e-UKCG common mistakes
e-UKCG · Understand the traps and how to avoid them.
Choosing the option that is the most definitive or specialist answer rather than the safest next step, for example selecting urgent referral or a definitive investigation when the scenario asks what should be done now in a stable primary care patient.
Candidates carry a 'hospital mindset' into general practice items and equate thoroughness with immediate escalation, missing that UK primary care answers reward safety-netting, watchful waiting and appropriate use of primary care pathways.
What to do instead
Before answering, identify the setting (primary, secondary, emergency) and the question stem's time frame ('now', 'next', 'initial management'). Practise ranking options by 'safest appropriate next step in this setting' rather than 'most complete work-up', and review NICE guidance on referral thresholds for common presentations.
Failing to recognise a deteriorating or critically unwell patient because the stem presents subtle or 'normal-looking' observations, and selecting a ward-based or outpatient action instead of a structured resuscitation approach.
Candidates focus on the single abnormal value or the diagnosis rather than applying an ABCDE/resuscitation hierarchy, and under-recognise tracks such as sepsis, anaphylaxis, major haemorrhage and diabetic ketoacidosis presenting atypically.
What to do instead
Drill a fixed ABCDE and 'assess severity first' routine: for every emergency item, decide whether the patient is stable or peri-arrest before considering diagnosis. Learn physiological triggers (NEWS2 thresholds, sepsis criteria, shock index) and rehearse the first three actions for each time-critical presentation.
Making prescribing decisions without checking interactions, renal or hepatic dosing, pregnancy and breastfeeding status, or the patient's age, and choosing a contraindicated drug in a scenario where a safer alternative exists.
Candidates anchor on the 'textbook drug' for the diagnosis and skim past the comorbidity, co-prescription or gestation detail inserted into the stem specifically to rule that drug out.
What to do instead
Build a pre-answer checklist: allergies, interactions, renal/hepatic function, pregnancy and lactation, age extremes, and monitoring requirements. Learn high-yield contraindications (for example NSAIDs in renal impairment and late pregnancy, ACE inhibitors in renal artery stenosis, tetracyclines in children, sodium valproate in women of childbearing potential).
Ignoring the patient's ideas, concerns and expectations and the ICE/consultation-model cues in communication items, instead selecting the option that gives information or a management plan immediately.
Candidates treat communication questions as clinical questions and pick the clinically correct content, missing that the mark is for the process of the consultation and the order of the steps.
What to do instead
For every consultation item, ask 'what stage of the consultation is this?' and choose the option that matches that stage. Learn a consultation framework (for example Calgary-Cambridge or the traditional ICE, explore, explain, plan sequence) and practise identifying the option that explores before it explains.
Breaching confidentiality or consent principles by disclosing information without the patient's agreement, or by failing to recognise the specific legal exceptions where disclosure is justified despite refusal.
Candidates over-apply 'confidentiality is absolute' or, conversely, over-apply 'public interest' to situations that do not meet the threshold, and they confuse capacity, consent and best-interests frameworks.
What to do instead
Learn the GMC consent and confidentiality decision trees and apply them as a checklist: does the patient have capacity, is there a legal obligation to disclose, is there a serious risk to others, is disclosure the minimum necessary? Practise distinguishing notifiable diseases, safeguarding, and fitness-to-drive reporting from ordinary disclosures.
Misinterpreting screening and diagnostic test statistics, particularly confusing sensitivity with positive predictive value, or confusing relative with absolute risk reduction when the stem gives prevalence or baseline risk.
Candidates memorise definitions without practising calculation in context, and stems deliberately supply prevalence to test whether the candidate knows that predictive values depend on it.
What to do instead
Practise constructing a 2x2 table for every statistics stem and calculate sensitivity, specificity, PPV, NPV, likelihood ratios and NNT from it. Learn the rule that PPV rises with prevalence and that relative risk reduction must be converted to absolute risk reduction and NNT before judging clinical importance.
Applying adult medicine criteria to children or pregnant patients, for example using adult vital-sign ranges in a febrile infant, adult drug doses in paediatrics, or missing pregnancy-specific diagnoses such as pre-eclampsia and ectopic pregnancy.
Candidates revise specialties in isolation and do not rehearse age- and pregnancy-specific thresholds, so they default to adult norms under time pressure.
What to do instead
Build age-banded reference ranges for heart rate, respiratory rate and blood pressure, and learn paediatric red flags (for example fever under three months, non-blanching rash, reduced feeding) and obstetric red flags (headache with hypertension after 20 weeks, pain with bleeding in early pregnancy). Practise weight-based paediatric dosing and pregnancy-safe alternatives.
Missing the diagnosis in an 'atypical' or multimorbid presentation because the stem does not contain the classic triad, and choosing a rare diagnosis over a common one without weighing pre-test probability.
Candidates pattern-match to classic textbook vignettes and do not apply probabilistic reasoning, so common conditions presenting atypically (for example depression in an older adult, coeliac disease, hypothyroidism, or myocardial infarction without chest pain) are overlooked.
What to do instead
Practise 'what is the most likely diagnosis in this age and setting?' before 'what is the most serious?'. Learn atypical presentations of high-frequency conditions and apply Occam's and probabilistic reasoning explicitly; when two options are plausible, choose the one with higher prevalence unless a red flag mandates excluding the dangerous one.
Answering ethics and professionalism items with a personal moral view or a rule applied rigidly, rather than following the professional framework, and failing to recognise when to escalate to a senior, a safeguarding lead or a regulatory body.
Candidates rely on intuition and general ethical principles without knowing the specific GMC guidance and statutory duties, and they are unsure when a concern crosses the threshold for formal reporting.
What to do instead
Learn the GMC's Good Medical Practice domains and the specific duties around candour, raising concerns, consent, and boundaries. For each ethics item, identify the relevant duty, the least restrictive option consistent with it, and the correct escalation route; practise scenarios on duty of candour, whistleblowing and impaired colleagues.
Ignoring the quality-improvement, evidence-appraisal or patient-safety lens in items that ask about system failures, audit, guideline implementation or incident analysis, and instead answering with an individual clinical action.
Candidates read these items as clinical vignettes and miss the cue words (audit, significant event, root cause, guideline, implementation), so they do not apply QI or safety frameworks.
What to do instead
Learn to distinguish audit from research, and to select the correct QI tool (PDSA cycle, process mapping, significant event analysis, root cause analysis, morbidity and mortality review) for the question. Practise identifying the level of the intervention (individual, team, system) and matching the answer to it.