The idea that organises this guide is consultation-anchored generalism: the same clinical fact produces a different consultation depending on the patient in front of you. That makes attaching context to knowledge, not collecting facts, the central revision task for MRCGP. Start by learning one detailed consultation model as your structure, convert each clinical topic into the questions, screens, shared options and safety nets a real appointment would need, and rehearse the conversion through recorded self-audits and applied scenarios rather than passive reading.
Turning clinical knowledge into consultation behaviours
Consultation-anchored generalism is the discipline of attaching every clinical fact to a decision inside a consultation: what you ask, what you exclude, what you offer, and what you agree. Revision that stops at facts leaves that conversion undone.
The concept is easier to see as a contrast. A statement such as 'drug Y is first-line for condition X' is stable, but the decision is not: for a woman planning pregnancy, an older patient on interacting medication, or a patient whose real fear is a specific serious disease, the same guideline sentence produces different consultations. Generalism is exactly this work of binding evidence to context, so each revision topic needs the context attached at study time rather than reconstructed under pressure.
Practically, convert topics with a four-part template. After reading a topic, write down: three questions that would change your working diagnosis; the red flags you would screen for; two options you could legitimately offer and what decides between them; and a safety-net sentence. The RCGP's own learning resources, including its Essential Knowledge Updates on new and changing aspects of practice, are built around changes to practice rather than stable facts, which fits this template well - each update simply becomes fresh material for the same four boxes.
Separating data gathering from management in scenarios and rehearsals
Data gathering builds a focused history, screens for red flags, and surfaces the patient's agenda. Management negotiates a plan. Revising them as one block produces fast-sounding consultations that commit to plans on incomplete information.
In written items, this separation is a classification skill. Before answering, label the vignette: is it asking you to gather (the next best question, the investigation that clarifies) or to commit (the next management step, the best plan)? If a stem leaves a critical uncertainty open - an unexplored red flag, an unstated patient factor, an unvoiced concern - the defensible next step is usually information. Make the labelling explicit during practice sets, and the choice stops feeling like guesswork.
In rehearsed consultations, keep the two phases as distinct blocks: open with the patient's agenda, take a targeted history, screen for red flags, summarise and check, then move to shared management. If you hold the boundary deliberately, you create a checkpoint where you can notice drift - for example, offering prescribing advice while the patient's real worry about cancer has not yet been spoken - and pause to return to data gathering instead of ploughing on.
Choosing between Calgary-Cambridge, Pendleton, Neighbour and RICE
Calgary-Cambridge gives the most granular structure; Pendleton foregrounds the patient's perspective; Neighbour offers five quick checkpoints; RICE compresses patient-perspective work into one step. Learn one detailed and one compact model rather than all four superficially.
Calgary-Cambridge segments the consultation into initiating the session, gathering information, physical examination, explanation and planning, and closing, with tasks such as agenda-setting and building shared understanding running throughout. Pendleton asks for the patient's reasons for attending, their ideas about the problem, and its impact on their life before any joint plan is made. Neighbour's five checkpoints - connecting, summarising, handover, safety netting, housekeeping - are short enough to run silently during a live consultation.
RICE - reason, ideas, concerns, expectations - compresses the perspective work into a single step, which fits the management discussion or a written plan. A workable pairing is one detailed model (Calgary-Cambridge) as your rehearsal structure and one compact prompt (RICE or Neighbour's checkpoints) as an in-consultation checklist. The table below shows how the four models differ in emphasis, so you can match the scaffold to the task instead of juggling all four at once.
| Model | Core stages or tasks | Best revision use |
|---|---|---|
| Calgary-Cambridge | Initiating, information gathering, examination, explanation and planning, closing; agenda-setting and shared understanding throughout | Building the full consultation structure and checking no phase is skipped |
| Pendleton | Reason for attending, patient's ideas, impact on daily life, joint management | Practising a patient-centred opening and exploring illness impact |
| Neighbour | Connecting, summarising, handover, safety netting, housekeeping | A fast in-consultation checklist, especially for closing well |
| RICE | Reason, ideas, concerns, expectations | Compressing patient-perspective exploration into the management discussion |
Keeping red-flag screening and safety netting distinct
Red-flag screening is a data-gathering act: systematically excluding serious causes while forming a diagnosis. Safety netting is a management act: naming what could worsen, what to watch for, and how to return. Scenario 1 shows both.
Because the two sound similar in everyday speech, they tend to collapse together in revision. Keep them apart by timing: screening questions belong while the history is still open; the safety-net plan belongs after you and the patient have agreed on management. In written items the distinction changes the answer - a stem asking what to do if symptoms change is asking for contingency advice, not for the urgent action you would have taken had the red flag been present now.
Worked scenario 1 (simulated consultation). A 54-year-old man describes three weeks of chest discomfort he attributes to indigestion. Plausible mistake: reassure on the strength of a normal examination and move straight to antacids. Better decision: ask about exertional relationship, radiation, breathlessness and risk factors, then frame the cardiac question openly with appropriate onward assessment, and finish with explicit safety netting - what to do if pain occurs at rest or intensifies. Why it matters: a plausible benign explanation never removes the duty to exclude the serious one and to state the contingency plan aloud.
Checking patient factors before committing to guideline answers
Run a two-step habit for prescribing decisions: name the guideline-concordant option, then check it against the specific patient's contraindications, interactions, organ function, pregnancy plans and existing therapy before committing to it.
Train the check deliberately on primary care drug classes where patient factors commonly overturn the default: antihypertensives in women who may become pregnant, NSAIDs in renal impairment, anticoagulants with interacting therapy or falls risk, and antibiotics in pregnancy or allergy. For each class, keep a one-line list of the patient factors that force a rethink. In practice sets, run the check before you look at the options, so the vignette's patient details are read as data rather than decoration.
Worked scenario 2 (applied item). A 30-year-old woman with newly diagnosed hypertension is planning a pregnancy. Plausible mistake: choosing the drug class most often associated with uncomplicated hypertension. Better decision: pause on two factors - several standard antihypertensive classes carry fetal risk, so the option consistent with safer use in pregnancy planning is the correct pick - and note that pre-conception review of the regimen is part of the plan. Why it matters: the reflexive guideline answer is evidence-based in general and unsafe for this patient, and the check is what converts one into the other.
Scoring your own consultations against a fixed rubric
Record a simulated consultation on an unseen written case, then score it immediately against a fixed rubric. Self-observation converts vague advice such as 'be more patient-centred' into specific behaviours you can rehearse and re-test.
Set up with a study partner playing the patient, or alone: read a written case you have not seen, then improvise the consultation aloud as if the case details were answering you, recording as you go. Score immediately afterwards while memory is fresh. Repeat the same case a week later only if the first attempt exposed a structural fault such as missing agenda-setting; otherwise move to a new case, which tests whether the fix transfers.
Rate each rubric item from 1 (absent) to 4 (routine and fluent). Treat a total around 20 or above as a learning milestone showing the structure is in place - a study indicator, not a prediction of any assessment outcome. Expected observations when the exercise is working: your first red-flag questions arrive within the opening minute, the patient's own words for their concern appear in your summary, and your closing contains a named symptom, a named action, and a route back.
- Opens with the patient's agenda before closed questions begin
- History is targeted to the presentation rather than exhaustive
- Red-flag screening is audible as explicit questions
- Ideas, concerns and expectations are explored in the patient's own terms
- Management is shared, with options, rationale and agreement checked
- Safety netting names a symptom, an action, and a route to follow-up
Sequencing three revision passes and readiness checks
Sequence preparation as structure, then knowledge, then integration: consultation models and phase discipline first, applied clinical content second, mixed timed scenarios with self-audit third. Adapt the durations to your calendar; the order is the transferable part.
Pass one: learn one detailed and one compact model, then run three recorded consultations purely for structure, ignoring content quality - the aim is fluency of phases. Pass two: work clinical topics through the four-part template and the prescribing check, one presentation at a time. Pass three: interleave timed written scenario sets with recorded consultations scored on the rubric, so classification, patient-factor checks and structure are exercised together under time pressure.
Readiness checks: you can name the stages of your two chosen models from memory and say which tasks sit in each; given any vignette, you can state within seconds whether it calls for more data gathering or a management decision; you can generate a safety-net plan unprompted for any common presentation; and your last three self-audits reach the rubric milestone. When those hold, test yourself against practice questions - the free practice sets for this credential and the wider study guide collection are the natural next step.
References and further reading
Use these references to explore the concepts and check the latest information from the relevant organizations.