OMSB Family Medicine (OEFM) common mistakes
OMSB Family Medicine (OEFM) · Understand the traps and how to avoid them.
In an adult vignette built around one abnormal number (TSH 6.4 mIU/L, Hb 10.8 g/dL, Na 129 mmol/L), they choose the option that treats that number directly - starting thyroxine, transfusing, or ordering hypertonic saline - instead of first confirming the value and looking for the single explanation that ties the whole stem together.
Years of 'lab value → drug' pattern drilling create a reflex answer; candidates under-weight transient and non-thyroidal illness, and never check whether the abnormal result is a screening test needing confirmation or a confirmed diagnosis.
What to do instead
Before selecting, ask two questions: is this value confirmed (repeat, add free T4, check the medication list), and does my chosen option explain every finding in the stem? For a TSH of 6.4 with normal free T4, the answer is repeat in 6–8 weeks, not thyroxine; treatment is justified when TSH exceeds about 10 mIU/L, or the patient is symptomatic with positive thyroid antibodies.
They prescribe adult-style doses and volumes to children - a 500 mL fluid bolus for a 12 kg toddler, 1 g paracetamol, an adult adrenaline dose - or give 20 mL/kg boluses to a child in diabetic ketoacidosis or with severe acute malnutrition.
Adult prescribing is over-learned and children get treated as small adults; candidates also do not recall that DKA, cardiac disease and severe malnutrition are the recognised exceptions to the standard 20 mL/kg bolus rule.
What to do instead
Anchor the numbers before answering: 10 mL/kg for DKA, cardiac or trauma-suspected shock; 20 mL/kg for suspected septic shock except in severe acute malnutrition, where it is 15 mL/kg over 1 hour; adrenaline 0.01 mg/kg IM of 1:1000 (maximum 0.5 mg); paracetamol 15 mg/kg; maintenance fluids by the 4-2-1 rule. Sanity-check every paediatric dose against a stated or estimated weight before committing.
Faced with a pregnant woman at 33 weeks with BP 162/110, 3+ proteinuria and a headache, they select 'admit for observation' or 'oral antihypertensive as an outpatient' rather than recognising severe pre-eclampsia and choosing magnesium sulphate plus delivery planning.
The cut-offs that define 'severe' are remembered vaguely, so the candidate treats the blood pressure as the only problem and forgets that symptoms and laboratory values - not the number alone - drive the emergency pathway.
What to do instead
Memorise and apply the severe-feature set: BP ≥160/110, platelets <100, creatinine >90 micromol/L, doubled transaminases, or any symptom (headache, visual disturbance, right upper quadrant pain, clonus). The answer then includes magnesium sulphate 4 g IV loading over 5–10 minutes followed by 1 g/hour, controlled blood pressure reduction with labetalol, hydralazine or nifedipine, antenatal corticosteroids if 24–34 weeks, and planned delivery. Also check whether anti-D immunoglobulin is due after any sensitising event.
In time-critical stems they pick the confirmatory investigation over the immediate action - ordering a chest X-ray for a tension pneumothorax, serum tryptase before adrenaline in anaphylaxis, or troponin before reperfusion in a STEMI - and sometimes pick the right drug with the wrong route or concentration.
Clinic habits favour diagnostic certainty, and in a written vignette the 'most thorough' option feels safest; under time pressure the candidate also stops reading at the drug name and ignores the route and strength in the option.
What to do instead
Rank the options by what happens in the first sixty seconds: airway and high-flow oxygen, IM adrenaline 0.5 mg of 1:1000 into the anterolateral thigh, needle decompression before imaging, primary PCI within 120 minutes of first medical contact. Eliminate any option that delays definitive treatment for confirmation, then re-read the chosen option for dose, route and concentration before finalising.
They send a patient with low back pain to physiotherapy without screening for red flags (saddle anaesthesia, bladder or bowel dysfunction, fever, intravenous drug use, unexplained weight loss, new onset over age 50, trauma, long-term corticosteroid use), or label a hot swollen joint as gout and start colchicine without aspirating.
Most back and joint pain genuinely is benign, so pattern recognition defaults to a musculoskeletal label; red-flag lists surface only when the stem explicitly cues them, and the stem often buries the cue in a single clause.
What to do instead
Run the red-flag screen aloud before choosing any conservative option: cauda equina features mean urgent MRI and same-day surgical referral, not analgesia and review; fever, intravenous drug use or an immunocompromised state means septic arthritis until excluded, with joint aspiration for cell count and crystals before antibiotics. Use the Ottawa rules for ankle and knee imaging and remember that a normal X-ray does not exclude septic arthritis.
They diagnose major depression or a primary psychotic disorder in an older or acutely unwell patient without excluding delirium, and they complete the diagnostic formulation without asking about suicidal ideation, plan, intent, access to means and previous attempts - or they assume a patient with a psychiatric diagnosis automatically lacks capacity.
The stem foregrounds mood or psychotic symptoms and pattern matching goes straight to a primary psychiatric label; risk assessment is treated as a formality rather than the discriminator, and capacity is confused with the presence of mental illness.
What to do instead
Always apply two screens first: acute onset with fluctuating course, inattention and visual hallucinations points to delirium, so look for infection, drugs, hypoxia, pain and metabolic causes before psychiatry. Then ask directly about risk and protective factors and about access to means. Diagnose and assess capacity separately using the four-stage test - understand, retain, weigh, communicate - because a treatable mental illness does not by itself remove decision-making capacity.
They treat a high sensitivity as proof that a positive result means disease (confusing sensitivity with positive predictive value), recommend a screening test without checking the eligible age band and interval, or accept a relative risk reduction as if it were the benefit the individual patient will experience.
Sensitivity, specificity and predictive values blur under time pressure, prevalence is ignored entirely, and preventive-care schedules are memorised as fragments rather than as a coherent programme.
What to do instead
Sketch a 2x2 table with a plausible prevalence before answering; predictive value rises with prevalence, so a '99% sensitive' test in a rare condition still yields mostly false positives. For any screening question apply the Wilson–Jungner gates - detectable latent stage, accepted test, effective early treatment, agreed policy on whom to treat, cost-effective - and check the stated age and interval against the national programme rather than memory.
They treat an annular scaly plaque with a topical corticosteroid before confirming tinea, reassuring the patient when the itch settles, or they dismiss a changing pigmented lesion or a non-healing ulcer as benign without applying a structured checklist and asking about recent change.
Steroids reliably relieve itch and 'look right' for eczema, and the edge, scale and collarette are not examined in the compressed stem; mole assessment is skipped because the candidate relies on overall impression rather than a checklist.
What to do instead
If tinea is possible, take skin scrapings for potassium hydroxide microscopy or culture before any topical steroid, and treat with a topical azole or terbinafine; steroid-modified tinea spreads and becomes harder to diagnose. For any pigmented lesion apply ABCDE plus the ugly duckling sign and ask specifically about change in size, shape, colour, bleeding or itch; biopsy or refer urgently rather than reviewing. Any ulcer or lesion failing to heal within 4–6 weeks needs biopsy.
They attribute new confusion, a fall or functional decline in an older patient to ageing or known dementia, and they add a sedative, anticholinergic or NSAID without reviewing the existing medication list, anticholinergic burden, renal function and falls risk.
Atypical presentation is under-recognised and geriatric problems are multifactorial, so the candidate settles for the first plausible explanation instead of generating a differential; prescribing is inherited from previous clinicians and never audited.
What to do instead
For any older patient with a fall, delirium or decline, screen systematically for infection, infarction, injury (fracture, subdural haematoma), impaired glucose, intoxication or withdrawal, faecal impaction, urinary retention, immobility and drugs. Review every medicine against Beers or STOPP criteria, check doses against creatinine clearance, and offer deprescribing, a multifactorial falls assessment and bone health review rather than adding another drug.
They read a trial result as a headline relative risk reduction (for example 'reduced events by 30%') and select that as the answer without converting to absolute risk, without calculating the number needed to treat, and without checking whether the confidence interval or study design supports the claim.
Pharmaceutical-style phrasing is familiar and feels authoritative, so the candidate never reconstructs the numbers; confidence intervals, intention-to-treat analysis and surrogate endpoints are seen as research-methods detail rather than the core of the question.
What to do instead
Recompute before answering: absolute risk reduction equals control event rate minus intervention event rate, and number needed to treat equals 1 divided by the absolute risk reduction. A 30% relative reduction from 2% to 1.4% gives an absolute reduction of 0.6% and a number needed to treat near 167. Then check whether the confidence interval crosses no effect, whether analysis was by intention to treat, and whether the outcome was patient-important rather than a surrogate marker.