MICGP CKT common mistakes
MICGP CKT · Understand the traps and how to avoid them.
Choosing the single 'most likely' diagnosis and moving on, without first actively excluding the red-flag or 'must-not-miss' alternative (e.g. labelling chest pain as musculoskeletal without considering ACS, or headache as tension-type without considering subarachnoid haemorrhage).
Candidates are trained in pattern recognition and the commonest answer feels satisfying; under time pressure the first plausible diagnosis anchors the whole question and later clues are read as confirmation rather than as disconfirming evidence.
What to do instead
Before committing, run a two-line safety check: 'What is the most likely diagnosis?' and 'What is the one diagnosis I must not miss?' If the stem contains any red flag (exertional component, night symptoms, weight loss, focal neurology, fever, age over 50 with new symptom), the safe-management option usually outranks the commonest-diagnosis option. Practise by tagging every practice question with both answers.
Answering the clinical question that was asked in the stem but ignoring the question actually posed in the final line (e.g. stem describes heart failure but the question asks about the next investigation, the safety-netting advice, or the indication for referral).
Long vignettes consume working memory; the candidate forms a diagnosis early and then selects the option that matches the diagnosis rather than the task verb (diagnose, investigate, manage, refer, explain).
What to do instead
Read the final sentence first, underline the task verb, then read the vignette. Before selecting, restate the task in your own words in one clause. If two options are both true statements, the one that answers the stated task is correct and the other is a distractor.
Managing hypertension, type 2 diabetes or hyperlipidaemia with a single reading or single modality, and forgetting to apply risk stratification, target organ damage assessment, or the correct treatment threshold before choosing therapy.
Candidates memorise drug classes but not the decision architecture (confirm the diagnosis, stage the condition, assess total cardiovascular risk, then treat); stems deliberately give a borderline value to test whether the candidate knows the threshold.
What to do instead
Build a fixed sequence for each metabolic condition: confirm (repeat measurement/ambulatory or HbA1c), stage (grade, albuminuria, eGFR, ECG, fundoscopy), stratify (QRISK-type total risk or existing CVD), then treat to target with a named agent and a review interval. Rehearse the thresholds explicitly rather than the drug names alone.
Treating the older patient's presentation as an inevitable consequence of ageing and not pursuing a treatable cause (e.g. accepting 'acopia' or 'off legs' without screening for infection, dehydration, medication adverse effects, constipation, pain, or depression; missing delirium superimposed on dementia).
Geriatric presentations are atypical and the stem often lacks the classic symptom cluster; candidates default to the 'age-related decline' option because it requires no further thought, and they under-weight polypharmacy and drug withdrawal as causes.
What to do instead
Use a structured screen for any older patient with functional decline: review every medicine against the presentation, check hydration and bowels, screen for infection and pain, assess mood and cognition against baseline, and ask about falls and vision. Treat 'off legs' as a syndrome requiring a cause, not a diagnosis.
Recommending investigation or referral without first establishing whether the presentation is safely manageable in primary care, and without stating the safety-netting advice and review interval that make the decision safe.
Candidates equate thoroughness with referral and fear under-investigating, so they over-refer; equally they under-refer when a red flag is subtle. The exam rewards the option that names both the action and the follow-up plan.
What to do instead
For every management answer, check three elements: appropriate setting (primary care vs same-day vs emergency), explicit safety-netting (what symptom means return immediately), and a review timeline. An option that refers without safety-netting, or safety-nets without acting on a red flag, is usually wrong.
Missing the safeguarding, capacity or confidentiality dimension embedded in an otherwise routine consultation (e.g. a child with an injury inconsistent with the story, a patient with cognitive impairment declining a needed intervention, a request for information about a third party).
The clinical content dominates attention and the ethical layer is one clause in a long stem; candidates also conflate 'capacity to make this decision' with 'capacity in general' and assume a refusal ends the matter.
What to do instead
When a stem involves a child, a cognitively impaired adult, a third party, or a request to withhold information, pause and apply the four-stage capacity test or the safeguarding threshold explicitly. Remember capacity is decision-specific and time-specific; a capacitous refusal is respected, an incapacitous one triggers best-interests decision-making.
Describing a consultation as 'explain and reassure' without naming the specific communication technique, and failing to recognise when the correct answer is to explore the patient's agenda, beliefs or expectations before giving information.
Communication questions feel subjective, so candidates pick the option that sounds kindest rather than the one that matches a recognised model (agenda-setting, chunking and checking, ICE - ideas, concerns, expectations, signposting, summarising).
What to do instead
Learn the named micro-skills and match them to the scenario: opening (agenda-setting), eliciting ICE, chunking-and-checking for information-giving, motivational interviewing for behaviour change, SPIKES-type framing for breaking bad news, and explicit closure with safety-netting. Choose the option that demonstrates the skill rather than the one that merely sounds empathetic.
Confusing the correct sequence of investigation in a presentation where timing matters (e.g. requesting an exercise ECG rather than a same-day admission for suspected unstable angina; ordering an endoscopy before checking for alarm features; requesting imaging before excluding a time-critical diagnosis).
Candidates know the list of available tests but not the urgency hierarchy, and stems often include a plausible but slower test as a distractor.
What to do instead
Rank tests by what changes immediate management: same-day/emergency action first (ECG, troponin pathway, urgent referral), then confirmatory tests, then screening. Ask 'what would I do in the next hour?' before 'what would I do next week?' and choose the option that matches the required timescale.
Misapplying risk-factor or drug-safety knowledge in women's health and paediatrics - e.g. prescribing a contraindicated agent in pregnancy or lactation, or failing to adjust a paediatric dose by weight and age rather than using an adult dose.
Candidates learn adult dosing and adult contraindications as the default and do not routinely switch to a pregnancy/lactation or weight-based frame; stems rarely announce 'this patient is pregnant' in the first line.
What to do instead
Before selecting any drug in a woman of reproductive age, check pregnancy status and contraception, then check the safest agent in pregnancy and lactation. For children, always calculate from weight and check the age-specific formulation and maximum dose. Keep a short list of agents to avoid in pregnancy and of paediatric dose formulae you can reproduce from memory.
Answering governance, ethics or practice-management questions from personal opinion rather than from the framework (GMC-style duties, consent law, audit vs research, incident reporting, appraisal and revalidation, complaints handling), and failing to distinguish a clinical audit from research or a significant event from a complaint.
These domains feel like common sense, so candidates do not revise the definitions; the options are all plausible and differ only in the framework they invoke.
What to do instead
Learn the defining features as discriminators: audit measures practice against a standard and closes the loop; research generates new generalisable knowledge and needs ethics approval; a significant event analysis is confidential and systems-focused; a complaint follows the formal complaints procedure. For ethics, anchor on the four principles plus the duty to be open and honest when things go wrong, and choose the option that documents, discusses and escalates.