Before you decide you have "finished" the Concorso SSM, you should be able to tick a specific list of evidence — not a feeling, and not a percentage. This checklist states the minimum proof of coverage that a candidate for the Italian specialty entrance (Scuole di Specializzazione in Medicina) needs before concluding that more new questions no longer add value. It is written as a checklist, not a timetable. The single most important caveat up front: a high question-bank percentage on self-selected questions is the most misleading signal in SSM preparation, because you choose what you practise and you remember what you have seen.
Current exam snapshot
Verified against the 2026 arrangements (last checked 19 July 2026): the Concorso SSM is 140 multiple-choice questions, each with five options and one correct answer, sat in 210 minutes on computer, nationally. Scoring is +1 for a correct answer, −0.25 for a wrong answer and 0 for a blank, so the maximum test score is 140; up to 7 curriculum points (degree grade, grade average, thesis, doctorate) bring the theoretical maximum to 147. Placement is decided by a single national ranking (graduatoria) in which your total score and your ordered list of preferences together determine which specialty and which school you can take.
Two facts about the official material shape everything below. The exam is governed by an annual ministerial decree (bando) from the Ministero dell'Università e della Ricerca, which is the authoritative source for that year's rules — read it, do not rely on a forum summary. And there is no pre-published official question bank (banca dati): the actual exam papers (the prove ufficiali) are released after each sitting, and those past papers are the closest thing to an official calibration gold-standard. The examinable content is broad, spanning the whole medical degree — internal medicine, cardiology, surgery, gynaecology and obstetrics, paediatrics, neurology, pharmacology, infectious disease, psychiatry, emergency medicine and the basic and statistical sciences.
Build a blueprint coverage table
You cannot claim coverage you have not measured. Build a table with one row per domain and these columns: official or observed weight, questions attempted, first-attempt accuracy, last reviewed, and confidence. A worked fragment, using invented data, shows the shape:
| Domain | Observed weight | Attempted | First-attempt accuracy | Last reviewed | Confidence |
|---|---|---|---|---|---|
| Cardiology | ~13% (observed, past papers) | 320 | 71% | 3 days ago | Medium |
| Internal medicine (general) | high | 410 | 68% | 1 week ago | Medium |
| Surgery | moderate | 140 | 59% | 2 weeks ago | Low |
| Gynaecology and obstetrics | moderate | 95 | 62% | 2 weeks ago | Low |
| Statistics and methodology | low but reliable | 30 | 44% | not reviewed | Low |
Note the honest label on "weight": there is no official published per-domain blueprint for the Concorso SSM, so any weighting is observed from past papers, not decreed. Cardiology has repeatedly been among the heaviest domains in released papers, but treat all such figures as estimates, and have an exam-specific clinician sanity-check your domain list before you trust it.
The ten blind spots self-selected practice hides
These are the domain-level areas most likely to stay hidden when you practise what you enjoy. Verify each explicitly, and have them reviewed by someone who knows the exam before you call coverage complete:
- Statistics and study methodology — reliably present, widely under-practised, high yield per hour.
- Basic pharmacology — mechanism, interactions and adverse effects, not just indications.
- Infectious disease and antibiotic choice — including notification and public-health angles.
- Obstetric emergencies — hypertensive disorders, haemorrhage, prenatal diagnosis.
- Paediatric red flags — neonatology, the acutely unwell child, developmental milestones.
- Neurology beyond stroke — demyelinating disease, neurosurgical emergencies, headache.
- Emergency and critical care — the acute, time-critical decisions across systems.
- Endocrinology — thyroid, adrenal and diabetes complications.
- Dermatology and the "minor" specialties — small footprint, easy marks if prepared, easy losses if not.
- Ethics, medico-legal and consent — small in number, disproportionately missed.
Format checklist
Coverage of content is not the same as readiness for the format. Verify:
- Italian-language breadth — you must practise in Italian, across the full clinical vocabulary, because the exam is in Italian and unfamiliar terminology costs marks under time.
- Ministerial simulation — at least one full 140-question, 210-minute sitting under the real −0.25 penalty, not comfortable topic sets.
- National-ranking strategy — a considered approach to the guess-or-blank decision and to ordering your preferences, because the graduatoria, not a raw pass mark, decides your outcome.
On the penalty, do the arithmetic once and internalise it. With five options and −0.25 for a wrong answer, a purely blind guess has an expected value of (1/5 × +1) + (4/5 × −0.25) = 0 — neutral on average. Eliminate even one option and the expected value turns positive, so informed guessing is worth it while pure blind guessing on a question you cannot narrow is, on average, a wash rather than a gain. Know your own accuracy well enough to apply this calmly on the day.
Ranking strategy: your preference order is part of the outcome
The graduatoria does not merely pass or fail you; it places you. Your total score sets your position in a single national queue, and when your turn comes you can only take a specialty-and-school combination that still has a seat. That makes your ordered list of preferences a strategic instrument, not an afterthought. Two candidates with identical scores can end up in different specialties purely because one ordered their preferences carelessly. Verify, well before the exam, that you understand how and when preferences are expressed for your year — the bando is the authoritative source — and that your list reflects genuine priorities rather than prestige defaults, because you may be assigned the highest-ranked option still open when your number is called. Reading last year's cut-offs for your target specialties against your simulated score tells you whether your plan is ambitious, safe or unrealistic, and therefore whether the marginal study hour is better spent raising your score or broadening the list of options you would genuinely accept. This is a content gap of a different kind: a strategy gap that no amount of extra questions will close.
Interpretation checklist
Verify you have practised, not just read about, the interpretive tasks the exam can set: clinical images, ECGs, radiographs, laboratory trends, simple calculations, and the statistics items that reward method over recall. These are precisely the item types that self-selected text-based practice tends to skip.
Recency checklist
Identify the guidance-sensitive topics — oncology, cardiology, infectious disease, and anything where management has recently shifted — and record, for each source you rely on, its date and jurisdiction. The Concorso SSM expects current, Italian- and European-relevant knowledge; an answer that was right three years ago, or that reflects another country's practice, is a silent way to lose marks. Prefer recent, authoritative Italian and European sources for these topics.
Performance checklist
Content and format aside, verify the performance signals that actually predict a good ranking:
- Unseen, timed, mixed blocks — accuracy on fresh items you have not seen, not on your familiar practice pool.
- Speed — a sustainable pace of roughly 90 seconds per item across 140 questions.
- High-confidence errors — the answers you were sure of and got wrong; these are the most dangerous and the most instructive.
- Retention — whether last month's corrections still hold this month.
- Official-material calibration — your performance on a released prova ufficiale sat cold, which is the closest available proxy for the real thing.
Stop or continue: a decision tree
Use the measured gap, not the calendar or your mood, to choose the next action:
- Coverage gaps remain (cold or unreviewed domains) → continue new questions, targeted at those domains.
- Coverage is broad but accuracy is unstable → consolidate: fewer new items, more spaced review of past errors.
- Coverage and accuracy are solid but pacing is shaky → simulate: full 140-item, 210-minute mock papers under the penalty.
- Repeated high-confidence errors in one area → seek teaching, not more solo questions.
- All signals green but performance is declining → rest; fatigue, not ignorance, is now the limiting factor.
The one-page checklist
Copy this and tick it before you conclude you are "done":
- Blueprint table complete, every domain with an attempted count and first-attempt accuracy.
- Each of the ten blind spots explicitly reviewed and clinician-checked.
- At least one full 140-question, 210-minute simulation under the −0.25 penalty.
- Guess-or-blank strategy and preference ordering decided in advance.
- Image, ECG, radiograph, lab-trend, calculation and statistics items all practised.
- Guidance-sensitive topics dated and sourced to current Italian/European guidance.
- Unseen mixed-block accuracy stable at target, with pacing near 90 seconds per item.
- High-confidence errors driven down and retention re-checked after a spacing gap.
- At least one released prova ufficiale sat cold as calibration.
If any box is unticked, you have a content gap, and new questions still have work to do.
A worked example: Giulia decides whether to stop
Giulia is eight weeks from the Concorso SSM. Her overall bank percentage is 74%, which feels reassuring — so, rather than trust the number, she builds the coverage table. It changes the picture. Cardiology and general internal medicine are broad and sit near 70% first-attempt on fresh items, but three rows are cold: statistics and methodology (44%, never reviewed), surgery (59%, two weeks stale) and ethics (untouched). Her most recent unseen mixed block came in at 61%, well below her 74% familiar-pool figure. The gap between those two numbers is the recognition she has been mistaking for knowledge.
She works the checklist. The blind-spots list flags statistics, ethics and obstetric emergencies as under-practised, and all three are confirmed. A full 140-item, 210-minute simulation under the −0.25 penalty exposes a pacing problem: she leaves twelve questions blank, not by strategy but because she ran out of time at roughly 110 seconds per item, slower than the 90 she needs. Two of her wrong answers were high-confidence, both in infectious disease, where she had relied on a source she now finds is three years old.
Her decision follows the tree, not the calendar. Coverage gaps remain, so she does not stop new questions — but she aims them only at the cold domains, not at comfortable cardiology. She schedules weekly full simulations to fix pacing, re-sources her infectious-disease knowledge to current Italian and European guidance, and books a teaching session on statistics rather than grinding more solo items. She leaves one released prova ufficiale sealed for a cold calibration read at week six. The 74% never reappears in her planning; the coverage table and the unseen-block figure do. Every figure here is invented to illustrate the method, not to predict a score.
Three mistakes this checklist is designed to stop
Reading the percentage as the score. Your overall bank figure blends easy repeated items with hard new ones, and it rises as you memorise your own pool. The graduatoria is decided on unseen questions under a penalty, so measure that — first-attempt accuracy on fresh, timed blocks — not the flattering blend.
Confusing "questions attempted" with "coverage". High volume in three comfortable domains is not coverage; a single cold row — statistics, ethics, a minor specialty — is a real gap however large your total count. Coverage is a property of the table, not of the tally.
Never sitting the real format cold. Topic sets without a timer and without the −0.25 penalty do not test the two things that most separate candidates on the day: sustainable pace and disciplined guess-or-blank decisions. At least one full ministerial-format simulation, and one released official paper sat cold, are non-negotiable before you conclude you are done.
FAQ
How do I know whether I have covered the full Concorso SSM blueprint? You know it from a completed coverage table, not a feeling. Because the Ministry does not publish a per-domain blueprint, you construct one from released past papers, list every domain, and record for each an attempted count, a first-attempt accuracy and a review date. Coverage means every domain — including the uncomfortable ones like statistics, ethics and the minor specialties — has been sampled, measured and reviewed, and confirmed by someone who knows the exam. Empty rows are your gaps.
Can one question bank be enough for Concorso SSM? One bank can be your main engine, but relying on it alone has a specific weakness: you eventually recognise its items, so your rising percentage measures familiarity rather than knowledge. The fix is not necessarily a second paid bank for its own sake, but a clean source of unseen items for measurement — released official papers sat cold, and a second bank you keep uncontaminated — so that at least some of what you sit is genuinely new. Volume from one bank; a clean measurement from another.
What should I measure instead of my overall Q-bank percentage for Concorso SSM? Measure first-attempt accuracy on unseen, timed, mixed blocks; per-domain coverage and accuracy; your high-confidence error rate; retention across a spacing gap; and your pace against the 90-seconds-per-item budget. Your overall percentage blends easy repeated items with hard new ones and flatters you. The graduatoria rewards accurate, fast performance on unseen questions under a marking penalty, so those are the numbers to track.
When should I stop doing new Concorso SSM questions? Stop adding new questions when your coverage table has no cold rows, your unseen mixed-block accuracy is stable at your target, your high-confidence errors are rare, and your pacing is comfortable under the penalty. At that point the marginal new question teaches little, and your time is better spent on full simulations, spaced review of past errors, and rest. Doing new questions past this point is usually reassurance-seeking rather than learning.
Which Concorso SSM resource should I use for my weakest component? Match the resource to the failure type. For a knowledge gap, go to a current, authoritative Italian or European source for that domain, then re-test. For interpretation weaknesses (ECGs, imaging, statistics), use targeted item sets that drill exactly those tasks. For pacing and stamina, use full ministerial-format simulations. And for a true calibration read, sit a released prova ufficiale cold — it is the closest proxy for the real exam. 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 rules are set each year by ministerial decree, and any per-domain weighting quoted here is observed from released past papers rather than an official published blueprint — confirm the current year's rules in the official bando. Disclosure: iatroX operates a competing Concorso SSM question bank, so this checklist is deliberately vendor-neutral and confines iatroX to the job of unseen, timed baseline measurement rather than presenting it as sufficient on its own. Corrections are welcome via the feedback route on iatrox.com.
References: Ministero dell'Università e della Ricerca — annual Concorso SSM bando and released prove ufficiali; iatroX — Your Q-Bank Percentage Is Not Your Exam Score; iatroX — question-bank completion is not coverage; iatroX — Concorso SSM: struttura, punteggio e difficoltà; iatroX comparison hub.