There is no single best resource for the Diploma in Geriatric Medicine (DGM), and the honest answer to "which should I use?" depends on your profile, your time and your budget. This is a decision tree, not a league table. Before you spend anything, one fact resolves half the confusion in this market: the DGM is run by the Royal College of Physicians (RCP), developed with the British Geriatrics Society (BGS) — the RCP administers the exam and publishes the official material, and the BGS does not run it or sell a question bank. Calibrate against the RCP. iatroX supplies the unseen-measurement baseline that tells you which branch of the tree you are on; it covers the Part 1 KBA knowledge layer, not the Part 2 Clinical.
Start here: the exam decides the shape of your stack
The DGM has two components with different resource needs (RCP, last checked 21 July 2026). Part 1, the KBA, is an online paper of 100 best-of-five questions over three hours, once a year, no negative marking (2026 fee £336). Part 2, the Clinical, is a four-station OSCE of roughly 90 minutes at the RCP in Liverpool (2026 fee £642), with a four-year window from KBA success. The KBA is a knowledge problem, best served by banks and official material; the Clinical is a performance problem, best served by coached practice. Buying an MCQ bank for a Clinical gap is the most common resource mistake in DGM.
The KBA blueprint (RCP syllabus, February 2025) also tells you where to spend effort:
| Domain | Weight (of 100) |
|---|---|
| Common geriatric problems / syndromes | 40 |
| Other illnesses affecting older people | 15 |
| Pharmacology and therapeutics | 15 |
| Administrative aspects, ethics and law | 10 |
| Principles of rehabilitation | 8 |
| Epidemiological, demographic and social factors | 5 |
| Transfers of care and community care | 5 |
| Basic science and biology of ageing | 2 |
Segment yourself first: six candidate profiles
| Profile | Dominant risk | First resource move |
|---|---|---|
| First attempt | Unknown coverage; over-buying in a thin market | Unseen baseline first; map weak domains before spending |
| Retake | Repeating the same errors | Error-log review + fresh unseen items, not a re-run |
| Doctor caring for older adults (not a geriatrician) | Patchy exposure to specialist syndromes and law | Prioritise syndromes, pharmacology and capacity; book Clinical practice early |
| Weak foundations | Core knowledge gaps | Teaching/reference source first, then a bank |
| Strong knowledge, poor pacing | Timing out on a 100-item, 3-hour paper | Timed mixed blocks, not more content |
| Strong MCQ, weak practical | KBA-ready but under-prepared for the OSCE | Coached station practice; a bank will not help |
Define the minimum stack
The DGM market is genuinely thin — there are few dedicated banks — so the temptation is either to over-buy or to rely on a single small resource. The minimum effective stack is four slots, two often empty:
- One primary Q-bank for KBA knowledge and retrieval.
- Official RCP calibration material — syllabus, blueprint and sample questions. Free and authoritative.
- One teaching/reference source, only where foundations are weak — BGS clinical guidance, a geriatric medicine reference, or e-learning.
- One modality tool for the Clinical, only if the OSCE is in scope — coached practice, a course, or supervised sessions.
A second MCQ bank is worth adding only to measure yourself on unseen items, never to re-read the same content (the two-Q-bank rule).
Budget bands (verify every price on the day you buy)
| Band | KBA knowledge | Clinical | Notes |
|---|---|---|---|
| Free / low-cost | RCP sample questions + Munro free items + iatroX free UK-core for unseen measurement | Study partner + supervised clinics | Viable if knowledge is strong and you have geriatric contact time |
| One premium resource | One DGM bank (Munro £10 mock, or an iatroX subscription — verify current price) | One coached OSCE block | The common sensible stack |
| Comprehensive stack | Premium bank + iatroX for unseen transfer practice and analytics + reference source | Course + repeated supervised station practice | Justified for weak foundations or a prior fail |
The Munro DGM Question Bank publishes free sample questions and a 100-question mock for £10 (vendor-reported, 21 July 2026); it has no published adaptive features and no stated total count, so verify current terms on the product page. Confirm any iatroX subscription price on its page before you buy.
Time bands: what to omit, not what to add
Under four weeks. Do not start a new full bank. Take one unseen timed block to find your two weakest high-weight domains (usually syndromes, pharmacology or capacity), drill only those, and do daily short mixed sets for pacing. If the Clinical is also within four weeks, protect two coached sessions and cut KBA volume to fit. Omit reference reading.
Four to twelve weeks. The standard window. Run one primary bank as retrieval — topic-filtered only where a domain is weak, mixed thereafter — calibrate weekly against unseen items and the RCP samples, and begin Clinical practice at week two if in scope. Omit the second bank.
More than twelve weeks. Room for a reference source to rebuild foundations before the bank, plus spaced review. Guard against passive re-reading by scheduling unseen measurement fortnightly. Omit nothing essential, but do not let extra time turn into re-reading instead of retrieval.
The decision matrix: platform → best job
| Resource | Best job | Not the right tool for |
|---|---|---|
| RCP official material (syllabus, blueprint, sample questions) | Official calibration and format truth | High-volume retrieval |
| Munro DGM Question Bank | Low-cost dedicated DGM MCQ volume and mocks (KBA) | The Clinical; large-scale adaptive analytics |
| iatroX | Unseen-MCQ measurement, UK-core knowledge, Socratic review of misses | The Part 2 Clinical OSCE (it does not simulate stations) |
| BGS clinical guidance / geriatric reference / e-learning | Teaching and content for weak foundations | Timed exam-style measurement |
| Coached OSCE practice / course / supervised clinics | Realism and performance for the Clinical | Building KBA knowledge |
Cannibalisation guardrail. This hub keeps platform descriptions deliberately short. For the detailed evidence — how a bank maps to the blueprint, how to run its first-pass-review-exit workflow, and what MCQs cannot train — read the narrow child articles: the Munro first-pass, review and exit plan, the DGM content-gap checklist, the DGM modality-gap guide, and the iatroX comparison hub. This page makes the decision; those pages carry the evidence.
Three worked candidate walk-throughs
Profile A — doctor caring for older adults, weak foundations, ten weeks, KBA only. Week one: RCP sample questions plus a 40-item unseen baseline; the coverage table shows pharmacology and capacity floored. Weeks two to four: a geriatric reference plus topic-filtered blocks on those two domains, error-coding every miss, sourcing medicines from the SmPC/eMC. Weeks five to eight: switch to mixed timed blocks; calibrate on unseen items and RCP samples each weekend. Weeks nine to ten: two timed mocks, reviewing high-confidence errors first. Exit criterion: unseen timed accuracy stable across two mixed blocks with no floored domain.
Profile B — strong knowledge, poor pacing, five weeks, KBA. Content is not the issue; the three-hour, 100-item clock is. Skip the reference source. Do daily timed mixed blocks with strict pacing discipline, reviewing only slow or wrong items, on unseen stems. Exit criterion: completing full-length timed blocks with margin while accuracy holds — pacing is the readiness signal, not percentage.
Profile C — passed the KBA, anxious about the Clinical, eight weeks. A bank is the wrong tool. Spend the eight weeks on the four-station format: coached history and communication practice with a role-player, a data/prescribing station rehearsed against the SmPC/eMC and NICE/CKS, capacity and frailty stations marked against the RCP descriptors, and real geriatric clinics for exposure. Use iatroX only to keep KBA knowledge warm. Exit criterion: completing a full mock circuit to time with a colleague marking against the rubric.
The decision tree in one page
| If this is true | Then your next move is |
|---|---|
| No unseen baseline yet | Take one before buying anything |
| Weak foundations, > 8 weeks | Reference source first, then one bank |
| Strong knowledge, timing out | Timed mixed blocks; no new content |
| KBA-ready, Clinical soon | Stop buying banks; book coached OSCE practice |
| A high-weight domain is floored | Targeted blocks on that domain only |
| Under four weeks, unprepared | Triage to two weakest high-weight domains |
| Everything green on unseen timed blocks | Stop adding questions; consolidate and rest |
The evidence hierarchy behind every recommendation
Rank the evidence when sources conflict. Official RCP material first for format and blueprint — the College defines the exam. Primary UK guidance next for content: NICE, CKS, SIGN, BGS clinical guidance, the SmPC/eMC for medicines, and NHS clinical content for every medicines and management fact. Vendor pages for product facts only, dated and labelled vendor-reported. Independent testing and your own unseen scores last for user experience and readiness. A vendor's claim is marketing; your unseen timed accuracy is data (your Q-bank percentage is not your exam score).
Three mistakes this decision tree is designed to stop
The first is buying an MCQ bank to fix a Clinical gap — different problem, different tool. The second is over-relying on a single small resource in a thin market and mistaking "finished it" for "covered the blueprint", when the 40-mark syndromes domain may still be under-practised. The third is skipping the free, authoritative RCP material because a paid bank feels more serious — the official samples are the only items guaranteed to match the real interface and standard.
FAQ
How do I know whether I have covered the full DGM blueprint? Build a coverage table from the RCP February 2025 blueprint and populate it as you revise — a row per domain, columns for questions attempted, first-attempt accuracy, last reviewed and confidence. You have covered the blueprint when every domain, weighted by its share of the 100 marks, has adequate unseen sample volume at a stable accuracy and none is floored. Because syndromes, other illnesses and pharmacology carry 70 marks between them, coverage is about weighting, not raw totals (completion is not coverage).
Can one question bank be enough for DGM? For the KBA, one blueprint-matched bank plus the RCP official samples and some unseen measurement can be enough. For the diploma as a whole it cannot, because no bank prepares you for the Part 2 Clinical OSCE. And in a thin market like DGM, even for the KBA it is wise to calibrate against the RCP samples and measure on unseen items from a second source so your score reflects reasoning rather than recall.
What should I measure instead of my overall Q-bank percentage for DGM? Measure per-domain unseen accuracy against the blueprint weighting, pacing against the three-hour budget, your high-confidence error rate, and retention of missed items after spacing. The overall percentage averages away floored domains, over-confidence and slow pacing — the three things that actually fail candidates — so it is the least informative single figure to watch.
When should I stop doing new DGM questions? Stop when your KBA coverage is even, your unseen timed accuracy is stable across two or more mixed blocks, your pacing is comfortable and your error log is quiet. Beyond that, new questions mostly reassure, and your time is better spent consolidating, resting, or — if the Clinical is in scope — practising stations. Continuing to drill past that point is sunk-cost studying.
Which DGM resource should I use for my weakest component? Match the tool to the weakness. Weak KBA knowledge in a domain: targeted blocks in a blueprint-mapped bank plus unseen transfer practice. Weak pacing: timed mixed blocks. Weak Clinical performance: coached, rubric-anchored OSCE practice with a clinician and role-players — not a question bank. iatroX is a strong option for the knowledge and unseen-measurement jobs and is explicitly not an OSCE simulator, so pair it with real performance practice for the Clinical.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 21 July 2026. The Munro DGM Question Bank figures (free samples, £10 mock, no published adaptive features or total count) and any iatroX subscription price are vendor-reported as of 21 July 2026 and change without notice — verify each on the product page on the day you buy. Disclosure: iatroX operates a competing UK question bank that includes DGM-level content; this article confines the iatroX role to unseen-MCQ measurement and KBA-level knowledge and states plainly that it does not reproduce the Part 2 Clinical OSCE. Note that the DGM is administered by the RCP, not the BGS, which neither runs the exam nor sells a bank. Corrections via the feedback route on iatrox.com. References: RCP Diploma in Geriatric Medicine and February 2025 syllabus (rcp.ac.uk); BGS DGM page (bgs.org.uk); Munro DGM Question Bank (dgmquestionbank.com); and, internally, the iatroX DGM bank, the iatroX comparison hub and Your Q-Bank Percentage Is Not Your Exam Score.