critical and 40% emergent — are each represented rather than averaged away; a single completion percentage cannot tell you any of that, which is why the table, not the dashboard, is the evidence.
Can one question bank be enough for ABEM? One well-constructed bank can carry the majority of your breadth, but "enough" is a claim about coverage and calibration rather than about a single product, and no bank should also be the thing that measures whether you are ready, because you will have seen its answers; the defensible approach is to learn from your main bank and then confirm transfer on a second, unseen set, which is the logic of the two-Q-bank rule.
What should I measure instead of my overall Q-bank percentage for ABEM? Measure first-attempt accuracy by EM Model domain, your accuracy on unseen and timed blocks, your high-confidence error rate, your speed against the roughly one-minute-per-item pace, and your retention of previously mastered topics on spaced re-testing; these are the signals that predict exam-day performance, whereas a blended percentage mixes seen and unseen items and hides exactly the domain-level gaps you need to find.
When should I stop doing new ABEM questions? Stop adding new questions when your coverage table is proportional and green, your retention is holding on spaced re-tests, and your unseen, timed, mixed performance is adequate and stable — at that point more novel questions add little, and your time is better spent on consolidation and full simulation; if any of those conditions is unmet, the honest answer is that you are not finished, however high the completion number looks.
Which ABEM resource should I use for my weakest component? Match the resource to the deficit the table reveals: for a knowledge and breadth gap, targeted questions with strong explanations in that domain; for an interpretation gap, a source rich in ECGs, radiographs and clinical images with worked reasoning; for a pace or stamina gap, full-length timed simulation; and for a persistent conceptual error, structured teaching or a review text rather than more random blocks — the weakest component defines the tool, not the other way around.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 19 July 2026; exam figures (question count, EM Model content areas, acuity distribution) are board- and vendor-reported and can change — confirm the current specifications on the official pages before you rely on any number. Disclosure: iatroX operates a competing ABEM question bank; in this article its role is confined to jobs an audit of your own coverage requires — a fresh, unseen, timed baseline and blueprint-weighted measurement — and it is not the source of your blueprint weights, which should come from the EM Model. Corrections are welcome via the feedback route on iatrox.com. References: ABEM Qualifying Examination and Exam Content (abem.org); The Model of the Clinical Practice of Emergency Medicine (EM Model); iatroX, Your Q-Bank Percentage Is Not Your Exam Score; iatroX, question-bank completion is not coverage; iatroX comparison hub.